The Health Consequences of Smoking 50 Years of Progress: A Report of the Surgeon General

Transcript

1 The Health Consequences of Smoking—50 Years of Progress A Report of the Surgeon General U.S. Department of Health and Human Services

2 The Health Consequences of Smoking —50 Years of Progress A Report of the Surgeon General 2014 U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Public Health Service Office of the Surgeon General Rockville, MD

3 National Library of Medicine Cataloging in Publication The health consequences of smoking – 50 years of progress: a report of the Surgeon General. – Atlanta, GA. : U.S. Department of Health and Human Services, Centers for Disease Control and Preven- tion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2014. p. 944 Includes bibliographical references. Smoking – adverse effects. 2. Tobacco Use Disorder – complications. 3. Smoking – preven - 1. tion & control. 4. Health Promotion – trends. I. United States. Public Health Service. Office of the Surgeon General. II. National Center for Chronic Disease Prevention and Health Promotion (U.S.). Office on Smoking and Health. QV 137 U.S. Department of Health and Human Services Centers for Disease Control and Prevention National Center for Chronic Disease Prevention and Health Promotion Office on Smoking and Health For more information . For more information about the Surgeon General’s report, visit www.surgeongeneral.gov To download copies of this document, go to www.cdc.gov/tobacco. To order copies of this document, go to www .cdc.gov/tobacco and click on Publications Catalog or call 1-800-CDC-INFO (1-800-232-4636); TTY: 1-888-232-6348. Suggested Citation U.S. Department of Health and Human Services. The Health Consequences of Smoking: 50 Years of Progress. A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2014. Printed with corrections, January 2014. Use of trade names is for identification only and does not constitute endorsement by the U.S. Department of Health and Human Services.

4 Message from Kathleen Sebelius Secretary of Health and Human Services Fifty years after the release of the first Surgeon General’s report warning of the health hazards of smoking, we have learned how to end the tobacco epidemic. Over the past five decades, scientists, researchers and policy makers have determined what works, and what steps must be taken if we truly want to bring to a close one of our nation’s most tragic battles—one that has killed ten times the num - ber of Americans who died in all of our nation’s wars combined. In the United States, successes in tobacco control have more than halved smoking rates since the 1964 landmark Surgeon General’s report came out. Americans’ collective view of smoking has been transformed from an accepted national pastime to a discouraged threat to individual and public health. Strong policies have largely driven cigarette smoking out of public view and public air space. Thanks to smokefree laws, no longer is smoking allowed on airplanes or in a growing number of restaurants, bars, college campuses and government buildings. Evidence in this new report shows tobacco’s continued, immense burden to our nation—and how essential ending the tobacco epidemic is to our work to increase the life expectancy and quality of life of all Americans. This year alone, nearly one-half million adults will still die prematurely because of smok - ing. Annually, the total economic costs due to tobacco are now over $289 billion. And if we continue on our current trajectory, 5.6 million children alive today who are younger than 18 years of age will die prematurely as a result of smoking. I believe that we can make the next generation tobacco-free. And I am extremely proud of the Children’s Health Insurance Obama Administration’s tobacco-control record. For example, the 2009 Program Reauthorization Act included an unprecedented $0.62 tax increase that raised the federal excise tax to $1.01 per pack of cigarettes; we know that increasing the cost of cigarettes is one of the most powerful interventions we can make to prevent smoking and reduce prevalence. Building on this knowledge, the President’s Fiscal Year 2014 Budget includes a $0.94 per pack Federal tobacco tax increase. For the first time in history, the 2009 Family Smoking Prevention and Tobacco Control Act Tobacco Control Act) g ave the U.S. Food and Drug Administration comprehensive authority to regu - ( late tobacco products, which will play a critical role in reducing the harm caused by these products. The Tobacco Control Act also provided for user fees to be paid by tobacco manufacturers that can sup - port sustained public education media campaigns targeting youth prevention and cessation. The 2010 Affordable Care Act (ACA) expands access to smoking cessation services and now requires most insur - ance companies to cover cessation treatments. The Public Health and Prevention Affordable Care Act’s Fund is supporting innovative and effective community-based programs as well as public education campaigns promoting prevention and helping people to quit. All of these tobacco control interventions are known to reduce tobacco use and, as a result, tobacco’s extraordinary toll of death and disease. But in order to free the next generation from these burdens, we must redouble our tobacco control efforts and enlist nongovernmental partners—and society as a whole—to share in this responsibility. Ending the devastation of tobacco-related illness and death is not in the jurisdiction of any one entity. We must all share in this most worthwhile effort to end the tobacco epidemic.

5

6 Message from Howard Koh Assistant Secretary for Health The nation stands poised at the crossroads of tobacco control. On one hand, we can celebrate tremendous progress 50 years after the landmark 1964 Surgeon General’s report: . Smoking and Health Adult smoking rates have fallen from about 43% (1965) to about 18% today. Mortality rates from lung cancer, the leading cause of cancer death in this country, are declining. Most smokers visiting health care settings are now routinely asked and advised about tobacco use. On the other hand, cigarette smoking remains the chief preventable killer in America, with more than 40 million Americans caught in a web of tobacco dependence. Each day, more than 3,200 youth (younger than 18 years of age) smoke their first cigarette and another 2,100 youth and young adults who are occasional smokers progress to become daily smokers. Furthermore, the range of emerging tobacco products complicates the current public health landscape. In this context, the 50th Anniversary of the Surgeon General’s report prompts us to pause and ask why this addiction persists when proven interventions can eliminate it. Of great concern, too many in our nation assume that past success in tobacco control guarantees future progress; nothing can be further from the truth. To rejuvenate and reinvigorate national efforts, in 2010, the U.S. Department of Health and Human Services unveiled its first ever strategic plan for tobacco control. Ending the Tobacco Epidemic: A Tobacco Control Strategic Action Plan provides a critical framework to guide efforts to rapidly drop prevalence rates of smoking among youth and adults. A major foundation and pillar of the plan is to encourage and promote leadership throughout all sectors of society. Now, this current 2014 Surgeon General’s report can accelerate that leadership to fully implement the life-saving prevention that can make the next generation free of tobacco-related death and disease. We have many tools that we know work. A comprehensive public policy approach emphasizing mass media campaigns to encourage prevention and quit attempts, smokefree policies, restrictions on youth access to tobacco products, and price increases can collectively drive further meaningful reduc - tions in tobacco use. Furthermore, we can accelerate progress through full commitment to clinical and public health advances; including the widespread use of telephone quit lines and science-based counseling and medications for tobacco users. Promoting progress today also requires recognizing that tobacco use has evolved from being an equal-opportunity killer to one threatening the most vulnerable members of our society. We must confront, and reverse, the tragically higher tobacco use rates that threaten persons of low socioeconomic status, sexual minorities, high school dropouts, some racial/ ethnic minority groups, and those living with mental illness and substance use disorders. Of all the accomplishments of the 20th century, historians rank the 1964 Surgeon General’s report as one of the seminal public health achievements of our time. Armed with both science and resolve, we can continue to honor the legacy of the report by completing the work it began in the last century. The current 2014 Surgeon General’s report represents a national vision for getting the job done. With strategy, commitment, and action, our nation can leave the crossroads and move forward to end the tobacco epidemic once and for all.

7

8 Foreword Fifty years have passed since publication of the landmark report of the Surgeon General’s Advisory Committee on smoking and health. This report highlights both the dramatic progress our nation has made reducing tobacco use and the continuing burden of disease and death caused by smoking. As a physician, when I think about smoking, I recall the patients I have cared for. The man who had a leg amputated. The woman who had to gasp for every single breath that she took. The man with heart disease who hoped to see his son graduate, but didn’t live long enough to do so. That’s the reality of smoking that health care providers see every day. The prevalence of current cigarette smoking among adults has declined from 42% in 1965 to 18% in 2012. However, more than 42 million Americans still smoke. Tobacco has killed more than 20 million people prematurely since the first Surgeon General’s report in 1964. The findings in this report show that the decline in the prevalence of smoking has slowed in recent years and that burden of smoking- attributable mortality is expected to remain at high and unacceptable levels for decades to come unless urgent action is taken. Recent surveys monitoring trends in tobacco use indicate that more people are using multiple tobacco products, particularly youth and young adults. The percentage of U.S. middle and high school students who use electronic, or e-cigarettes, more than doubled between 2011 and 2012. We need to monitor patterns of use of an increasingly wide array of tobacco products across all of the diverse seg- ments of our society, particularly because the tobacco industry continues to introduce and market new products that establish and maintain nicotine addiction. Tobacco control efforts need to not only address the general population, but also to focus on populations with a higher prevalence of tobacco use and lower rates of quitting. These populations include people from some racial/ethnic minority groups, people with mental illness, lower educational levels and socioeconomic status, and certain regions of the country. We now have proven interventions and policies to reduce tobacco initiation and use among youth and adults. With intense use of proven interventions, we can save lives and reduce health care costs. In 2012, the Centers for Disease Control and Prevention (CDC) launched the first-ever paid national tobacco education campaign — Tips From Former Smokers Tips ) — to raise awareness of the harms to health ( caused by smoking, encourage smokers to quit, and encourage nonsmokers to protect themselves and their families from exposure to secondhand smoke. It pulled back the curtain in a way that numbers alone cannot, and showed the tobacco-caused tragedies that we as health care professionals see and are saddened by every day. As a result of this campaign, an estimated 1.6 million smokers made an - attempt to quit and, based on a conservative estimate, at least 100,000 smokers quit for good. Addition ally, millions of nonsmokers talked with friends and family about the dangers of smoking and referred smokers to quit services. In 2013, CDC launched a new round of advertisements that helped even more people quit smoking by highlighting the toll that smoking-related illnesses take on smokers and their loved ones. CDC has also established reducing tobacco use as one of its “Winnable Battles.” These are public health priorities with large-scale impact on health that have proven effective strategies to address them. CDC believes that with additional effort and support for evidence-based, cost-effective policy and pro- gram strategies to reduce tobacco use, we can reduce smoking substantially, prevent millions of people from being killed by tobacco, and protect future generations from smoking. i

9 While we have made tremendous progress over the past 50 years, sustained and comprehensive efforts are needed to prevent more people from having to suffer the pain, disability, disfigurement, and death that smoking causes. Most Americans who have ever smoked have already quit, and most smokers who still smoke want to quit. If we continue to implement tobacco prevention and cessation strategies that have proven effective in reducing tobacco use, people throughout our country will live longer, healthier, more productive lives. Thomas R. Frieden, M.D., M.P.H. Director Centers for Disease Control and Prevention ii

10 Preface from the Acting Surgeon General, U.S. Department of Health and Human Services On January 11, 1964, Luther L. Terry, M.D., the 9th Surgeon General of the United States, released Smoking and Health: Report of the Advisory the first report on the health consequences of smoking: Committee of the Surgeon General of the Public Health Service. That report marked a major step to reduce the adverse impact of tobacco use on health worldwide. Over the past 50 years, 31 Surgeon General’s reports have utilized the best available evidence to expand our understanding of the health consequences of smoking and involuntary exposure to tobacco smoke. The conclusions from these reports have evolved from a few causal associations in 1964 to a robust body of evidence documenting the health consequences from both active smoking and exposure to secondhand smoke across a range of diseases and organ systems. The 2004 report concluded that smoking affects nearly every organ of the body, and the evidence in this report provides even more support for that finding. A half century after the release of the first report, we continue to add to the long list of diseases caused by tobacco use and exposure to tobacco smoke. This report finds that active smoking is now causally associated with age-related macular degeneration, diabetes, colorectal cancer, liver cancer, adverse health outcomes in cancer patients and survivors, tuberculosis, erectile dysfunction, orofacial clefts in infants, ectopic pregnancy, rheumatoid arthritis, inflammation, and impaired immune function. In addition, exposure to secondhand smoke has now been causally associated with an increased risk for stroke. Smoking remains the leading preventable cause of premature disease and death in the United States. The science contained in this and prior Surgeon General’s reports provide all the information we need to save future generations from the burden of premature disease caused by tobacco use . How- ever, evidence-based interventions that encourage quitting and prevent youth smoking continue to be underutilized. This report strengthens our resolve to work together to accelerate and sustain what works—such as hard-hitting media campaigns, smokefree air policies, optimal tobacco excise taxes, barrier-free cessation treatment, and comprehensive statewide tobacco control programs funded at CDC-recommended levels. At the same time, we will explore “end game” strategies that support the goal of eliminating tobacco smoking, including greater restrictions on sales. It is my sincere hope that 50 years from now we won’t need another Surgeon General’s report on smoking and health, because tobacco-related disease and death will be a thing of the past. Working together, we can make that vision a reality. Boris D. Lushniak, M.D., M.P.H. Rear Admiral, U.S. Public Health Service Acting Surgeon General U.S. Department of Health and Human Services iii

11

12 —50 Years of Progress The Health Consequences of Smoking Acknowledgments Leslie A. Norman, M.B.A., Managing Editor, Office on This report was prepared by the U.S. Department of Health Smoking and Health, National Center for Chronic Disease and Human Services under the general direction of the Prevention and Health Promotion, Centers for Disease - Centers for Disease Control and Prevention, National Cen Control and Prevention, Atlanta, Georgia. ter for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. Peter L. Taylor, M.B.A., Technical Editor, Fairfax, Virginia. Rear Admiral (RADM) Boris D. Lushniak, M.D., M.P.H. , Acting Surgeon General, U.S. Public Health Service, Contributing authors were Office of the Assistant Secretary for Health, Office of the Surgeon General, Office of the Secretary, U.S. Department David B. Abrams, Ph.D., Executive Director, The Schroeder of Health and Human Services, Washington, D.C. Institute for Tobacco Research and Policy Studies, Pro- fessor, Department of Health, Behavior and Society, The Howard K. Koh, M.D., M.P.H., Assistant Secretary for Johns Hopkins Bloomberg School of Public Health, Bal - Health, U.S. Department of Health and Human Services, timore, Maryland, and Professor of Oncology (adjunct), Washington, D.C. Georgetown University Medical Center, Lombardi Com - prehensive Cancer Center, Washington, D.C. Thomas R. Frieden, M.D., M.P.H., Director, Centers for Disease Control and Prevention, Atlanta, Georgia. Anthony J. Alberg, Ph.D., M.P.H., Blatt Ness Professor, - Department of Public Health Sciences, Associate Direc Ursula E. Bauer, Ph.D., M.P.H., Director, National Cen - tor of Cancer Prevention and Control, Hollings Cancer ter for Chronic Disease Prevention and Health Promo- Center, Medical University of South Carolina, Charleston, tion, Centers for Disease Control and Prevention, Atlanta, South Carolina. Georgia. Gary P. Anderson, Ph.D., Director, Centre for Lung Health Samuel F. Posner, Ph.D., Associate Director for Science Research, Department of Pharmacology and Therapeutics Preventing Chronic Disease: (Acting), Editor in Chief, and Department of Medicine, Faculty of Medicine, Den- , National Public Health Research, Practice and Policy tistry and Health Sciences, The University of Melbourne, - Center for Chronic Disease Prevention and Health Promo Australia. tion, Centers for Disease Control and Prevention, Atlanta, Georgia. Brent M. Ardaugh, M.P.H., M.S., Science Writer, Framing - ham Heart Study, Boston University School of Medicine, Tim A. McAfee, M.D., M.P.H., Director, Office on Smoking Boston, Massachusetts. and Health, National Center for Chronic Disease Preven - tion and Health Promotion, Centers for Disease Control Kathy B. Baumgartner, Ph.D., Professor and Associate and Prevention, Atlanta, Georgia. Dean for Faculty Affairs, Department of Epidemiology and - Population Health, School of Public Health and Infor mation Sciences, The Henry Vogt Professorship in Can- The editors of the report were cer Prevention and Control, James Graham Brown Cancer Center, University of Louisville, Louisville, Kentucky. Jonathan M. Samet, M.D., M.S., Senior Scientific Editor, Professor and Flora L. Thornton Chair, Department of Richard N. Baumgartner, Ph.D., Professor and Chair, Preventive Medicine, Keck School of Medicine; Director, Department of Epidemiology and Population Health, Institute for Global Health, University of Southern Cali- School of Public Health and Information Sciences, Uni - fornia, Los Angeles, California. versity of Louisville, Louisville, Kentucky. Terry F. Pechacek, Ph.D., Senior Associate Editor, Deputy - Neal L. Benowitz, M.D., Professor of Medicine, Bioengi Director for Research Translation, Office on Smoking and neering, and Therapeutic Sciences, Chief, Division of Health, National Center for Chronic Disease Prevention - Clinical Pharmacology, University of California, San Fran and Health Promotion, Centers for Disease Control and cisco, California. Prevention, Atlanta, Georgia. v

13 Surgeon General’s Report MeLisa R. Creamer, M.P.H., Ph.D. Candidate and Michael Rebecca H. Bitsko, Ph.D., Health Scientist, Division of - Human Development and Disability, National Center on and Susan Dell Health Scholar, Division of Epidemiol ogy, University of Texas School of Public Health, Austin Birth Defects and Developmental Disabilities, Centers for Regional Campus, Michael and Susan Dell Center for Disease Control and Prevention, Atlanta, Georgia. Healthy Living, Austin, Texas. Georgiy V. Bobashev, Ph.D., Senior Research Statistician, Mary E. Cronin, M.D., Rheumatology Fellowship Program Research Triangle Institute, Durham, North Carolina. - Director, Professor of Medicine, Medical College of Wis consin, Milwaukee, Wisconsin. Kari Bohlke, Sc.D., Consultant Epidemiologist, Alvin J. Siteman Cancer Center, Washington University School of K. Michael Cummings, Ph.D., M.P.H., Co-director, Tobacco , St. Louis, Missouri. Medicine in St. Louis Research Program, Hollings Cancer Center and Professor, Department of Psychiatry and Behavioral Sciences, Medi Ross C. Brownson, Ph.D., Professor and Co-Director, - - cal University of South Carolina, Charleston, South Caro Prevention Research Center in St. Louis, Brown School; Division of Public Health Sciences and Alvin J. Siteman lina. - Cancer Center, School of Medicine, Washington Univer Nancy E. Davidson, M.D., Director, University of Pitts sity in St. Louis, St. Louis, Missouri. - burgh Cancer Institute and University of Pittsburgh Medi- cal Center, Associate Vice Chancellor for Cancer Research, Arthur L. Burnett, M.D., M.B.A., F.A.C.S., Patrick C. Walsh Hillman Professor of Oncology, Distinguished Professor of Distinguished Professor of Urology, Cellular and Molecu- Medicine and Pharmacology and Chemical Biology, Uni- lar Medicine, Director, Sexual Medicine Division, Depart - versity of Pittsburgh, Pittsburgh, Pennsylvania. ment of Urology, Johns Hopkins University, Baltimore, Maryland. Richard A. Daynard, J.D., Ph.D., President, Public Health Advocacy Institute, University Distinguished Professor of David M. Burns M.D., Professor Emeritus, Department of Law, Northeastern University, Boston, Massachusetts. Family and Preventive Medicine, University of California San Diego, School of Medicine, San Diego, California. Cristine D. Delnevo, Ph.D., M.P.H., Director, Center for Tobacco Surveillance and Evaluation Research, Professor - Angelika H. Claussen, Ph.D., Research Psychologist, Divi and Chair, Department of Health Education/Behavioral sion of Human Development and Disabilities, National Science, Rutgers School of Public Health, New Bruns- Center on Birth Defects and Developmental Disabilities, wick, New Jersey. Centers for Disease Control and Prevention, Atlanta, Georgia. - Dawn L. DeMeo, M.D., M.P.H., Channing Division of Net work Medicine, Associate Professor of Medicine, Harvard Graham A. Colditz, M.D., Dr. P.H., Fellow of the Australian Faculty of Public Health Medicine, Niess-Gain Professor - Medical School, Associate Physician, Division of Pulmo and Chief, Division of Public Health Sciences, Department nary and Critical Care, Brigham and Women’s Hospital, Boston, Massachusetts. of Surgery, Washington University School of Medicine, and Associate Director, Prevention and Control, Siteman Cancer Center, St. Louis, Missouri. Mark D. Eisner, M.D., M.P.H., Associate Professor of Medicine and Anesthesia, Division of Occupational and Environmental Medicine and Division of Pulmonary and David Coultas, M.D., Vice President for Clinical and Aca - - - Critical Care Medicine, Department of Medicine, Univer demic Affairs, Professor and Chair, Department of Medi cine, The University of Texas Health Science Center, Tyler sity of California, San Francisco, California. Texas. Lucinda England, M.D., M.S.P.H., Medical Epidemiolo - James D. Crapo M.D., Professor of Medicine, National Jew- gist, Office on Smoking and Health, National Center for ish Health and University of Colorado, Denver, Colorado. Chronic Disease and Health Promotion, Centers for Dis - ease Control and Prevention, Atlanta, Georgia. vi

14 The Health Consequences of Smoking —50 Years of Progress Rachel Kaufmann, M.P.H., Ph.D., Associate Director for Jonathan Golub, Ph.D, M.P.H., Associate Professor, Medi- - Science, Center for Surveillance, Epidemiology, and Labo cine, Epidemiology and International Health, Center for - ratory Services, Centers for Disease Control and Preven - Tuberculosis Research, Johns Hopkins University, Balti more, Maryland. tion, Atlanta, Georgia. Brian A. King, Ph.D., M.P.H., Epidemiologist, Office on Mark Gottlieb, J.D., Executive Director, Public Health Smoking and Health, National Center for Chronic Disease Advocacy Institute, Northeastern University School of Prevention and Health Promotion, Centers for Disease Law, Boston, Massachusetts. Control and Prevention, Atlanta, Georgia. - Stephen S. Hecht, Ph.D., Wallin Professor of Cancer Pre vention, Masonic Cancer Center, University of Minnesota, Gregory D. Kirk, M.D., M.P.H., Ph.D., Associate Professor of Epidemiology, Medicine and Oncology, Johns Hopkins Minneapolis, Minnesota. University, Baltimore, Maryland. Jack E. Henningfield, Ph.D., Vice President, Research and Kim Klausner, Manager, Industry Documents Digital Health Policy, Pinney Associates, Bethesda, Maryland, and Libraries, University of California San Francisco, San Professor, Adjunct, Behavioral Biology, Department of Francisco, California . Psychiatry and Behavioral Sciences, The Johns Hopkins University School of Medicine, Baltimore, Maryland. Robert C. Klesges, Ph.D., Professor, Department of Pre- ventive Medicine, The University of Tennessee Health Sci - Theodore R. Holford, Ph.D., Professor, Department of Bio - statistics, Yale School of Public Health, New Haven, Con - ence Center, Memphis, Tennessee. necticut. Darwin R. Labarthe, M.D., M.P.H., Ph.D., F.A.H.A., Pro- Stephen T. Holgate M.D., Professor, Fellow of the Academy fessor of Preventive Medicine, Department of Preventive of Medical Sciences, Medical Research Council, Clinical Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois. Professor of Immunopharmacology, University of South- ampton, United Kingdom. Anne Landman, B.A., Landman Research and Communi- John W. Holloway, Ph.D., Professor of Allergy and Respira- cations, Grand Junction, Colorado. tory Genetics, Faculty of Medicine, University of South- David T. Levy, Ph.D., Professor, Cancer Control, Lombardi ampton, United Kingdom. Comprehensive Cancer Center, Georgetown University, David M. Homa, Ph.D., M.P.H., Epidemiologist, Office on Washington, D.C. Smoking and Health, National Center for Chronic Disease Carolyn A. Lullo, Ph.D., M.S., Public Health Analyst, Divi - Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia. sion of Human Development and Disability, National Center on Birth Defects and Developmental Disabilities, Margaret A. Honein, Ph.D., M.P.H., Chief, Birth Defects Centers for Disease Control and Prevention, Atlanta, - Branch, National Center on Birth Defects and Develop Georgia. mental Disabilities, Centers for Disease Control and Pre- vention, Atlanta, Georgia. Patricia L. Mabry, Ph.D., Senior Advisor and Acting Deputy Director, Office of Behavioral and Social Sciences - Frank B. Hu, M.D., Ph.D., Professor of Nutrition and Epi Research, Office of the Director, National Institutes of demiology, Harvard School of Public Health; Professor Health, Bethesda, Maryland. of Medicine, Harvard Medical School and Department of Medicine, Brigham and Women’s Hospital, Boston, Mas - David M. Mannino, M.D., Professor and Chair, Department of Preventive Medicine and Environmental Health, Direc sachusetts. - tor, Pulmonary Epidemiology Research Laboratory, Uni - versity of Kentucky College of Public Health, Lexington, Susan Hutfless, Ph.D., Assistant Professor, Division of - Kentucky. Gastroenterology and Hepatology, Department of Medi cine, Johns Hopkins University, Baltimore, Maryland. vii

15 Surgeon General’s Report Paul Melstrom, Phar.D., Ph.D., Toxicologist/Epidemiolo- Ann K. Rosenthal, M.D., Will and Cava Ross Professor of gist, Office on Smoking and Health, National Center for Medicine, Chief of Rheumatology and Vice Chair, Depart- - Chronic Disease Prevention and Health Promotion, Cen ment of Medicine, Medical College of Wisconsin, Milwau - ters for Disease Control and Prevention, Atlanta, Georgia. kee, Wisconsin. - David Méndez, Ph.D., Associate Professor, Health Manage- Hossein Sadeghi-Nejad, M.D., F.A.C.S., Professor of Sur ment and Policy, School of Public Health, University of gery, Division of Urology, Rutgers New Jersey Medical Michigan, Ann Arbor, Michigan. School, Chief of Urology, Virginia-New Jersey Health Care System; Director, Center for Male Reproductive Medicine, Hackensack University Medical Center, Hackensack, New Nathaniel D. Osgood, Ph.D., Director, Computational Epidemiology and Public Health Informatics Laboratory, Jersey. Associate Professor, Department of Computer Science, Associate Faculty, Department of Community Health and Jonathan M. Samet, M.D., M.S., Professor and Flora L. Thornton Chair, Department of Preventive Medicine, Keck Epidemiology, University of Saskatchewan, Saskatoon, Canada. School of Medicine; Director, Institute for Global Health, University of Southern California, Los Angeles, California. Mark Parascandola, Ph.D., M.P.H., Epidemiologist, Tobacco Control Research Branch, Behavioral Research Elizabeth Selvin, Ph.D., M.P.H., Associate Professor of Epidemiology and Medicine, Welch Center for Preven - Program, Division of Cancer Control and Population Sci - ences, National Cancer Institute, Bethesda, Maryland. tion, Epidemiology and Clinical Research, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland. Terry F. Pechacek, Ph.D., Deputy Director for Research Steven D. Shapiro, M.D., Executive Vice President and Translation, Office on Smoking and Health, National Cen- Chief Medical and Scientific Officer, University of Pitts - ter for Chronic Disease Prevention and Health Promo- burgh Medical Center; President, Physician Services tion, Centers for Disease Control and Prevention, Atlanta, Georgia. Division; Professor of Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania. Cheryl L. Perry, Ph.D., Professor and Regional Dean, Uni - versity of Texas School of Public Health, Austin Regional Peter G. Shields, M.D., Deputy Director, Comprehensive Cancer Center, Professor, College of Medicine, James Can - Campus, Michael and Susan Dell Center for Healthy Liv - cer Hospital, The Ohio State University Wexner Medical ing, Austin, Texas. Center, Columbus, Ohio. Keydra L. Phillips, M.Sc., Health Scientist, Division of Donald R. Shopland, U.S. Public Health Service (retired), Birth Defects and Developmental Disabilities, National Ringgold, Georgia. Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Atlanta, David Sidransky, M.D., Director, Head and Neck Cancer Georgia. Research Division, Professor in Otolaryngology, Cellular Elizabeth A. Platz, Sc.D., M.P.H., Professor, Department of and Molecular Medicine, The Johns Hopkins University School of Medicine, Baltimore, Maryland. Epidemiology, Johns Hopkins Bloomberg School of Pub - lic Health and the Sidney Kimmel Comprehensive Cancer Edwin K. Silverman, M.D., Ph.D., Associate Professor of Center at Johns Hopkins University, Baltimore, Maryland. Medicine, Harvard Medical School, Associate Physician, Brigham and Women’s Hospital, Boston, Massachusetts. - Jennita Reefhuis, Ph.D., Research Health Scientist, Divi sion of Reproductive Health, National Center for Chronic - Disease Prevention and Health Promotion, Centers for Martin R. Stampfli, Ph.D., Professor, Department of Pathol ogy and Molecular Medicine, McMaster Immunology Disease Control and Prevention, Atlanta, Georgia. Research Centre, and Department of Medicine, Firestone Institute of Respiratory Health at St. Joseph’s Healthcare, Nancy A. Rigotti, M.D., Director, Tobacco Research and McMaster University, Hamilton, Ontario, Canada. Treatment Center and Associate Chief, General Medicine Division, Department of Medicine, Massachusetts General Hospital; Professor of Medicine, Harvard Medical School, Boston, Massachusetts. viii

16 —50 Years of Progress The Health Consequences of Smoking Kala Visvanathan, M.B.B.S., F.R.A.C.P., M.H.S., Associ - Melissa H. Stigler, Ph.D., M.P.H., Assistant Professor, The ate Professor, Department of Epidemiology, Bloomberg University of Texas Health Science Center at Houston, School of Public Health, Johns Hopkins University, Bal - School of Public Health, Austin Regional Campus, Michael and Susan Dell Center for Healthy Living, Austin, Texas. timore, Maryland. Warner, Ph.D., Avedis Donabedian Distin Edward L. Sweda, Jr., J.D., Senior Attorney, Public Health Kenneth E. - guished University Professor of Public Health, Depart - Advocacy Institute, Northeastern University School of ment of Health Management and Policy, School of Public Law, Boston, Massachusetts. Health, University of Michigan, Ann Arbor, Michigan. Rachel Taketa, Library Specialist, Industry Documents Graham Warren M.D., Ph.D., Vice Chairman for Research Digital Libraries, Library and Center for Knowledge Man - agement, University of California, San Francisco, Califor - in Radiation Oncology, Associate Professor, Department nia. of Radiation Oncology, Department of Cell and Molecular Pharmacology, Hollings Cancer Center, Medical Univer - Gerald W. Talcott, Ph.D., ABPP, Col USAF (ret), Director of sity of South Carolina, Charleston, South Carolina. - Military Research, Associate Professor, Department of Pre Sheila West, Ph.D., El Maghraby Professor, Vice Chair ventive Medicine, University of Tennessee Health Science for Research, Wilmer Eye Institute, School of Medicine, Center, Memphis, Tennessee. Johns Hopkins University, Baltimore, Maryland. Leslie A. Thompson, M.D., Division of Rheumatology, Heather Wipfli, Ph.D., Associate Director, Institute for Department of Medicine, Medical College of Wisconsin, Global Health, Assistant Professor, Department of Preven - Milwaukee, Wisconsin. tive Medicine and School of International Relations, Keck - Michael J. Thun, M.D., Vice President Emeritus, Epidemi School of Medicine, University of Southern California, Los ology and Surveillance Research, American Cancer Soci- Angeles, California. ety, Atlanta, Georgia. Xin Xu, Ph.D., Senior Economist, Office on Smoking and - Hilary A. Tindle, M.D., M.P.H, Assistant Professor of Medi Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and cine, Epidemiology, and Clinical and Translational Sci- Prevention, Atlanta, Georgia. ence, University of Pittsburgh, Pittsburgh, Pennsylvania. - Mitchell Zeller, J.D., Senior Vice President, Pinney Associ - Scott L. Tomar, D.M.D., M.P.H., Dr. P.H., Professor, Depart ment of Community Dentistry and Behavioral Science, ates, Bethesda, Maryland. University of Florida College of Dentistry, Gainesville, Florida. Reviewers were Van T. Tong, M.P.H., Epidemiologist, Division of Reproduc- Executive Director, The David B. Abrams, Ph.D., - tive Health, National Center for Chronic Disease Preven Schroeder Institute for Tobacco Research and Policy tion and Health Promotion, Centers for Disease Control of Health, Behavior and Studies, Professor, Department and Prevention, Atlanta, Georgia. Society, The Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, and Professor of Oncology Andrea C. Villanti, Ph.D., M.P.H., Associate Director for - (adjunct), Georgetown University Medical Center, Lom Regulatory Science and Policy, The Schroeder Institute bardi Comprehensive Cancer Center, Washington, D.C. for Tobacco Research and Policy Studies, American Legacy FoundationS, Washington, D.C. Anthony J. Alberg, Ph.D., M.P.H., Blatt Ness Professor, - Department of Public Health Sciences, Associate Direc Susanna N. Visser, M.S., Lead Epidemiologist, Division of tor of Cancer Prevention and Control, Hollings Cancer Human Development and Disability, National Center on Center, Medical University of South Carolina, Charleston, Birth Defects and Developmental Disabilities, Centers for South Carolina. Disease Control and Prevention, Atlanta, Georgia. ix

17 Surgeon General’s Report Ashwin N. Ananthakrishnan, M.D., M.P.H, Assistant Pro Thomas H. Brandon, Ph.D., Director, Tobacco Research - fessor of Medicine, Division of Gastroenterology, Massa and Intervention Program, and Chair, Department of - chusetts General Hospital and Harvard Medical School, Health Outcomes and Behavior, Moffitt Cancer Center; Professor of Psychology and Oncologic Sciences, Univer - Boston, Massachusetts. sity of South Florida, Tampa, Florida. Benjamin Apelberg, Ph.D., M.H.S., Chief, Epidemiology - Otis W. Brawley, M.D., Chief Medical and Scientific Offi Branch, Division of Population Health Sciences, Office of cer, American Cancer Society; Professor, Departments of Science, Center for Tobacco Products, U.S. Food and Drug Hematology and Oncology, Medicine and Epidemiology, Administration, Rockville, Maryland. Emory University, Atlanta, Georgia. Erik M. Augustson, Ph.D., M.P.H., Behavioral Scientist, Peter A. Briss, M.D., M.P.H., Medical Director, National Tobacco Control Research Branch, Behavioral Research Center for Chronic Disease Prevention and Health Promo Program, Division of Cancer Control and Population Sci - - tion, Centers for Disease Control and Prevention, Atlanta, ences, National Cancer Institute, Rockville, Maryland. Georgia. Olalekan A. Ayo-Yusuf, D.D.S., M.P.H., Ph.D., Head of Gerald Brock, M.D., Professor of Surgery, Urology Pro Clinical Unit and Associate Professor, Department of Com- - munity Dentistry, University of Pretoria, Pretoria, South gram Director, The University of Western Ontario, Vice President of Education, Canadian Urological Association, Africa. London, Ontario, Canada. Cathy L. Backinger, Ph.D., M.P.H., Deputy Director for Paul A. Bunn, Jr., M.D., James Dudley Chair in Cancer Research, Office of Science, Center for Tobacco Products, U.S. Food and Drug Administration, Rockville, Maryland. Research, University of Colorado Cancer Center, Aurora, Colorado. John A. Baron, M.D., Professor of Medicine, University of North Carolina, Chapel Hill, North Carolina. - Gregory L. Burke, M.D., M.Sc., Professor and Chair, Divi sion of Public Health Sciences, Wake Forest University School of Medicine, Winston-Salem, North Carolina. Leslie Bernstein, Ph.D., Director and Professor, Division of Cancer Etiology, Department of Population Sciences, Beckman Research Institute, City of Hope, and Dean for David M. Burns M.D., Professor Emeritus, Department of Faculty Affairs, Office of Academic Affairs, City of Hope, Family and Preventive Medicine, University of California San Diego School of Medicine, San Diego, California. Duarte, California. - Tim Byers M.D., M.P.H., Associate Director for Cancer Douglas W. Bettcher, M.D., M.P.H., Ph.D., Director, Pre Prevention and Control, University of Colorado Cancer vention of Noncommunicable Diseases Department, Non- Center, Associate Dean for Public Health Practice at the communicable Diseases and Mental Health Cluster, World Colorado School of Public Health, Aurora, Colorado. Health Organization, Geneva, Switzerland. Mary Beth Bigley, Dr.P.H., M.S.N., A.N.P., Director, Divi Neil E. Caporaso, M.D., Chief, Genetic Epidemiology - Branch, Division of Cancer Epidemiology and Genetics, sion of Science and Communications, and Acting Editor, Public Health Reports, Office of the Surgeon General, National Cancer Institute, Rockville, Maryland. Office of the Assistant Secretary for Health, Office of the Secretary, U.S. Department of Health and Human Ser - Richard Casaburi, Ph.D., M.D., Medical Director, Reha - bilitation Clinical Trials Center, Los Angeles Biomedical vices, Washington, D.C. Research Institute at Harbor-University of California Los Angeles Medical Center; Professor of Medicine, University Michele Bloch, M.D., Ph.D., Chief, Tobacco Control Research Branch, Behavioral Research Program, Division of California Los Angeles, School of Medicine, Torrance, California. of Cancer Control and Population Sciences, National Can- cer Institute, Rockville, Maryland. Frank J. Chaloupka, Ph.D., Director, Health Policy Center, Ron Borland Ph.D., Nigel Gray Distinguished Fellow in University of Illinois at Chicago, Distinguished Professor, Department of Economics, University of Illinois, Chicago, Cancer Prevention, Research Division, Cancer Council Illinois. Victoria, Australia. x

18 The Health Consequences of Smoking —50 Years of Progress June M. Chan, Sc.D., Professor, Epidemiology and Biosta Hashem B. El-Serag, M.D., M.P.H., Professor, Department - of Medicine, Chief of Gastroenterology and Hepatology, tistics, and Urology, Steven and Christine Burd-Safeway - Baylor College of Medicine, and Chief, Clinical Epidemi Distinguished Professor, University of California, San ology and Outcomes, Michael E. DeBakey Department of Francisco, California. Veterans Affairs Medical Center, Houston, Texas. Fellow of the Australian Graham A. Colditz, M.D., Dr. P.H., Faculty of Public Health Medicine , Niess-Gain Professor Michael Eriksen, Sc.D., Dean and Professor, School of and Chief, Division of Public Health Sciences, Department Public Health, Georgia State University, Atlanta, Georgia. of Surgery, Washington University School of Medicine, Eric J. Feuer, Ph.D., Chief, Statistical Methodology and and Associate Director, Prevention and Control, Siteman Applications Branch, Division of Cancer Control and Cancer Center, St. Louis, Missouri. Population Sciences, National Cancer Institute, Rockville, Maryland. D.M.D., M.P.H., Professor of Public , Gregory N. Connolly Health Practice, Director, Center for Global Tobacco Con - Lawrence J. Fine, M.D., Dr. P.H., Chief, Clinical Applica trol, School of Public Health, Harvard University, Boston, - tions and Prevention Branch, Division of Cardiovascu - Massachusetts. lar Sciences, National Heart, Lung, and Blood Institute, Karen H. Costenbader, M.D., M.P.H., Associate Professor Bethesda, Maryland. of Medicine, Harvard Medical School; Associate Physician - David W. Fleming, M.D., Director and Health Officer for and Co-Director of Lupus Center, Division of Rheuma Public Health, Seattle and King County, Seattle, Washing tology, Immunology and Allergy, Brigham and Women’s - ton. Hospital, Boston, Massachusetts. James D. Crapo, M.D., Professor of Medicine, National Silvia Franceschi, M.D., Special Advisor, Head, Infections and Cancer Epidemiology Group, International Agency for Jewish Health and University of Colorado Denver, Denver, Research on Cancer, Lyon, France. Colorado. Neal D. Freedman, Ph.D., M.P.H., Investigator, Nutritional Amanda J. Cross, Ph.D., Investigator, Nutritional Epide- Epidemiology Branch, Division of Cancer Epidemiology miology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, Maryland. and Genetics, National Cancer Institute, Rockville, Mary - land. K. Michael Cummings, Ph.D., M.P.H., Co-director, Tobacco Research Program, Hollings Cancer Center and Professor, Gary D. Friedman, M.D., M.S., Consulting Professor of Epidemiology, Department of Health Research and Policy, Department of Psychiatry and Behavioral Sciences, Medi - - Stanford University School of Medicine; Adjunct Investi - cal University of South Carolina, Charleston, South Caro gator (Former Director), Division of Research, Kaiser Per - lina. manente Medical Care Program of Northern California, Thomas Dietrich, D.M.D., M.D., M.P.H., Professor and Stanford, California. Chair, Department of Oral Surgery, School of Dentistry, Philip S. Gardiner, Dr. P.H., Program Officer, Policy and - University of Birmingham, Birmingham, United King dom. Regulatory Sciences, Nicotine Dependence and Neurosci- ences, Tobacco-Related Disease Research Program, Office Carolyn Dresler, M.D., M.P.A., Associate Director of Medi - of the President, University of California, Oakland, Cali - fornia. cal and Health Sciences, Office of Science, Center for Tobacco Products, U.S. Food and Drug Administration, Joseph Gfroerer, Chief, Population Survey Branch, Divi - Rockville, Maryland. sion of Surveillance and Data Collection, Center for Behavioral Health Statistics and Quality, Substance Abuse Bruce A. Dye, D.D.S., M.P.H., U.S. Public Health Service, and Mental Health Services Administration, Rockville, Dental Epidemiology Officer, Centers for Disease Control and Prevention, National Center for Health Statistics, Maryland. Division of Health and Nutrition Examination Statistics, Hyattsville, Maryland. xi

19 Surgeon General’s Report Gretchen Gierach, Ph.D., M.P.H., Investigator, Hormonal Susan Marsiglia Gray, M.P.H., Coordinator, National Synar - - Program, Center for Substance Abuse Prevention, Sub and Reproductive Epidemiology Branch, Division of Can stance Abuse and Mental Health Services Administration, cer Epidemiology and Genetics, National Cancer Institute, Rockville, Maryland. Rockville, Maryland. John D. Groopman, Ph.D., Anna M. Baetjer Professor of Elizabeth Ginexi, Ph.D., Health Scientist/Program Environmental Health Sciences, Department of Envi - Director, Tobacco Control Research Branch, Behavioral Research Program, Division of Cancer Control and Pop- ronmental Health Sciences, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland. ulation Sciences, National Cancer Institute, Rockville, Maryland. Beth Han, M.D., Ph.D., M.P.H., Statistician, Center for Behavioral Health Statistics and Quality, Substance Abuse Edward L. Giovannucci, M.D., Sc.D., Professor of Nutri- tion and Epidemiology, Harvard School of Public Health; and Mental Health Services Administration, U.S. Depart - ment of Health and Human Services, Rockville, Maryland. Associate Professor of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Harvard University, Craig L. Hanis, Ph.D., Professor, Division of Epidemiology, Boston, Massachusetts. Human Genetics and Environmental Sciences, University of Texas Health Science Center, Houston, Texas. - Gary A. Giovino, Ph.D., M.S., Professor and Chair, Depart ment of Community Health and Health Behavior, School - Curtis C. Harris, M.D., Chief, Laboratory of Human Car of Public Health and Health Professions, The State Univer - cinogenesis, Center of Cancer Research, National Cancer sity of New York at Buffalo, Buffalo, New York. - Institute, National Institutes of Health, Bethesda, Mary land. Stanton A. Glantz, Ph.D., Professor of Medicine and American Legacy Foundation Distinguished Professor in - Patricia Hartge, Sc.D., M.A., Deputy Director, Epide Tobacco Control; Director, Center for Tobacco Control Research and Education, University of California, San miology and Biostatistics Program, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Francisco, California. Bethesda, Maryland. Thomas J. Glynn, Ph.D., Director, Cancer Science and Trends; Director, International Cancer Control, American Dorothy K. Hatsukami, Ph.D., Forster Family Professor in Cancer Prevention, Professor, Department of Psychiatry, Cancer Society, Washington, DC. University of Minnesota, Minneapolis, Minnesota. - David C. Goff, Jr., M.D., Ph.D., Dean and Professor, Colo - rado School of Public Health, Aurora, Colorado. Stephen S. Hecht, Ph.D., Wallin Professor of Cancer Pre vention, Masonic Cancer Center, University of Minnesota, Minneapolis, Minnesota. Diane R. Gold, M.D., M.P.H., Professor of Medicine, Chan - ning Laboratory, Department of Medicine, Brigham and Jack E. Henningfield, Ph.D., Vice President, Research and Women’s Hospital; Professor of Environmental Health, Department of Environmental Health, Harvard School of Health Policy, Pinney Associates, Bethesda, Maryland, and Public Health, Boston, Massachusetts. Professor, Adjunct, Behavioral Biology, Department of Psychiatry and Behavioral Sciences, School of Medicine, Richard A. Goodman, M.D., M.P.H., Senior Medical Advi - The Johns Hopkins University, Baltimore, Maryland. sor, Office of the Medical Director, National Center for - James C. Hogg, M.D., Ph.D., Emeritus Professor of Pathol - Chronic Disease Prevention and Health Promotion, Cen ters for Disease Control and Prevention, Atlanta, Georgia. ogy, Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, British Colum - Gottesman, M.D., Ph.D., Associate Professor of Rebecca F. bia, Canada. Neurology and Epidemiology, Johns Hopkins University School of Medicine, Baltimore, Maryland. V. Michael Holers, M.D., Head, Division of Rheumatology; Professor of Medicine and Immunology, University of Col - orado School of Medicine, Denver, Colorado. Nigel Gray, AO, MBBS, FRACP, Honorary Senior Associate, Cancer Council Victoria, Carlton, Victoria, Australia. xii

20 The Health Consequences of Smoking —50 Years of Progress Annette Kaufman, Ph.D., M.P.H., Health Scientist/Pro- Stephen T. Holgate M.D., Professor, Fellow of the Academy gram Director, Tobacco Control Research Branch, Behav - of Medical Sciences, Medical Research Council, Clinical ioral Research Program, Division of Cancer Control and Professor of Immunopharmacology, University of South- Population Sciences, National Cancer Institute, Rockville, ampton, United Kingdom. Maryland. Philip P. Huang, M.D., M.P.H., Medical Director, Health Authority, Austin, Travis County Health and Human Ser Talmadge E. King, Jr., M.D., Chair, Department of Medi- - cine, Julius R. Krevans Distinguished Professorship in vices Department, Austin, Texas. Internal Medicine, Department of Medicine, University of Yvonne Hunt, Ph.D., M.P.H., Program Director, Tobacco California, San Francisco, California. Control Research Branch, Behavioral Research Program, Division of Cancer Control and Population Sciences, Barbara B. E. Klein, M.D., M.P.H., Professor, School of National Cancer Institute, Rockville, Maryland. - Ophthalmology and Visual Sciences, University of Wiscon sin, Madison, Wisconsin. Richard D. Hurt, M.D., Professor of Medicine, College of Ronald Klein, M.D., M.P.H., Professor, Department of Oph Medicine, Director, Nicotine Dependence Center, Mayo - Clinic, Rochester, Minnesota. thalmology and Visual Sciences, School of Medicine and - Public Health, University of Wisconsin, Madison, Wiscon Corrine Husten, M.D., M.P.H., Senior Medical Advisor, sin. Center for Tobacco Products, Food and Drug Administra - tion, Rockville, Maryland. Jennie Kline, Ph.D., Senior Research Scientist, Imprints Center, New York State Psychiatric Institute, New York, New York; Adjunct Professor of Epidemiology, Depart- Andrew Hyland, Ph.D., Chair, Department of Health ment of Epidemiology at Mailman School of Public Health Behavior, Roswell Park Cancer Institute, Buffalo, New and the Sergievsky Center, Columbia University, New York. York, New York. Eric J. Jacobs, Ph.D., Strategic Director of Pharmacoepi - Nino Künzli, M.D., Ph.D., Deputy Director and Head of demiology, Epidemiology Research Program, American Department, Swiss Tropical and Public Health Institute, Cancer Society, Atlanta, Georgia. Basel, Switzerland; Professor, Department of Public Prabhat Jha, O.C., M.D., D.Phil., Professor and Chair in Health, University of Basel, Basel, Switzerland. Disease Control, University of Toronto, Executive Direc - tor, Centre for Global Health Research, St. Michael’s Hos - Stephanie Land, Ph.D., Statistician, Tobacco Control pital, University of Toronto, Ontario, Canada. Research Branch, Behavioral Research Program, Division of Cancer Control and Population Sciences, National Can- cer Institute, Rockville, Maryland. Kenneth C. Johnson, Ph.D., Senior Epidemiologist/ - Research Scientist, Centre for Chronic Disease, Preven Kate Langley, Ph.D., Research Associate, The Medical tion and Control, Public Health Agency of Canada, Ottawa, Ontario, Canada. Research Council Centre for Neuropsychiatric Genetics and Genomics, Neuroscience and Mental Health Research Institute, Cardiff University, Cardiff, United Kingdom. Ron Johnson, Ph.D., Program Director, Cancer Etiology - Branch, Division of Cancer Biology, National Cancer Insti Scott J. Leischow, Ph.D., Department of Health Sciences tute, Rockville, Maryland. Research, Mayo Clinic, Scottsdale, Arizona. Amy E. Kalkbrenner, Ph.D., M.P.H., Assistant Professor, Loftus, Jr., M.D., Consultant, Division of Gas- Zilber School of Public Health, University of Wisconsin- Edward V. troenterology and Hepatology, Director of Inflammatory Milwaukee, Milwaukee, Wisconsin. Bowel Disease Interest Group, Professor of Medicine, Mayo Clinic, Rochester, Minnesota. Gilaad G. Kaplan, M.D., M.P.H., Associate Professor, Departments of Medicine and Community Health Sci- ences, Division of Gastroenterology, University of Calgary, Calgary, Alberta, Canada. xiii

21 Surgeon General’s Report - David A. Lomas, Ph.D., Sc.D., F.R.C.P., FMedSci., Dean, J. Michael McGinnis, M.D., M.A., M.P.P., Executive Direc tor, Value and Science-Driven Health Care; Senior Scholar, Faculty of Medical Sciences, University College London, - Institute of Medicine, The National Academies, Washing London, United Kingdom. ton, D.C. Jay H. Lubin, Ph.D., Special Volunteer, Biostatistics - Paul Melstrom, Pharm.D., Ph.D., Toxicologist/Epidemi Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, Maryland. ologist, Epidemiology Branch, Office on Smoking and Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Russell V. Luepker, M.D., M.S., Mayo Professor, Division of Epidemiology and Community Health, School of Public Prevention, Atlanta, Georgia. Health, University of Minnesota, Minneapolis, Minnesota. Martin M. Miner, M.D., Co-Director, Men’s Health Center, Jing Ma, M.D., Ph.D., Associate Professor, Channing Divi - - The Miriam Hospital, Clinical Associate Professor of Fam - - sion of Network Medicine, Brigham and Women’s Hospi ily Medicine and Urology, Warren Alpert School of Medi tal, Harvard Medical School; Department of Epidemiology, cine, Brown University, Providence, Rhode Island. Harvard School of Public Health, Harvard University, Bos - Glen Morgan, Ph.D., Psychologist, Tobacco Control ton, Massachusetts. Research Branch, Behavioral Research Program, Division Patricia L. Mabry, Ph.D., Senior Advisor and Acting Dep - of Cancer Control and Population Sciences, National Can- cer Institute, Rockville, Maryland. uty Director, Office of Behavioral and Social Sciences Research, Office of the Director, National Institutes of Health, Bethesda, Maryland. Matthew L. Myers, J.D., President, Campaign for Tobacco- Free Kids, Washington, D.C. - Judith Mackay, F.R.C.P (Edin), F.R.C.P (Lon), Senior Advi sor, World Lung Foundation, New York, USA Thomas E. Novotny, M.D., M.P.H., Professor and Associate ; Director, Asian Consultancy on Tobacco Control, Hong Kong. Director for Border and Global Health, Division of Epi- demiology, Graduate School of Public Health, San Diego Ruth E. Malone, R.N., Ph.D., F.A.A.N., Professor and Chair, State University, San Diego, California. Department of Social and Behavioral Sciences, School of Sarah Nyante, Ph.D., M.S.P.H, Postdoctoral Fellow, Hor - Nursing, University of California, San Francisco, Editor, San Francisco, California. monal and Reproductive Epidemiology Branch, Division Tobacco Control, of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, Maryland. Stephen E. Marcus, Ph.D., Epidemiologist and Program - Director, Social and Behavioral Modeling Research Pro Andrew F. Olshan, Ph.D., Professor and Chair, Department gram, Division of Biomedical Technology, Bioinformatics, of Epidemiology, Gillings School of Global Public Health, and Computational Biology, National Institute of General Medical Sciences, National Institutes of Health, Bethesda, University of North Carolina, Chapel Hill, North Carolina. Maryland. Jamie S. Ostroff, Ph.D., Chief, Behavioral Sciences Ser - vice, Department of Psychiatry and Behavioral Sciences, Joseph B. Margolick, M.D., Ph.D., Professor, Department Memorial Sloan-Kettering Cancer Center; Professor of of Molecular Microbiology and Immunology, Bloomberg Psychology and Public Health, Weill Cornell Medical Col - School of Public Health, Johns Hopkins University, Balti - lege, New York, New York. more, Maryland. - Fernando J. Martinez, M.D., M.S., Professor of Internal Yikyung Park, Sc.D., Staff Scientist, Nutritional Epide miology Branch, Division of Cancer Epidemiology and Medicine, Associate Division Chief of Clinical Research, Genetics, National Cancer Institute, Rockville, Maryland. - Division of Pulmonary and Critical Care Medicine, Univer sity of Michigan, Ann Arbor, Michigan. John P. Pierce, Ph.D., Professor of Cancer Research, Department of Family and Preventive Medicine, Director Carol A. Mathews, M.D., Associate Professor, Department of Population Research, Moores Cancer Center, University of Psychiatry, University of California, San Francisco; Co- - of California San Diego, La Jolla California. Chair Tourette Syndrome Association International Con sortium for Genetics, San Francisco, California. xiv

22 —50 Years of Progress The Health Consequences of Smoking Donald R. Shopland, U.S. Public Health Service (retired), Elizabeth A. Platz, Sc.D., M.P.H., Professor, Department of Epidemiology, Johns Hopkins Bloomberg School of Pub - Ringgold, Georgia. lic Health and the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University, Baltimore, Maryland. Alfred Sommer, M.D., M.H.S., Dean Emeritus, Bloomberg - School of Public Health; Professor, Departments of Epide miology, Ophthalmology and International Health, Johns Robert N. Proctor, Ph.D., Professor of the History of Sci - Hopkins University, Baltimore, Maryland. ence and Professor, by courtesy, of Pulmonary and Critical Care, Stanford University, Stanford, California. - Frank E. Speizer, M.D., E. H. Kass Distinguished Profes Stephen I. Rennard, M.D., Larson Professor, Division of sor of Medicine, Harvard Medical School, Channing Labo - ratory, Department of Medicine, Brigham and Women’s Pulmonary, Critical Care, Allergy and Sleep Medicine, Hospital, Boston, Massachusetts. University of Nebraska Medical Center, Omaha, Nebraska. Meir J. Stampfer, M.D., Dr.P.H., Head, Chronic Disease William N. Rom, M.D., M.P.H., Director, Division of Pul - Epidemiology Unit, Channing Division of Network Medi - monary, Critical Care, and Sleep Medicine, Sol and Judith cine, Department of Medicine, Brigham and Women’s Bergstein Professor of Medicine and Environmental Medi - cine, New York University School of Medicine, New York, - Hospital and Harvard Medical School; Professor of Epide miology and Nutrition, Harvard School of Public Health; New York. and Professor of Medicine, Harvard Medical School, Bos- ton, Massachusetts. Julia H. Rowland, Ph.D., Director, Office of Cancer Survi - vorship, Division of Cancer Control and Population Sci - Gary D. Stoner, Ph.D., Professor, Department of Medicine, ences, National Cancer Institute, Rockville, Maryland. Division of Hematology and Oncology, Medical College of Marc B. Schenker, M.D., M.P.H., Director, Center for Occu Wisconsin, Milwaukee, Wisconsin. - pational and Environmental Health; Professor, Depart - Catherine M. Stoney, Ph.D., Program Director, Division of ment of Public Health Sciences, University of California, Davis, California. Cardiovascular Sciences, National Heart, Lung, and Blood - Institute, National Institutes of Health, Bethesda, Mary land. Steven A. Schroeder, M.D., Distinguished Professor of Health and Healthcare and Director, Smoking Cessation Leadership Center, Department of Medicine, University of - Ira B. Tager, M.D., M.P.H., Professor Emeritus of Epide California-San Francisco, San Francisco, California. miology, School of Public Health, University of California, Berkeley, California. - David A. Schwartz, M.D., Professor of Medicine and Immu Prue Talbot, Ph.D., Professor of Cell Biology, Director of nology, Chair, Department of Medicine, University of Colo- - the Stem Cell Center and Core, Department of Cell Biol rado, Aurora, Colorado. ogy and Neuroscience, University of California, Riverside, Donald W. Sexton, M.B.A., M.H.A., Doctoral Student, Uni California. - versity of Michigan School of Public Health, and United States Army Medical Department, Ann Arbor, Michigan. Crystal E. Tan, M.D., M.S., Center for Tobacco Control Research and Education, University of California, San A. Richey Sharrett, M.D., Dr. P.H., Professor, Department Francisco, California. of Epidemiology, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland. Michael J. Thun, M.D., M.S., Emeritus Vice President, - Epidemiology and Surveillance Research, American Can Peter G. Shields, M.D., Deputy Director, Comprehensive cer Society, Atlanta, Georgia. - Cancer Center, Professor, College of Medicine, James Can cer Hospital, The Ohio State University Wexner Medical Douglas Tipperman, M.S.W., Lead Public Health Advisor, Center, Columbus, Ohio. Center for Substance Abuse Prevention, Substance Abuse and Mental Health Services Administration, Rockville, Maryland. xv

23 Surgeon General’s Report - Benjamin Toll, Ph.D., Program Director, Smoking Cessa Walter C. Willett, M.D., Dr. P.H., Departments of Nutri - tion Service at Smilow Cancer Hospital and Yale Cancer tion and Epidemiology, Harvard School of Public Health, Center; Associate Professor, Department of Psychiatry, Channing Division of Network Medicine, Department of School of Medicine, Yale University, New Haven, Con - Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts. necticut. Nick Wilson, M.B.Ch.B., M.P.H., Associate Professor of Edward J. Trapido, Sc.D., F.A.C.E., Associate Dean for - Public Health, Co-Director of the Burden of Disease Epi Research, Professor and Wendell H. Gauthier Chair of demiology, Equity and Cost-effectiveness Programme, Cancer Epidemiology, Louisiana State University School of Public Health; Deputy Director, Stanley S. Scott Cancer University of Otago, Wellington, New Zealand. Center; Professor of Medicine and Senior Liaison to the Dean of the School of Medicine, Louisiana State University Gayle C. Windham, Ph.D., M.S.P.H., Chief, Epidemio - logical Surveillance Unit, Division of Environmental and Health Sciences Center, New Orleans, Louisiana. Occupational Disease, California Department of Public Bruce J. Trock, Ph.D., Professor, Departments of Urology, Health, Richmond, California. Oncology, Epidemiology, and Environmental Health Sci- ences, Director, Division of Epidemiology of the James Anna H. Wu, Ph.D., Professor, Keck School of Medicine, Buchanan Brady Urological Institute, Johns Hopkins Department of Preventive Medicine, University of South - ern California, Los Angeles, California. School of Medicine, Baltimore, Maryland. Mandi Yu, Ph.D., M.S., Mathematical Statistician, Data Donna Vallone, Ph.D., M.P.H., Senior Vice President, Analysis and Interpretation Branch, Division of Cancer American Legacy Foundation; Assistant Professor (Adjunct), Department of Health, Behavior and Society, Control and Population Sciences, National Cancer Insti - Johns Hopkins Bloomberg School of Public Health, Bal tute, Rockville, Maryland. - timore, Maryland. , Rohit Varma, M.D., M.P.H. Other contributors were Professor and Chair, Illinois Eye and Ear Infirmary, Department of Ophthalmology and Julie M. Carter, M.P.H., Public Health Analyst, Office on Visual Sciences; Associate Dean for Strategic Planning, Smoking and Health, National Center for Chronic Disease College of Medicine, Chicago, Illinois. Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia. Robert E. Vollinger, M.S.P.H., Public Health Advisor, Tobacco Control Research Branch, Behavioral Research Stephanie Denkhoff Boone, Ph.D., M.P.H., Postdoctoral - Program, Division of Cancer Control and Population Sci Associate, Department of Epidemiology and Population ences, National Cancer Institute, Rockville, Maryland. - Health, School of Public Health and Information Sci ences, James Graham Brown Cancer Center, University of Melanie A. Wakefield, Ph.D., Director, Center for Behav- Louisville, Louisville, Kentucky. ioral Research in Cancer, Cancer Council Victoria, Mel - bourne, Australia. Athena Foong, M.S., Research Writer/Research Assistant - to Jonathan M. Samet, M.D., M.S., Department of Pre Robert B. Wallace, M.D., M.Sc., Director, Center on Aging, ventive Medicine, Keck School of Medicine, University of Professor, Department of Epidemiology, College of Public Southern California, Los Angeles, California. Health, and Professor, Department of Internal Medicine, Carver College of Medicine, University of Iowa, Iowa City, Earl S. Ford, M.D., M.P.H., Medical Officer, Epidemiology Iowa. and Surveillance Branch, Division of Population Health, National Center for Chronic Disease Prevention and - Kenneth E. Warner, Ph.D., Avedis Donabedian Distin Health Promotion, Centers for Disease Control and Pre - - guished University Professor of Public Health, Depart vention, Atlanta, Georgia. ment of Health Management and Policy, School of Public Health, University of Michigan, Ann Arbor, Michigan. xvi

24 The Health Consequences of Smoking —50 Years of Progress Chronic Disease Prevention and Health Promotion, Cen Joseph Gfroerer, Chief, Population Survey Branch, Divi - - sion of Surveillance and Data Collection, Center for ters for Disease Control and Prevention, Atlanta, Georgia. Behavioral Health Statistics and Quality, Substance Abuse Gabbi Promoff, M.A., Associate Director for Policy, Office - and Mental Health Services Administration, U.S. Depart on Smoking and Health, National Center for Chronic Dis - ment of Health and Human Services, Rockville, Maryland. - ease Prevention and Health Promotion, Centers for Dis Beth Han, M.D., Ph.D., M.P.H., Statistician, Center for ease Control and Prevention, Atlanta, Georgia. Behavioral Health Statistics and Quality, Substance Abuse Hilda Razzaghi, Ph.D., M.S.P.H., Epidemiologist-ORISE - and Mental Health Services Administration, U.S. Depart ment of Health and Human Services, Rockville, Maryland. Fellow, Birth Defects Branch, National Center on Birth - Defects and Developmental Disabilities, Centers for Dis S. Jane Henley, M.S.P.H., Epidemiologist, - Division of Can ease Control and Prevention, Atlanta, Georgia. cer Prevention and Control, National Center for Chronic Lisa C. Richardson, M.D., M.P.H., Associate Director for Disease Prevention and Health Promotion, Centers for Science, Division of Cancer Prevention and Control, Disease Control and Prevention, Atlanta, Georgia. National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Pre - Carissa Baker Holmes, M.P.H., Public Health Analyst, vention, Atlanta, Georgia. Office on Smoking and Health, National Center for - Chronic Disease Prevention and Health Promotion, Cen Emily Ridgway, M.P.H., ORISE Fellow, Office on Smoking ters for Disease Control and Prevention, Atlanta, Georgia. - and Health, National Center for Chronic Disease Preven - Lynn A. Hughley, Health Communications Special tion and Health Promotion, Centers for Disease Control ist, Office on Smoking and Health, National Center for and Prevention, Atlanta, Georgia. - Chronic Disease Prevention and Health Promotion, Cen Lauren K. Shaw, M.S., Statistician, Office on Smoking and ters for Disease Control and Prevention, Atlanta, Georgia. Health, National Center for Chronic Disease Prevention Joel London, M.P.H., C.H.E.S., Health Communications and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia. Specialist, Office on Smoking and Health, National Cen - ter for Chronic Disease Prevention and Health Promo- tion, Centers for Disease Control and Prevention, Atlanta, Dana Shelton, M.P.H., Deputy Director, National Center Georgia. for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Kristy Marynak, M.P.P., Public Health Analyst, Office on Georgia. Smoking and Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Kathryn E. Szynal, Editorial Assistant, Office on Smoking and Health, National Center for Chronic Disease Preven - Control and Prevention, Atlanta, Georgia. tion and Health Promotion, Centers for Disease Control Luz M. Moncayo, Executive Assistant to Jonathan M. and Prevention, Atlanta, Georgia. Samet, M.D., M.S., Department of Preventive Medicine, - Keck School of Medicine, Institute for Global Health, Uni Cheryll C. Thomas, M.S.P.H., Deputy Associate Director versity of Southern California, Los Angeles, California. for Science, Office of the Director, Division of Cancer Pre - vention and Control, National Center for Chronic Disease Jake Nikota, Ph.D. Candidate, Department of Pathology Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia. and Molecular Medicine, McMaster Immunology Research Centre, McMaster University, Hamilton, Ontario, Canada. Dale J. Willett, J.D., Candidate Class of 2013, Legal Research and Writing Assistant, Northeastern University Erin L. O’Connor, M.S., Data Analyst, Office on Smoking - School of Law, Boston, Massachusetts. and Health, National Center for Chronic Disease Preven tion and Health Promotion, Centers for Disease Control - and Prevention, Atlanta, Georgia. Dongyan Yang, M.D., M.S., Assistant Professor, Depart ment of Epidemiology and Population Health, School of Alissa O’Halloran, M.S.P.H., Health Research Analyst, Public Health, University of Louisville, Louisville, Ken - tucky. Office on Smoking and Health, National Center for xvii

25

26 The Health Consequences of Smoking —50 Years of Progr ess Chapter 1. Introduction, Summary, and Conclusions 1 Introduction 3 Organization of the Report 4 6 Development of the Report 7 Scientific Basis of the Report Major Conclusions from the Report 7 8 Chapter Conclusions References 14 Chapter 2. Fifty Years of Change—1964–2014 15 17 Introduction Tobacco Control: Before the 1964 Report 19 21 Scientific Judgment and the 1964 Report 23 Remedial Action and Change Following the 1964 Report Passive Smoking and Environmental Change 26 Nicotine and Addiction 30 Denormalization and the Tobacco Industry 31 Summary 33 References 35 Chapter 3. Producing the Surgeon General’s Report From 1964–2014: Process and Purpose 43 Introduction 45 Development of a Scientific Consensus 45 Development of the Criteria for Causation Used in the 1964 Report 51 Methods for Reviewing the Evidence and Developing Conclusions 53 Process of Ensuring Consensus and Strength of the Peer Review 55 Separation of Scientific Conclusions and the Formation of Policy 56 References 57 Chapter 4. Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014 61 Introduction 63 Evolution of Conclusions on Cigarette Smoking amd Exposure to Secondhand Smoke as a Cause of Disease 67 Summary 69 104 References xix

27 Chapter 5. Nicotine 107 109 Introduction Toxicokinetics and Acute Toxicity of Nicotine 109 Pathophysiology of Nicotine Addiction 112 Health Consequences of Nicotine Exposure 113 125 Evidence Summary 126 Conclusions Implications 126 127 References Chapter 6. Cancer 139 Introduction 143 148 Mechanisms of Cancer Induction by Tobacco Smoke Changing Cigarettes and Risk for Lung Cancer Over Time 151 187 Liver Cancer 197 Colorectal Cancer 204 Prostate Cancer 209 Breast Cancer 284 Adverse Health Outcomes in Cancer Patients and Survivors 292 Evidence Summary Chapter Conclusions 293 References 294 Chapter 7. Respiratory Diseases 353 Introduction 355 Chronic Obstructive Pulmonary Disease 355 Asthma 371 Tuberculosis 376 386 Idiopathic Pulmonary Fibrosis Impact of Smoke-free Policies on Respiratory Outcomes 389 391 Evidence Summary Implications 391 Chapter Conclusions 391 References 393 Chapter 8. Cardiovascular Diseases 411 Introduction 413 Tobacco Use and Cardiovascular Diseases: Evolution of the Evidence 418 Mechanisms by Which Smoking Causes Cardiovascular Diseases 419 Updated Summaries of the Evidence: Active Smoking 423 Updated Evidence Reviews 430 Impact of Smokefree Laws on Acute Cardiovascular Events 434 Evidence Summary 443 Chapter Conclusions 444 Implications 445 446 References xx

28 Chapter 9. Reproductive Outcomes 459 461 Introduction Surveillance 461 Cessation 463 Advances in the Understanding of Tobacco and Reproductive Health 464 471 Updated Evidence Reviews 499 Chapter Conclusions References 500 523 Chapter 10. Other Specific Outcomes Introduction 527 Eye Disease: Age-Related Macular Degeneration 527 532 Dental Disease Diabetes 537 545 Immune Function and Autoimmune Disease Rheumatoid Arthritis and Systemic Lupus Erythematosus 572 576 Inflammatory Bowel Disease 584 Chapter Conclusions 585 References Chapter 11. General Morbidity and All-Cause Mortality 623 625 Introduction 626 Smoking and General Morbidity and Economic Costs All-Cause Mortality 633 Summary 641 642 Chapter Conclusions 642 Implications References 643 Chapter 12. Smoking-Attributable Morbidity, Mortality, and Economic Costs 647 649 Introduction Methodology Used by CDC to Compute Smoking-Attributable Mortality in the United States 650 651 Methodological Issues in SAM Calculation 2013 Update to SAMMEC Methodology 655 Smoking-Attributable Mortality in Adults and Infants, United States, 2005–2009 659 Projected Smoking-Related Deaths Among Youth, United States, 2012 666 Smoking-Attributable Morbidity Estimates 667 676 Total Smoking-Attributable Mortality, 1965–2014 Summary 678 Conclusions 679 Implications 679 References 680 Chapter 12 Appendices 685 Appendix 12.1 687 Appendix 12.2 691 698 References xxi

29 Chapter 13. Patterns of Tobacco Use Among U.S. Youth, Young Adults, and Adults 701 703 Introduction 705 Historical Trends in Tobacco Use Cigarette Smoking 708 730 Other Tobacco Products Conclusions 761 761 Summary and Implications Appendix 13.1: Sources of Data 762 Appendix 13.2: Measures of Tobacco Use 763 767 References 771 Chapter 14. Current Status of Tobacco Control 773 Introduction The Changing Public Image of Tobacco 774 Changes in the Tobacco Industry, Products, and Product Regulations 779 788 Tobacco Control Policies 803 Clinical and Educational Approaches for Tobacco Cessation International Tobacco Control 818 826 Summary Conclusions 827 References 828 Chapter 15. The Changing Landscape of Tobacco Control—Current Status and Future Directions 843 Introduction 845 845 The Tobacco Control Landscape: Over a Hundred Years and Counting 852 Potential End Game Strategies Chapter Summary 857 Conclusions 858 Implications for Ending the Tobacco Epidemic 859 References 860 865 Chapter 16. A Vision for Ending the Epidemic: Toward a Society Free of Tobacco-caused Death and Disease A Vision for Ending the Tobacco Epidemic 867 Accelerating the National Movement to Reduce Tobacco Use 875 References 876 List of Abbreviations 879 List of Tables and Figures 883 List of Gene Tables 895 903 Index xxii

30 Chapter 1 Introduction, Summary, and Conclusions 3 Introduction 4 Organization of the Report Development of the Report 6 7 Scientific Basis of the Report 7 Major Conclusions from the Report Chapter Conclusions 8 8 Chapter 5: Nicotine Chapter 6: Cancer 8 8 Lung Cancer L Cancer 8 iver 8 Colorectal Cancer 8 Prostate Cancer Breast Cancer 8 Adverse Health Outcomes in Cancer Patients and Survivors 9 Chapter 7: Respiratory Diseases 9 Chronic Obstructive Pulmonary Disease 9 Asthma 9 9 Tuberculosis 10 Idiopathic Pulmonary Fibrosis Chapter 8: Cardiovascular Disease 10 Chapter 9: Reproductive Outcomes 10 Congenital Malformations 10 Neurobehavioral Disorders of Childhood 10 Ectopic Pregnancy 10 Spontaneous Abortion 10 Male Sexual Function 10 Chapter 10: Other Specific Outcomes 10 Eye Disease: Age-Related Macular Degeneration 10 Dental Disease 10 Diabetes 11 Immune Function and Autoimmune Disease 11 Rheumatoid Arthritis 11 Systemic Lupus Erythematosus 11 11 Inflammatory Bowel Disease 1

31 11 Chapter 11: General Morbidity and All-Cause Mortality Chapter 12: Smoking-Attributable Morbidity, Mortality, and Economic Costs 11 12 Chapter 13: Patterns of Tobacco Use Among U.S. Youth, Young Adults, and Adults Chapter 14: Current Status of Tobacco Control 12 Chapter 15: The Changing Landscape of Tobacco Control— Current Status and Future Directions 13 14 References 2

32 The Health Consequences of Smoking —50 Years of Progress Introduction . These laws required a health warning on On January 11, 1964, Luther L. Terry, M.D., Sur - ing Act of 1969 Smoking and geon General of the United States, released cigarette packages, banned cigarette advertising in the broadcasting media, and called for an annual report on the Health: Report of the Advisory Committee of the Surgeon General of the Public Health Service health consequences of smoking. Since then, there have . This report, writ- ten at the request of President John F. Kennedy, was in been several actions at the federal level—the enactment of the response to the evidence on smoking and lung cancer that Family Smoking Prevention and Tobacco Control had been accumulating since the 1950s (see Chapter 2, Ending the Tobacco in 2009, and the publication of Act “Fifty Years of Change 1964–2014”). This was the first in Epidemic: A Tobacco Control Strategic Plan for the U.S. (USDHHS Department of Health and Human Services the series that is now generally referred to as the Surgeon 2010a). General’s reports. On the basis of more than 7,000 articles Since that first report in 1964, knowledge of the in the biomedical literature relating to smoking and dis - - health consequences of smoking and involuntary expo - ease that were available at the time, the Advisory Commit tee concluded that cigarette smoking is: sure to tobacco smoke has expanded dramatically (see Chapter 4, “Advances in Knowledge on the Health Con - sequences of Smoking: From 1964–2014”). This series of • Associated with 70% higher all-cause mortality reports has provided definitive syntheses of the evolving rates among men evidence on smoking and health. The topics have ranged A cause of lung cancer and laryngeal cancer in men widely, including comprehensive coverage of the adverse • - health effects of active smoking and exposure to second hand smoke (USDHEW 1979; U.S. Department of Health • A probable cause of lung cancer in women and Human Services [USDHHS] 1986, 2004, 2006), the impact of tobacco control policies (USDHHS 2000), and • The most important cause of chronic bronchitis addiction (USDHHS 1988). A goal of these reports has (U.S. Department of Health, Education, and Welfare - been to synthesize available evidence to reach conclu [USDHEW] 1964). sions on causality that have public health implications. In - For several days, the report was the topic of news - reaching conclusions on causation, the reports have fol paper headlines across the country and lead stories on - lowed a model that originated with the 1964 report: com pilation of all relevant lines of scientific evidence, critical television newscasts (Parascandola 1997). Later, it was ranked among the top news stories of the 20th century assessment of the evidence, evaluation of the strength ( of evidence by using guidelines for evidence evaluation, USA Today 1999). The release of that report was one of the first in a series of steps, still being taken 50 years later, and a summary conclusion on causation (USDHEW 1964; USDHHS 2004; Table 1.1; Chapter 3, “Producing the Sur - to diminish the impact of tobacco use on the health of people worldwide. Ever since, individual citizens, private - geon General’s Report from 1964–2014: Process and Pur pose”). The Surgeon General’s reports have established a organizations, public agencies, and elected officials have pursued the Advisory Committee’s call for “appropriate long list of health consequences and diseases caused by tobacco use and exposure to tobacco smoke (see Chapter remedial action.” Early on, in response to the 1964 report, the U.S. 4). Fifty years later, this report documents that our knowl - - Congress passed the Federal Cigarette Labeling and Adver edge continues to expand as new causal conclusions are Public Health Cigarette Smok and the tising Act of 1965 - still being added to that long list (Figures 1.1A and 1.1B). Table 1.1 Four-level hierarchy for classifying the strength of causal inferences from available evidence Level 1 Evidence is sufficient to infer a causal relationship Level 2 Evidence is suggestive but not sufficient to infer a causal relationship Evidence is inadequate to infer the presence or absence of a causal relationship (which encompasses evidence that is Level 3 sparse, of poor quality, or conflicting) Evidence is Level 4 suggestive of no causal relationship Source : U.S. Department of Health and Human Services 2004. 3 Introduction, Summary, and Conclusions

33 Surgeon General’s Report Figure 1.1A The health consequences causally linked to smoking Source: USDHHS 2004, 2006, 2012. Note: Each condition presented in bold text and followed by an asteri s k (*) is a new disease that has been causally linked to smoking in this report. Organization of the Report in this section gives a 50-year perspective on the identi - This report is divided into three sections. Section 1 “Historical perspective, overview, and conclusions” pro - fication of the health consequences of active smoking vides an overall summary of the report and its conclu- and exposure to secondhand smoke. The other chapters sions. It also provides a summary of the history of this in this section provide updates on critical topics and on series of reports, moving from their origins in 1964 to the topics for which the evidence has advanced, since the previous reviews in the 2004 and 2006 Surgeon General’s present, contrasting what we knew in 1964 with what we reports, know now in 2014. Section 2 “The Health Consequences The Health Consequences of Smoking: A Report of Active and Passive Smoking: The Evidence in 2014” The Health Consequences of of the Surgeon General and Involuntary Exposure to Tobacco Smoke: A Report of the provides a direct link to the 1964 report, which addressed the health effects of active smoking only. The first chapter Surgeon General, including a brief review of the state of Chapter 1 4

34 The Health Consequences of Smoking —50 Years of Progress Figure 1.1B The health consequences causally linked to exposure to secondhand smoke USDHHS 2004, 2006. Source: Each condition presented in bold text and followed by an asteri Note: s k (*) is a new disease that has been causally linked to exposure to secondhand smoke in this report. T the evidence. Understanding of mechanisms, as laid out in he final chapter “A Vision for the Ending the Tobacco How Tobacco Smoke Causes Disease: The the 2010 report, Epidemic” outlines broad strategies and potential courses Biology and Behavioral Basis for Smoking-Attributable of action for tobacco control in the future. Disease, is also (USDHHS 2010b). Active smoking and Each section within the chapters on the health con- exposure to secondhand smoke are covered in the same sequences of smoking (Chapters 6 – 11) is accompanied chapters. Section 3 “Tracking and Ending the Epidemic” by evidence tables detailing the studies that were used to evaluate the evidence to assess causality. A supplement includes a descriptive chapter on the patterns of smoking, to this report is provided that contains these tables. The a chapter on the impact of the tobacco control environ - tables included in the supplement are indicated with an - ment on smoking since 1964, and additional chapters pro “ S ” where they are called out in the text. viding estimates of premature deaths that are avoidable. Introduction, Summary, and Conclusions 5

35 Surgeon General’s Report Development of the Report the Surgeon General has increasingly become involved The Surgeon General’s reports on smoking and health were previously mandated by the Cigarette Smok - in developing the messaging for the public release of the reports. Consistent with a primary duty of the Surgeon ing Act of 1969, Public Law 91-222, section 8 (a), which General to “Protect and advance the health of the Nation required that “The Secretary of Health, Education, and through educating the public, advocating for effective Welfare shall transmit a report to Congress not later than disease prevention and health promotion programs and January 1, 1971, and annually thereafter, concerning (A) activities, and, providing a highly recognized symbol of current information in the health consequences of smok - national commitment to protecting and improving the ing, and (B) such recommendations for legislation as he the Office of the Surgeon General (n.d.) may deem appropriate.” In addition, recent reports have public’s health,” has expanded the range of educational materials support - - also satisfied the statutory reporting required by the Com ing the release of the scientific report, particularly the prehensive Smokeless Tobacco Health Education Act of development of a consumer summary which is produced 1986, Public Law 99-252, which required that “The Sec - in nontechnical but scientifically valid language. retary of Health and Human Services shall transmit a As shown in Table 3.1, over time the size of the report to Congress no later than January 11, 1987, and - reports has grown, largely due to the increase in scien biennially thereafter, containing—(1) a description of the tific literature on the topics reviewed, but also as the scope effects of health education efforts on the use of smokeless of topics has grown from those addressed in the initial tobacco products, (2) a description of the use by the public charge provided by Secretary Califano in 1979 to address of smokeless tobacco products, (3) an evaluation of the the behavioral and tobacco control aspects of the problem. health effects of smokeless tobacco products and the iden - This broader focus is reflected in the 2012 report which tification of areas appropriate for further research, and (4) reviewed not only the epidemiology, causes, and health such recommendation for legislation and administrative effects of tobacco use among youth and young adults, - action as the Secretary considers appropriate.” These stat but also the interventions proven to prevent this problem utory requirements were sunsetted in 1999 and an annual (USDHHS 2012). report to Congress is no longer required by law. This report of the Surgeon General was prepared Initially, the annual reports to Congress on the by the Office on Smoking and Health, National Center health consequences of smoking were prepared by the for Chronic Disease Prevention and Health Promotion, - National Clearinghouse for Smoking and Health; how Centers for Disease Control and Prevention, USDHHS. ever, in 1978 Secretary of Health, Education, and Welfare Initial chapters were written between 2010–2011 by 75 Joseph Califano established the Office on Smoking and experts selected because of their knowledge of, and famil - Health in the Office of the Assistant Secretary of Health iarity with, the topics presented here. These contributions to coordinate the production of the annual report to Con - are summarized in 15 chapters, which were evaluated by gress that would review not only the biomedical but also more than 100 peer reviewers. The entire manuscript was the behavioral and control data about smoking and its then sent to more than 20 scientists and other experts, effects on health. The fifteenth anniversary report (USD- who examined it for scientific integrity. After each review HEW 1979) was the first report produced by the Office cycle, the drafts were revised by the editors on the basis on Smoking and Health (see Table 3.1 for a full listing of of the reviewers’ comments. Subsequently, the report was reports from 1964–2012). - reviewed by various institutes and agencies within USD Beginning with Dr. Luther L. Terry, each Surgeon HHS. Publication lags, even short ones, prevent an up-to- General has released the reports to the public and served the-minute inclusion of all recently published articles and as the primary spokesperson of the findings. However, the data. Therefore, by the time the public reads this report, preparation of these reports, starting with the 1964 Advi - additional studies or data may have been published. sory Committee, has been conducted with a high degree The methodology for evidence compilation, review, of independence, in order to protect their scientific integ- and synthesis draws on the approach of the 1964 Surgeon rity. Although the public may assume that the individual General’s report (USDHEW 1964), as further modified in Surgeon Generals have been active in the authoring of the 2004 report (USDHHS 2004). That report also refined the reports, their role has remained largely at the level the methodology for causal inference and set out a clas- - of approving topics and reviewing drafts before the vol sification of strength of evidence for causal inference. ume is published. Nevertheless, over time, the Office of 6 Chapter 1

36 The Health Consequences of Smoking —50 Years of Progress Scientific Basis of the Report The statements and conclusions throughout this - owing to the delay between acceptance and final publica tion, the study is referred to as “in press.” This report also report are documented by the citation of studies published refers, on occasion, to unpublished research such as a pre - in the scientific literature. For the most part, this report cites peer-reviewed journal articles, including reviews that sentation at a professional meeting or a personal commu - nication from the researcher. These personal references integrate findings from numerous studies, and books by recognized experts. When a study has been accepted for are to acknowledge experts whose research is in progress. publication, but the publication has not yet been issued, Major Conclusions from the Report 1. The century-long epidemic of cigarette smoking has In addition to causing multiple diseases, cigarette 6. smoking has many other adverse effects on the body, caused an enormous avoidable public health tragedy. such as causing inflammation and impairing immune Since the first Surgeon General’s report in 1964 more function. than 20 million premature deaths can be attributed to cigarette smoking. 7. Although cigarette smoking has declined signifi - cantly since 1964, very large disparities in tobacco use The tobacco epidemic was initiated and has been 2. - sustained by the aggressive strategies of the tobacco remain across groups defined by race, ethnicity, edu cational level, and socioeconomic status and across industry, which has deliberately misled the public on regions of the country. the risks of smoking cigarettes. 3. Since the 1964 Surgeon General’s report, cigarette 8. Since the 1964 Surgeon General’s report, compre - hensive tobacco control programs and policies have smoking has been causally linked to diseases of nearly all organs of the body, to diminished health status, been proven effective for controlling tobacco use. Further gains can be made with the full, forceful, and and to harm to the fetus. Even 50 years after the sustained use of these measures. first Surgeon General’s report, research continues to newly identify diseases caused by smoking, including such common diseases as diabetes mellitus, rheuma - 9. The burden of death and disease from tobacco use in the United States is overwhelmingly caused by ciga- toid arthritis, and colorectal cancer. rettes and other combusted tobacco products; rapid Exposure to secondhand tobacco smoke has been 4. elimination of their use will dramatically reduce this causally linked to cancer, respiratory, and cardiovas - burden. cular diseases, and to adverse effects on the health of - For 50 years the Surgeon General’s reports on smok infants and children. 10. - ing and health have provided a critical scientific foun dation for public health action directed at reducing 5. The disease risks from smoking by women have risen tobacco use and preventing tobacco-related disease sharply over the last 50 years and are now equal to those for men for lung cancer, chronic obstructive and premature death. pulmonary disease, and cardiovascular diseases. 7 Introduction, Summary, and Conclusions

37 Surgeon General’s Report Chapter Conclusions The evidence is not sufficient to specify which design Note: Chapters 2-4 do not have conclusions. 3. changes are responsible for the increased risk of adenocarcinoma, but there is suggestive evidence that ventilated filters and increased levels of tobacco- Chapter 5: Nicotine specific nitrosamines have played a role. 1. The evidence is sufficient to infer that at high-enough 4. The evidence shows that the decline of squamous cell doses nicotine has acute toxicity. carcinoma follows the trend of declining smoking prevalence. 2. The evidence is sufficient to infer that nicotine acti- vates multiple biological pathways through which Liver Cancer smoking increases risk for disease. 1. - The evidence is sufficient to infer a causal relation The evidence is sufficient to infer that nicotine expo - 3. ship between smoking and hepatocellular carcinoma. sure during fetal development, a critical window for brain development, has lasting adverse consequences Colorectal Cancer for brain development. The evidence is sufficient to infer a causal relation - 1. ship between smoking and colorectal adenomatous The evidence is sufficient to infer that nicotine 4. polyps and colorectal cancer. adversely affects maternal and fetal health during - pregnancy, contributing to multiple adverse out Prostate Cancer comes such as preterm delivery and stillbirth. The evidence is suggestive of no causal relationship 1. The evidence is suggestive that nicotine exposure 5. between smoking and the risk of incident prostate during adolescence, a critical window for brain devel - cancer. opment, may have lasting adverse consequences for brain development. 2. The evidence is suggestive of a higher risk of death from prostate cancer in smokers than in nonsmokers. The evidence is inadequate to infer the presence or 6. absence of a causal relationship between exposure to - In men who have prostate cancer, the evidence is sug 3. nicotine and risk for cancer. gestive of a higher risk of advanced-stage disease and less-well-differentiated cancer in smokers than in nonsmokers, and—independent of stage and histo - logic grade—a higher risk of disease progression. Chapter 6: Cancer Breast Cancer Lung Cancer 1. The evidence is sufficient to identify mechanisms by The evidence is sufficient to conclude that the risk 1. which cigarette smoking may cause breast cancer. - of developing adenocarcinoma of the lung from ciga rette smoking has increased since the 1960s. 2. The evidence is suggestive but not sufficient to infer a causal relationship between tobacco smoke and The evidence is sufficient to conclude that the 2. breast cancer. increased risk of adenocarcinoma of the lung in - smokers results from changes in the design and com 3. The evidence is suggestive but not sufficient to infer position of cigarettes since the 1950s. a causal relationship between active smoking and breast cancer. 8 Chapter 1

38 The Health Consequences of Smoking —50 Years of Progress The evidence is sufficient to infer that severe 4. 4. The evidence is suggestive but not sufficient to infer a causal relationship between exposure to secondhand 1-antitrypsin deficiency and cutis laxa are genetic α tobacco smoke and breast cancer. causes of chronic obstructive pulmonary disease. Adverse Health Outcomes in Cancer Patients Asthma and Survivors 1. The evidence is suggestive but not sufficient to infer 1. In cancer patients and survivors, the evidence is suf - a causal relationship between active smoking and the - incidence of asthma in adolescents. ficient to infer a causal relationship between ciga rette smoking and adverse health outcomes. Quitting 2. smoking improves the prognosis of cancer patients. The evidence is suggestive but not sufficient to infer a causal relationship between active smoking and exac - 2. erbation of asthma among children and adolescents. In cancer patients and survivors, the evidence is suf - ficient to infer a causal relationship between cigarette smoking and increased all-cause mortality and can - The evidence is suggestive but not sufficient to infer 3. cer-specific mortality. a causal relationship between active smoking and the incidence of asthma in adults. In cancer patients and survivors, the evidence is suf 3. - - 4. The evidence is sufficient to infer a causal relation ficient to infer a causal relationship between cigarette ship between active smoking and exacerbation of smoking and increased risk for second primary can - asthma in adults. cers known to be caused by cigarette smoking, such as lung cancer. Tuberculosis 4. In cancer patients and survivors, the evidence is sug - The evidence is sufficient to infer a causal relation - 1. gestive but not sufficient to infer a causal relation- ship between smoking and an increased risk of Myco- ship between cigarette smoking and (1) the risk of disease. bacterium tuberculosis recurrence, (2) poorer response to treatment, and (3) increased treatment-related toxicity. 2. The evidence is sufficient to infer a causal relationship between smoking and mortality due to tuberculosis. Chapter 7: Respiratory Diseases The evidence is suggestive of a causal relationship 3. between smoking and the risk of recurrent tubercu - Chronic Obstructive Pulmonary Disease losis disease. The evidence is sufficient to infer that smoking is the 1. 4. The evidence is inadequate to infer the presence or dominant cause of chronic obstructive pulmonary absence of a causal relationship between active smok- disease (COPD) in men and women in the United ing and the risk of tuberculosis infection. States. Smoking causes all elements of the COPD phenotype, including emphysema and damage to the 5. The evidence is inadequate to infer the presence or airways of the lung. absence of a causal relationship between exposure to - secondhand smoke and the risk of tuberculosis infec - Chronic obstructive pulmonary disease (COPD) mor 2. tion. tality has increased dramatically in men and women since the 1964 Surgeon General’s report. The number The evidence is inadequate to infer the presence or 6. of women dying from COPD now surpasses the num- absence of a causal relationship between exposure to ber of men. secondhand smoke and the risk of tuberculosis dis - ease. The evidence is suggestive but not sufficient to infer 3. that women are more susceptible to develop severe chronic obstructive pulmonary disease at younger ages. 9 Introduction, Summary, and Conclusions

39 Surgeon General’s Report Idiopathic Pulmonary Fibrosis 2. The evidence is insufficient to infer the presence or absence of a causal relationship between maternal 1. The evidence is suggestive but not sufficient to infer - prenatal smoking and anxiety and depression in chil a causal relationship between cigarette smoking and dren. idiopathic pulmonary fibrosis. The evidence is insufficient to infer the presence or 3. absence of a causal relationship between maternal Chapter 8: Cardiovascular Disease prenatal smoking and Tourette syndrome. 1. The evidence is sufficient to infer a causal relation - The evidence is insufficient to infer the presence or 4. ship between exposure to secondhand smoke and absence of a causal relationship between maternal increased risk of stroke. prenatal smoking and schizophrenia in her offspring. 2. The estimated increase in risk for stroke from expo - The evidence is insufficient to infer the presence or 5. sure to secondhand smoke is about 20-30%. absence of a causal relationship between maternal prenatal smoking and intellectual disability. - The evidence is sufficient to infer a causal relation 3. ship between the implementation of a smokefree law Ectopic Pregnancy or policy and a reduction in coronary events among The evidence is sufficient to infer a causal relation 1. - people younger than 65 years of age. ship between maternal active smoking and ectopic pregnancy. The evidence is suggestive but not sufficient to infer 4. a causal relationship between the implementation of Spontaneous Abortion a smokefree law or policy and a reduction in cerebro - vascular events. The evidence is suggestive but not sufficient to infer a 1. causal relationship between maternal active smoking The evidence is suggestive but not sufficient to infer 5. and spontaneous abortion. a causal relationship between the implementation of a smokefree law or policy and a reduction in other Male Sexual Function heart disease outcomes, including angina and out-of- The evidence is sufficient to infer a causal relation 1. - hospital sudden coronary death. ship between smoking and erectile dysfunction. Chapter 9: Reproductive Outcomes Chapter 10: Other Specific Congenital Malformations Outcomes 1. - The evidence is sufficient to infer a causal relation Eye Disease: Age-Related Macular Degeneration ship between maternal smoking in early pregnancy and orofacial clefts. - The evidence is sufficient to infer a causal relation 1. ship between cigarette smoking and neovascular and The evidence is suggestive but not sufficient to infer 2. atrophic forms of age-related macular degeneration. a causal relationship between maternal smoking in early pregnancy and clubfoot, gastroschisis, and atrial The evidence is suggestive but not sufficient to infer 2. septal heart defects. that smoking cessation reduces the risk of advanced age-related macular degeneration. Neurobehavioral Disorders of Childhood Dental Disease The evidence is suggestive but not sufficient to infer a 1. causal relationship between maternal prenatal smok - 1. The evidence is suggestive but not sufficient to infer a ing and disruptive behavioral disorders, and attention causal relationship between active cigarette smoking deficit hyperactivity disorder in particular, among and dental caries. children. 10 Chapter 1

40 —50 Years of Progress The Health Consequences of Smoking Inflammatory Bowel Disease 2. The evidence is suggestive but not sufficient to infer a causal relationship between exposure to tobacco The evidence is suggestive but not sufficient to infer 1. smoke and dental caries in children. a causal relationship between cigarette smoking and Crohn’s disease. 3. The evidence is suggestive but not sufficient to infer a causal relationship between cigarette smoking and 2. The evidence is suggestive but not sufficient to infer a failure of dental implants. causal relationship between cigarette smoking and a protective effect for ulcerative colitis. Diabetes 1. The evidence is sufficient to infer that cigarette smok - ing is a cause of diabetes. Chapter 11: General Morbidity and All-Cause Mortality 2. The risk of developing diabetes is 30–40% higher for active smokers than nonsmokers. 1. The evidence is sufficient to infer a causal relation - ship between smoking and diminished overall health. 3. There is a positive dose-response relationship between Manifestations of diminished overall health among the number of cigarettes smoked and the risk of devel- smokers include self-reported poor health, increased oping diabetes. absenteeism from work, and increased health care utilization and cost. Immune Function and Autoimmune Disease 1. The evidence is sufficient to infer that components of The evidence is sufficient to infer that cigarette 2. cigarette smoke impact components of the immune smoking increases risk for all-cause mortality in men system. Some of these effects are immune activating and women. and others are immune-suppressive. The evidence is sufficient to infer that the relative 3. - The evidence is sufficient to infer that cigarette smok 2. risk of dying from cigarette smoking has increased ing compromises the immune system and that altered over the last 50 years in men and women in the immunity is associated with increased risk for pulmo- United States. nary infections. 3. The evidence is sufficient to infer that cigarette smoke Chapter 12: Smoking-Attributable compromises immune homeostasis and that altered Morbidity, Mortality, and Economic - immunity is associated with an increased risk for sev eral disorders with an underlying immune diathesis. Costs Rheumatoid Arthritis 1. Since the first Surgeon General’s report on smoking and health in 1964, there have been more than 20 The evidence is sufficient to infer a causal relationship 1. million premature deaths attributable to smoking and between cigarette smoking and rheumatoid arthritis. exposure to secondhand smoke. Smoking remains the leading preventable cause of premature death in the - The evidence is sufficient to infer that cigarette smok 2. United States. ing reduces the effectiveness of the tumor necrosis factor-alpha (TNF- α ) inhibitors. Despite declines in the prevalence of current smoking, 2. the annual burden of smoking-attributable mortality Systemic Lupus Erythematosus in the United States has remained above 400,000 for The evidence is inadequate to infer the presence or 1. more than a decade and currently is estimated to be absence of a causal relationship between cigarette about 480,000, with millions more living with smok - smoking and systemic lupus erythematosus (SLE), ing-related diseases. the severity of SLE, or the response to therapy for SLE. 11 Introduction, Summary, and Conclusions

41 Surgeon General’s Report Due to the slow decline in the prevalence of current 3. Chapter 14: Current Status of smoking, the annual burden of smoking-attributable Tobacco Control mortality can be expected to remain at high levels for - decades into the future, with 5.6 million youth cur The evidence is sufficient to conclude that there are 1. rently 0 to 17 years of age projected to die prema - diverse tobacco control measures of proven efficacy at turely from a smoking-related illness. the population and individual levels. Annual smoking-attributable economic costs in the 4. 2. The evidence is sufficient to conclude that advertising United States estimated for the years 2009–2012 were and promotional activities by the tobacco companies between $289–332.5 billion, including $132.5–175.9 cause the onset and continuation of smoking among billion for direct medical care of adults, $151 billion adolescents and young adults. for lost productivity due to premature death estimated from 2005–2009, and $5.6 billion (in 2006) for lost 3. Tobacco product regulation has the potential to productivity due to exposure to secondhand smoke. contribute to public health through reductions in tobacco product addictiveness and harmfulness, and by preventing false or misleading claims by the Chapter 13: Patterns of Tobacco Use tobacco industry of reduced risk. Among U.S. Youth, Young Adults, 4. The evidence is sufficient to conclude that litigation and Adults against tobacco companies has reduced tobacco use in the United States by leading to increased product - In the United States, the prevalence of current ciga 1. prices, restrictions on marketing methods, and mak - rette smoking among adults has declined from 42% ing available industry documents for scientific analy - in 1965 to 18% in 2012. sis and strategic awareness. 2. The prevalence of current cigarette smoking declined 5. The evidence is sufficient to conclude that increases first among men (between 1965 and the 1990s), and in the prices of tobacco products, including those then among women (since the 1980s). However, resulting from excise tax increases, prevent initiation declines in the prevalence of smoking among adults of tobacco use, promote cessation, and reduce the (18 years of age and older) have slowed in recent years. prevalence and intensity of tobacco use among youth and adults. Most first use of cigarettes occurs by 18 years of 3. age (87%), with nearly all first use by 26 years of 6. The evidence is sufficient to conclude that smokefree age (98%). indoor air policies are effective in reducing exposure to secondhand smoke and lead to less smoking among Very large disparities in tobacco use remain across 4. covered individuals. racial/ethnic groups and between groups defined by educational level, socioeconomic status, and region. 7. The evidence is sufficient to conclude that mass - media campaigns, comprehensive community pro In the United States, there are now more former 5. grams, and comprehensive statewide tobacco con - smokers than there are current smokers. More than trol programs prevent initiation of tobacco use and half of all ever smokers have quit smoking. reduce the prevalence of tobacco use among youth and adults. The rate of quitting smoking among recent birth 6. cohorts has been increasing, and interest in quitting 8. The evidence is sufficient to conclude that tobacco is high across all segments of society. cessation treatments are effective across a wide popu - lation of smokers, including those with significant 7. Patterns of tobacco use are changing, with more mental and physical comorbidity. intermittent use of cigarettes and an increase in use of other products. 12 Chapter 1

42 The Health Consequences of Smoking —50 Years of Progress by the Centers for Disease Control and Prevention. Chapter 15: The Changing Implementing tobacco control policies and programs Landscape of Tobacco Control— Ending the Tobacco Epidemic: A as recommended by Current Status and Future by the U.S. Depart- Tobacco Control Strategic Plan ment of Health and Human Services and the End- Directions ing the Tobacco Problem: A Blueprint for the Nation by the Institute of Medicine on a sustained basis at Together, experience since 1964 and results from 1. high intensity would accelerate the decline of tobacco models exploring future scenarios of tobacco control use in youth and adults, and also accelerate progress indicate that the decline in tobacco use over coming toward the goal of ending the tobacco epidemic. decades will not be sufficiently rapid to meet targets. The goal of ending the tragic burden of avoidable New “end game” strategies have been proposed with 3. disease and premature death will not be met quickly the goal of eliminating tobacco smoking. Some of enough without additional action. these strategies may prove useful for the United States, particularly reduction of the nicotine con - Evidence-based tobacco control interventions that 2. tent of tobacco products and greater restrictions - are effective continue to be underutilized and imple on sales (including bans on entire categories of to- mented at far below funding levels recommended bacco products). 13 Introduction, Summary, and Conclusions

43 Surgeon General’s Report References U.S. Department of Health and Human Services. The Family Smoking Prevention and Tobacco Control Act, Public Law 111-31, 123 Health Consequences of Involuntary Exposure to U.S. Statutes at Large 1776 Tobacco Smoke: A Report of the Surgeon General (2009). . Atlanta (GA): U.S. Department of Health and Human Federal Cigarette Labeling and Advertising Act of 1965, Public Law 89-92, U.S. Statutes at Large Services, Centers for Disease Control and Prevention, 79 (1965). Parascandola J. The Surgeons General and Smoking. Coordinating Center for Health Promotion, National Pub- lic Health Reports - Center for Chronic Disease Prevention and Health Pro 1997;112:439-41. Public Law motion, Office on Smoking and Health, 2006. Public Health Cigarette Smoking Act of 1969, U.S. Department of Health and Human Services. , 84 U.S. Statutes at Large 87 (1970). 91-222 Ending the Tobacco Epidemic: A Tobacco Control Strategic 100 events that shifted history. February 24, USA Today. Action Plan for the U.S. Department of Health and 1999. U.S. Department of Health and Human Services. The Human Services. Washington: Office of the Assistant Secretary for Health, 2010a. Health Consequences of Involuntary Smoking. A U.S. Department of Health and Human Services. Report of the Surgeon General. Rockville (MD): U.S. How Department of Health and Human Services, Public Tobacco Smoke Causes Disease—The Biology and Health Service, Centers for Disease Control, Center Behavioral Basis for Tobacco-Attributable Disease: A Report of the Surgeon General . Atlanta (GA): U.S. - for Health Promotion and Education, Office on Smok Department of Health and Human Services, Centers ing and Health, 1986. DHHS Publication No. (CDC) for Disease Control and Prevention, National Center 87-8398. for Chronic Disease Prevention and Health Promotion, The U.S. Department of Health and Human Services. Office on Smoking and Health, 2010b. - Health Consequences of Smoking: Nicotine Addic Pre- U.S. Department of Health and Human Services. Atlanta (GA): tion. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public venting Tobacco Use Among Youth and Young Adults: Health Service, Centers for Disease Control, National A Report of the Surgeon General. Atlanta (GA): U.S. - Department of Health and Human Services, Centers Center for Chronic Disease Prevention and Health Pro for Disease Control and Prevention, National Center motion, Office on Smoking and Health, 1988. DHHS Publication No. (CDC) 88-8406. for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2012. U.S. Department of Health and Human Services. Reducing Atlanta Smok- Tobacco Use. A Report of the Surgeon General. U.S. Department of Health, Education, and Welfare. ing and Health: Report of the Advisory Committee to (GA): U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National the Surgeon General of the Public Health Service. Washington: U.S. Department of Health, Education, - Center for Chronic Disease Prevention and Health Pro and Welfare, Public Health Service, Center for Disease motion, Office on Smoking and Health, 2000. Control, 1964. PHS Publication No. 1103. The U.S. Department of Health and Human Services. Health Consequences of Smoking: A Report of the Sur - U.S. Department of Health, Education, and Welfare. . Atlanta (GA): U.S. Department of Health geon General Smoking and Health. A Report of the Surgeon Gen - - and Human Services, Centers for Disease Control and Washington: U.S. Department of Health, Educa eral. tion, and Welfare, Public Health Service, Office of the Prevention, National Center for Chronic Disease Pre - vention and Health Promotion, Office on Smoking and Assistant Secretary for Health, Office on Smoking and Health, 2004. Health, 1979. DHEW Publication No. (PHS) 79-50066. Chapter 1 14

44 Chapter 2 Fifty Years of Change 1964–2014 Introduction 17 Tobacco Control: Before the 1964 Report 19 Scientific Judgment and the 1964 Report 21 Remedial Action and Change Following the 1964 Report 23 Passive Smoking and Environmental Change 26 Nicotine and Addiction 30 Denormalization and the Tobacco Industry 31 Summary 33 35 References 15

45

46 The Health Consequences of Smoking —50 Years of Progress Introduction Tobacco, a New World plant, was used by the native nomic, gender, race, and ethnicity boundaries. Cigarette smoking was widely accepted, highly prevalent, and not peoples of the Americas for millennia. Brought to the Old discouraged in homes, and it took place in public spaces of World by Christopher Columbus, tobacco and tobacco products soon spread worldwide. The manufactured ciga all kinds, including hospitals, restaurants, airplanes, and - medical conferences (Brandt 1990). Today, the prevalence rette has been the dominant form of tobacco use in the United States for only a century (Figure 2.1), surpass - of smoking among U.S. adults is about 20% (see Chapter 13, “Patterns of Tobacco Use Among U.S. Youth, Young ing other forms of use as the modern tobacco industry Adults, and Adults”), and state and local laws have prohib - was shaped by James B. Duke and his American Tobacco Company (Chandler 1977). During that century, referred ited smoking in workplaces, restaurants, and bars in many to as “The Cigarette Century” (Brandt 2007), there was a regions of the country (see Chapter 14, “Current Status of Tobacco Control”). The majority of households are smoke- sharp rise in tobacco consumption to a peak in the 1960s free and smoking is banned on airplanes worldwide (U.S. and then a decline that has continued over the last three USDHHS ] Department of Health and Human Services [ decades. This chapter addresses why this rise and fall of - 2006). Moreover, the rise and fall of smoking-caused dis cigarette smoking occurred, giving emphasis to the half- century since the 1964 report of the Advisory Committee eases and premature deaths during the twentieth century to the Surgeon General, , and to the generally follow patterns of changing tobacco-use behav Smoking and Health - impact of the reports of the Surgeon General on tobacco ior, albeit several decades later. Although there had been previous statements on the use in the United States. harms of using tobacco, the 1964 report was significant for This chapter provides a perspective on the tobacco epidemic, setting a context for this anniversary report by providing the most thorough and comprehensive review describing some of the most critical “lessons learned” with up to that time. However, translating this knowledge into regard to the factors driving tobacco use and the strategies action to benefit public health was not a simple or direct process. At the time of release of the 1964 report, the for ending it. The following chapter describes the Surgeon General’s reports, including the approach used to compile tobacco industry had a powerful influence and attempted to minimize the impact of the report using a broad set and synthesize scientific evidence to reach conclusions Kluger 1996; Brandt 2007; Proctor 2011) . (see Chap- of strategies ( that has been the foundation of these reports ter 3, “Producing the Surgeon General’s Report From That influence has now greatly declined, diminished by 1964–2014: Process and Purpose”) - many factors, including trends in American culture, poli . Two major sections follow: the first provides a comprehensive updating of the tics, economics, health care, and social life. This chapter health consequences of active smoking and exposure to addresses how the evolving scientific evidence on tobacco secondhand smoke, updating the many previous reviews; has been a key driver of the changes that have led to a and the second details the current status of the epidemic, dramatic shift in social norms around cigarette smoking. reviews the policy approaches that have proved effective During this same time span, 1964–2014, there have for tobacco control, and offers a strategy and a vision for been striking changes in mortality rates from major dis - eases and substantial improvements in life expectancy (see bringing this long-running epidemic to an end—the so- called “end game.” Chapter 4, “Advances in Knowledge of the Health Conse- In offering a perspective on the long and complex quences of Smoking: From 1964–2014”). These changes story of the tobacco epidemic, this chapter is necessarily have been driven by many factors, including patterns of limited in its historical detail and does not follow the for tobacco use across the twentieth century to the present. - ) dropped CVD Mortality from cardiovascular diseases ( mat of a detailed review of evidence that is typical of these reports. Lengthy and detailed historical accounts are sharply and progressively, and rates for a number of can - available elsewhere (Kluger 1996; Brandt 2007; Proctor cers peaked and began to decline, most notably in men. By 2011). Americans’ behaviors, perceptions, attitudes, and contrast, mortality from chronic obstructive pulmonary beliefs toward the cigarette have changed dramatically disease steadily climbed. Changes in the prevalence of since 1964 when the first report of the Surgeon General tobacco smoking contributed to these shifts, but patterns of other risk factors also changed over the last 50 years, as on smoking and health was released. At the time, 40% of programs addressed hypertension and other risk factors Americans were regular smokers, with the majority of men for CVD, and medications became available that reduced (53%) and about one-third of women being regular smok- CVD, such as statins (Feinlieb et al. 1979; Stern 1979; ers (U.S. Department of Health, Education, and Welfare [USDHEW] 1979). The smoking habit crossed socioeco Jemal et al. 2005; Ford and Capewell 2011). - 17 Fifty Years of Change 1964–2014

47 Surgeon General’s Report Adult* per capita cigarette consumption and major smoking and health events, United States, 1900–2012 Figure 2.1 Sources: Adapted from Warner 1985 with permission from Massachusetts Medical Society, ©1985; U.S. Department of Health and Human Services 1989; Creek et al. 1994; U.S. Department of Agriculture 2000; U.S. Census Bureau 2013; U.S. Department of the Treasury 2013. Adults ≥18 years of age as reported annually by the Census Bureau. * Chapter 2 18

48 —50 Years of Progress The Health Consequences of Smoking Tobacco Control: Before the 1964 Report To understand the transformative consequences of removed from the tobacco harmful corrosive acids (pun - gent irritants) present in cigarettes manufactured in the the 1964 report for tobacco control, this chapter begins old fashioned way” ( with a description of the developments in tobacco con FTC - , p. LBA -2); “Smoking Camels trol before 1964. Cigarette smoking grew rapidly in early stimulates the natural flow of digestive fluids ... increases alkalinity” (Camel, p. Kool menthol twentieth century America with the arrival of technology 1964). FTC -1a) ( LBA cigarettes, characterized by the cooling effect of this addi for mass production and the development of a consumer - tive, were offered to nose and throat specialists to hand culture and effective advertising and promotion on an unprecedented national scale (Figure 2.1) (Kluger 1996). out to their patients “suffering from colds and kindred At the same time, there was strong opposition to this trend 1948, Bates No. 400566440/6490, disorders” ( Information from some groups, but early condemnations were often p. 9). FTC brought legal action against each of the major based on concerns about adverse moral and social impact cigarette companies during the 1940s in an effort to curb rather than specific health effects (Best 1979). Addition- health claims in advertising, resulting in a series of cease- ally, concerns focused on specific groups seen to be espe and-desist orders. However, the agency’s power to con - - trol such advertising claims at the time was limited (FTC cially vulnerable to the social and psychological effects FTC 735 (1950); FTC 46 v. P. Lorillard Co., FTC 1950a,b; of chronic cigarette smoking, notably youth and women. FTC v. American v. R.J. Reynolds Co ., 46 FTC 706 (1950); And unrestricted tobacco advertising, often with health- 1393 (1951); related claims, was seen as taking unfair advantage of Tobacco Co. , 47 FTC FTC v. Philip Morris & ). those who were most vulnerable. In the first two decades 703 (1952) FTC , 49 Co. By the 1930s, however, American scholars and activ of the century, an organized antitobacco effort developed, - ists had become aware of increasing cancer death rates. composed of temperance advocates, religious leaders, and Statisticians in the insurance industry, such as Freder - . They were alarmed by the Kluger 1996) health reformers ( increase in cigarette smoking among youth and believed ick L. Hoffman at Prudential Insurance Company, had it to be associated with the abuse of alcohol and narcotic - amassed statistical data documenting the growing influ drugs. During this period, a total of 15 states ence of cancer since the turn of the century, and volun - banned the tary organizations like the American Cancer Society had sale, manufacture, possession, or use of cigarettes. Many - other states considered such legislation, and municipali been using these data to bring public attention to the cancer problem (Patterson 1987). In the late 1930s, the ties imposed additional restrictions on advertising, smok- government published cancer mortality statistics from ing near school buildings, and women smoking in public 1900–1935 based on U.S. Census data and subsequently (Tate 1999). - cause-specific mortality was tracked, providing an ongo - Warnings about tobacco were offered by the Sur geon General before 1964. In 1929, Surgeon General ing picture of mortality trends (Gover 1939). Hugh S. Cumming warned about the hazards of tobacco Complementing these mortality statistics, some cli- claiming that excessive smoking caused nervousness, nicians described a growing clinical experience with lung - insomnia, and other ill effects in young women (Burn cancer patients and the surgical treatment of the disease ham 1989). Cumming warned that smoking could lower by pneumonectomy, removal of a lung. Thoracic surgeon the “physical tone” of the nation. Like many physicians at Alton Ochsner recounted being called as a medical student in 1910 to see an autopsy of a patient with lung cancer the time, he believed that women were more susceptible than men to certain injuries, especially of the nervous sys because such cases were so rare (Ochsner 1973). Several - decades later, he began to see many such patients. Och- tem. But Cumming, a smoker, distanced himself from the sner and DeBakey (1939) reported their experience with more extreme antitobacco and temperance reformers of the time (Parascandola 1997). pneumonectomy for lung cancer and proposed that smok- Although physicians generally did not see a sig ing contributed to the development of this malignancy: - - “In our opinion the increase in smoking with the univer nificant health threat for most smokers, there was grow- sal custom of inhaling is probably a responsible factor, as ing concern over cigarette advertising during the 1930s the inhaled smoke, constantly repeated over a long period and 1940s that made a wide array of unfounded health of time, undoubtedly is a source of chronic irritation to claims. In the highly competitive branded cigarette mar - ket, prominent advertising campaigns included explicit the bronchial mucosa” (p. 109). At the same time, smok - ing was clearly linked to decreased life expectancy by Pearl health claims: “Not a cough in a carload” (Old Gold) (U.S. ] 1964, p. LBA-5); “we Federal Trade Commission [ FTC (1938), based on follow-up of adults in Baltimore. 19 Fifty Years of Change 1964–2014

49 Surgeon General’s Report Yet, there was also substantial skepticism within the ) and TIRC the Tobacco Industry Research Committee ( Collection medical community about whether the seeming increase DATTA later the Council for Tobacco Research ( in cancer deaths was real or an artifact of better diagnosis. 1953), and the nationwide publication of the “Frank State - ment,” which publicly stated the industry’s commitment The rise in lung cancer, a rare disease at the beginning - to public health (Pollay Advertising Collection, n.d.). Clar of the twentieth century, drew particular scrutiny (Wits- ence Cook Little, a leading researcher and academician, chi 2001). However, the possibility of diagnostic bias was was hired in 1954 as the first head of TIRC ; he assumed a set aside through appropriate research and the continu - public position of skepticism with regard to the evidence ing rise of lung cancer deaths made such diagnostic bias USDHEW improbable (Macklin 1942; 1964). A wide range on smoking and health, seeking to create doubt about the of possible industrial and environmental causes were cited harmful effects of smoking (Brandt 2007; Proctor 2011). as possibly contributing to the increase, including road For decades, the industry followed the strategies set out tars, vehicle exhaust, and air pollution, along with tobacco in the early 1950s: denying the harms of its products, dis - crediting the scientific evidence that showed these harms, smoking (Witschi 2001). funding research that was intended to divert attention Beginning as early as the 1920s, the rise of lung can - from cigarettes, and marketing new products with implied cer prompted epidemiologic research on its causes that lower risks than existing products ( United States v. Philip was carried out in the United States and Europe. These initial studies found an association between lung cancer 2006; Brandt 2007; Proctor 2011). Morris Inc. Generally, there was little response in the medical and tobacco smoking that was repeatedly confirmed in a community to the first wave of studies on the risks of wave of research that began in the 1940s and continued in smoking. In 1953, in the midst of early reports on ciga - the 1950s (Witschi 2001). These studies were of the case- rette smoking and lung cancer, the American Medical control design, involving comparison of the frequency and AMA - ) did announce that it would stop accept intensity of smoking by people with lung cancer to smok - Association ( ing cigarette (and alcohol) advertising in its journal begin - ing among comparable people without lung cancer—the 1953). However, the Advertising Age ning January 1, 1954 ( controls. By the early 1950s, in follow-up of the strong associations found in the case-control studies, cohort move was not an indication that AMA accepted that smok - or follow-up studies were initiated that compared rates ing was hazardous, but was primarily a response to the of lung cancer occurrence or death among smokers and medical claims increasingly seen in cigarette advertising; pharmaceutical companies had reportedly complained nonsmokers. These epidemiologic studies provided the that while their claims were subject to thorough pivotal evidence on smoking and lung cancer for the 1964 AMA to scrutiny, cigarette manufacturers’ claims were not ( report of the Surgeon General. The public responded to Adver- 1953). Cigarette manufacturers were also start- tising Age the new information on smoking and lung cancer with ing to worry that overt medical claims could backfire, a slight decrease in consumption (from 1953–1954) that drawing attention to the growing evidence of harms. was quickly followed by a sharp rise (Figure 2.1). In summary, in the first half century of the cigarette - The American tobacco industry’s strategies for deal epidemic, concerns about cigarette smoking often focused ing with scientific evidence documenting the harms of its By the early on the habit’s impact on the social and moral fabric of products also originated during the 1950s. 1950s, the epidemiologic evidence on lung cancer and society. Additionally, broader fears about the booming smoking was abundant and coherent, and Wynder and - ulture and the ubiquitous advertising associ consumer c ated with it led to attempts to control or warn the public colleagues’ (1953) mouse experiments had documented that cigarette smoke condensate caused tumors confirm- about misleading advertising claims. As long as consum- ing earlier work by Angel H. Roffo (Proctor 2006). In a ers were protected from misleading claims, the decision to now well-documented effort to counter this evidence and smoke or not smoke was one that the medical community - had little to say about. But the emergence of strong evi to minimize risk to the industry, the executives of the dence related to cancer and other health risks from ciga major tobacco companies met in December 1953 and, - with the guidance of the advertising firm Hill & Knowl rette smoking during the 1950s shifted the focus to the - - scientific evidence on its health effects, setting the stage ton, devised a unified strategy that included the found for evidence-based action. ing of an industry-funded research organization, initially Chapter 2 20

50 —50 Years of Progress The Health Consequences of Smoking Scientific Judgment and the 1964 Report By the late 1950s, the amassing evidence on smok diseases and to the overall death rate” ( - USDHEW 1964, - p. 31). However, the 1964 report went beyond these ear ing and lung cancer called for public health action. The - lier reviews in its transparent methodology and depth of Surgeon General was among the first authoritative fig ures to address the public health implications of the ris - analysis, including a systematic gathering and review of the data and a synthesis of the findings for causality based ing evidence on the health risks of smoking. Before the on prior criteria. The members of the Advisory Commit - 1964 report was released, there had been several previous tee were carefully selected to identify a panel that would statements from the Surgeon General, several consensus be considered as free of any bias as to the report’s find statements from groups of public health scientists, and - ings (Parascandola 1997). Its landmark status reflects a report from the Royal College of Physicians (1962), all this approach, which made it a model, not only for identifying cigarette smoking as a cause of lung cancer (Cutler 1955; Study Group on Smoking and Health 1957). future reports of the Surgeon General, but for reviews in other fields. These reports were based largely on epidemiologic stud- The Surgeon General’s emphasis on methodology - ies, both case-control and cohort; on findings from labo ratory studies using animals and pathology studies; on merits highlighting (see Chapter 3). The report devoted two chapters to describing the working methods of the chemical identification of known carcinogens in cigarette smoke; and on analyses of large-scale patterns of cigarette group, and the criteria they employed, in making infer - consumption and disease rates (Proctor 2011). Although ences about cause and effect relationships. The Com- mittee cited five criteria for making a determination of - the case-control studies were questioned on methodologi cal grounds, evidence from several cohort studies was causation from an observed association: consistency, reported in the 1950s that confirmed the strong associa- strength, specificity, temporal relationship, and coherence USDHEW tion between smoking and lung cancer. In June 1954, the ( 1964). For lung cancer in particular, the Com - mittee discussed a range of different types of evidence in results from the first cohort assembled by the American great detail, responding to alternative explanations for the Cancer Society, which included 180,000 older men, were high risk of lung cancer in smokers, other than smoking, announced (Hammond and Horn 1958). The study showed and addressing inconsistencies in the total body of evi that heavy smokers were dying of lung cancer at a rate 5 to - dence. Although previous reviews had covered some of the 16 times higher than that of similar people who were not same material and employed similar criteria, the Advisory smokers. At the same time, similar findings were reported Committee did so in a way that was more explicit and for - from studies of British physicians (Doll and Hill 1954) and U.S. veterans (Dorn 1958, 1959). By 1959, Surgeon mal than previous inquiries. In the end, it was no single General Leroy E. Burney declared cigarette smoking “the study, but the mass of cumulative evidence from diverse principal [sic] etiological factor in the increased incidence sources that made the case for smoking as a cause of lung of lung cancer” (Burney 1959, p. 1835). The same year, a cancer irrefutable (Parascandola et al. 2006). review by leading public health scientists assessed a range This approach successfully addressed the new prob - of potential criticisms of the research findings and con lem in public health of interpreting observational findings. - cluded that the evidence was overwhelming: “if the find - - The 1950s and 1960s were a critical time for a new applica a focus on chronic rather than tion of epidemiology with ings had been made on a new agent, to which hundreds of millions of adults were not already addicted, and on one infectious diseases, an emphasis on identifying individual - which did not support a large industry, skilled in the arts risk factors for disease, and the use of advanced quantita of mass persuasion, the evidence for the hazardous nature Chronic tive methodology (Morris 1957; Lilienfield 1978). - of the agent would generally be regarded as beyond dis diseases such as cancer and heart disease required a new pute” (Cornfield et al. 1959, p. 198). approach to understanding their etiology. Unlike tradi- Thus, the 1964 report’s most noteworthy finding— tional infectious disease research, where a single neces - sary causal agent or organism could be identified and “Cigarette smoking is causally related to lung cancer in studied in the laboratory and in the population, cancer was men; the magnitude of the effect of cigarette smoking far associated with a wide range of exposures and agents and outweighs all other factors. The data for women, though developed over decades. The picture was quite similar for less extensive, point in the same direction” ( USDHEW 1964, p. 31)—had been anticipated in prior reviews. The cardiovascular diseases and chronic lung disease. Human experiments could not be carried out to determine if par - - report also concluded that “...cigarette smoking con ticular agents had causal effects; instead, risk factors were tributes substantially to mortality from certain specific 21 Fifty Years of Change 1964–2014

51 Surgeon General’s Report “objective assessment of the nature and magnitude of the identified through observational epidemiologic research which is inherently subject to various sources of bias. health hazard” (USDHEW 1964, p. 8). The 10 Committee As a result, there was substantial debate about what members were selected from a list of about 150 eminent - - physicians and biomedical scientists from a variety of dif type of evidence was needed to declare cigarette smok ferent disciplines. Major medical associations, volunteer ing a cause of lung cancer (Brandt 1990; Parascandola public health organizations, the Tobacco Institute, the 2004). Some advocated for a narrow view of cause and Food and Drug Administration (FDA), FTC, and the Presi - effect, insisting it must be demonstrated that cigarette smoking is uniquely linked to lung cancer, the link must dent’s Office of Science and Technology were all given the - be demonstrated in a randomized trial, or additional evi opportunity to remove a name from the list for any rea - son (Terry 1983). Anyone who had taken a prior public dence demonstrating underlying biological mechanisms position on any question of smoking and health would be was required (Yerushalmy and Palmer 1959; Parascandola eliminated from the list (Terry 1983). The members of the 2011). The tobacco industry took advantage of the meth - Committee held their meetings at the National Library odologic divide, insisting that epidemiology and statistics - of Medicine in Bethesda, Maryland, with their delibera alone could not prove cause and effect and that a detailed tions under strict secrecy and documents under lock and understanding of the mechanisms of cancer etiology was key. Even the Surgeon General himself knew nothing of required to support such claims (Little 1961). The dis - missal of epidemiologic evidence as imperfect was a strat - the details of their work until the final report was being egy used repeatedly by the tobacco industry, particularly printed (Terry 1983). This approach, which did not directly in attempting to thwart the consequences of the studies urney’s statements during involve Terry, contrasted with B the 1950s, which had been presented as the “opinions” of linking exposure to secondhand smoke to lung cancer and - the Surgeon General and senior U.S. Public Health Ser other diseases (Kluger 1996; Brandt 2007; Proctor 2011). However, as the evidence on smoking and disease accumu- vice (PHS) leaders (U.S. Congress 1957). The process used for the report marks the beginning lated throughout the 1950s, many public health scientists of a new role for scientific experts in the United States. increasingly insisted that such “logically rigorous” proof Allan M. Brandt (2007) refers to the era of “procedural of causation, requiring demonstration of a necessary and science” and Robert N. Proctor calls the report a prod - sufficient cause, was not required (Cornfield et al. 1959). In the 1964 Surgeon General’s report, the Advisory - uct of an “administrative rather than a scientific consen Committee endorsed this conceptual approach, explain- sus” (Proctor 2011, p. 236). That is, the crucial science relied upon by the Advisory Committee had been already ing that, in the absence of experimentation, the “causal significance of an association is a matter of judgment” published; the authority of the report also rested on the - (USDHEW 1964, p. 20). Additionally, they employed a characteristics of the process used in reaching its conclu sions, which assured that conclusions were reached by more flexible, pragmatic definition of “cause,” which - - focused not on identifying a unique necessary and suf considering the full range of evidence available and judg ing the evidence in a transparent and consistent frame - ficient cause, as for infectious diseases, but on finding The explicit appeal to the process and criteria for work. the modifiable multifactorial determinants of health out- comes with the ultimate aim of supporting prevention, an judgment was novel at the time, but has since come to approach which was to be further developed by an emerg- - be standard practice for evidence reviews in controver sial areas of medicine and public health. The industry’s ing discipline of chronic disease epidemiology (MacMahon et al. 1960). The criteria for evidence evaluation offered documents provide insights into how the industry viewed flexibility for evidence interpretation that avoided the the 1964 report from the planning process through the rigid requirements of the Henle-Koch postulates long report’s development and release (Allen 1962; Cullman used for infectious organisms (Evans 1976, 1978, 1993; 1962; Hockett and Thompson 1962; Bass 1963; Hill & Susser 1995). Knowlton 1963; Council for Tobacco Research 1964; The mechanism by which the report was produced Cullman 1964; Haas 1964; Pacey 1964; Wakeham 1964; gave it a status and authority beyond the previous reviews. Weissman 1964). Notably, the industry was treated as When Surgeon General Luther Terry initiated the effort a stakeholder and given the opportunity to make rec - in 1962 at the request of President John F. Kennedy, he - ommendations on members of the Advisory Commit stated that the group would not conduct any new research tee and to provide research materials to the Committee (Terry 1983). or make any recommendations, but would provide an Chapter 2 22

52 The Health Consequences of Smoking —50 Years of Progress Remedial Action and Change Following the 1964 Report The 1964 Surgeon General’s report concluded that threat. Consequently, the release of the report was care- “Cigarette smoking is a health hazard of sufficient impor fully managed with the media response in mind. The press - conference was held on a Saturday to minimize the effects tance in the United States to warrant appropriate reme- dial action” (USDHEW 1964, p. 33). However, the report of the report on the stock market and to ensure coverage in the Sunday newspapers (Parascandola 1997). All of the did not specifically state what actions should be taken and lacking any precedent at the time, it was not immediately approximately 200 reporters attending were required to Surgeon General remain for the entire session. Each was given a copy of the clear what form this action should take. Luther Terry had initially outlined two distinct phases final report and allowed to study it for an hour. Report- ers were then permitted to question the Surgeon General of inquiry. The first was an expert committee to provide and the Administration. Finally, the doors were opened an “objective assessment of the nature and magnitude of the health hazard” ( USDHEW 1964, p. 8). The second and reporters raced out to file their stories (Parascandola 1997). The report received enormous publicity. phase, which would provide recommendations for action Newsweek lauded it as “monumental” and subsequently the report and require a different range of expertise, would follow, although this effort never fully materialized. has been named by the New York Public Library as one During the 1950s, federal public health officials saw of the top 100 books of the twentieth century (Diefendorf - their role as limited. Alexander Langmuir, who pioneered 1996). Terry made the Surgeon General into a public fig in disease surveillance at the Communicable Disease ure, no longer an anonymous government official; his Center (now known as the Centers for Disease Control use of the media to address national public health issues would be taken up and further developed by later Sur and Prevention [CDC]), viewed the role of public health - researchers as generating evidence for others who make geons General. Nevertheless, while the report was to lead to action, policy decisions: “When major health problems arise, someone must make decisions. This is not the primary health officials and political leaders still saw a carefully responsibility of the epidemiologist. Administrative and circumscribed role for federal intervention on smoking and health. Secretary of Health, Education, and Welfare political as well as technical considerations must also be brought to bear. It is the epidemiologists’ function to Anthony J. Celebrezze had already stated his views on the government’s responsibilities even before the Com - get the facts to the decision makers” (Langmuir 1963, p. 191). Testifying before Congress in 1957, Surgeon General mittee began its work: “I firmly believe that it is not the proper role of the federal government to tell citizens to Leroy Burney insisted it was the role of PHS to present the facts as they became available to state health agencies, The proposals that stop smoking” (Toth 1962, p. 20). and sometimes the national media, but not to undertake emerged were primarily aimed at ensuring that consum - e added, an organized national educational campaign. H - ers had accurate information with which to make deci “We should not go all out on a campaign and put stickers sions about their own behavior. At the time, of course, the addictive potential of nicotine in tobacco smoke was on cigarettes and certain other things” (Burney 1957b, p. When Burney released official statements on smoking not generally known. Government had a role in protect 24). - ing consumers from industry abuses, such as fraudulent and health in 1957 and 1959, they appeared in academic advertising, but not in intervening to change consumer medical journals and were sent out to state public health officers and to AMA, but not to the general public. The behavior. For example, Senator Maurine Neuberger urged statements received little public attention. Thus, although require cigarette manufacturers to state tar and to FTC Burney (1957a) was unequivocal on the weight of the evi- nicotine yields on advertisements and cigarette packages dence, this judgment on the association of smoking with to “stimulate the development of less hazardous cigarettes lung cancer did not necessarily translate into a call for and facilitate intelligent choice between competing brands New York Times action, even action to educate the public ( on the basis of relative safety” (Neuberger 1964, p. 1). But This approach contrasted sharply 1957; Fritschler 1969). FDA regulatory authority over tobacco proposals to give with Luther Terry’s dramatic, nationally televised press products were rejected by federal public health officials as conference in 1964. The 1964 report spoke with far more impractical and contrary to what the public would accept certainty than Burney’s earlier publications, which were (U.S. Congress 1964, 1965). brief and had a more limited evidence base. Additionally, Congress did enact legislation to educate consum- the 1964 report had been requested by President Kennedy ers about the hazards of smoking. In 1965, the Federal and it was an unprecedented review of a public health mandated Cigarette Labeling and Advertising Act of 1965 23 Fifty Years of Change 1964–2014

53 Surgeon General’s Report - the first Surgeon General’s warning to appear on cigarette F. Banzhaf III, a consumer lawyer. In 1967, Banzhaf suc - packages: “Caution: Cigarette Smoking May Be Hazard - cessfully petitioned the Federal Communications Com 1 ) to apply the Fairness Doctrine to cigarette - FCC mission ( ous to Your Health.” It called for an annual report to Con gress on the health consequences of smoking and for the advertising to counter the tobacco industry’s advertising FCC Secretary of Health to make recommendations for needed messages ( , 405 F.2d 1082, 1086 [D.C. Banzhaf v. legislation. In October 1965, PHS created the National Cir. 1968], cert. denied, 396 U.S. 842, 90 S. Ct. 50 [1969]; USDHHS 2000). After a court struggle, the national net - Clearinghouse on Smoking and Health. This office was to works were forced to air antismoking advertising spots play a key role in the development of the first 10 Surgeon in prime time, giving tens of millions of dollars’ worth of General’s reports (1967–1978) as well as development of - free airtime to antismoking efforts. In 1968, 1,300 antito national informational and educational programs about - bacco messages were aired by the three major networks the risks of smoking. However, at the same time it prohib ited FTC from taking any new regulatory action to control (Lewit et al. 1982). These public service announcements may have contributed to a reduction of overall consump cigarette advertising for 4 years. Contemporary observers - explained that the tobacco industry had decided it was tion; per capita cigarette consumption fell from 4,197 in in their interest to accept the warning label in exchange 1966 to 3,969 in 1970 (Figure 2.1). The effect was short- lived, however, as tobacco companies were mandated to for halting any regulatory efforts (Drew 1965). However, subsequent analyses have shown how the tobacco indus - take their ads off the airwaves in 1971 following the Public try used its connections within government to assure a , which included a Health Cigarette Smoking Act of 1969 weak bill and a weak warning label (Brandt 2007). The prohibition on broadcast advertising of cigarettes. Conse - wording of the label, “Caution: Cigarette Smoking May quently, the antismoking advertisements were no longer - Be Hazardous to Your Health,” contrasts sharply with the required under the Fairness Doctrine and cigarette con certainty of the 1964 report’s conclusion on smoking and sumption rose after they ended (Warner 1979). From about the time of the 1964 report, per capita lung cancer. Subsequent government actions were largely cigarette consumption began to decline in the United focused around promoting public information about the - States (Figure 2.1), but not uniformly across the popula risks of cigarette smoking and how they might be reduced. tion. Physicians and other health professionals had begun The Surgeon General convened another group of experts to accept the evidence and to stop smoking even before in 1966 to assess the importance of different constituents the release of the 1964 report. While 60% of physicians identified in cigarette smoke for disease risk; the group smoked in 1949, this figure declined to 30% by 1964 (Garfinkel and Stellman 1976). Surveys of Massachusetts recommended that actions be encouraged to progressively physicians during the 1950s found that by 1954 a major reduce the tar and nicotine content of cigarette smoke - ( ity of physicians (55% of smokers and 63% of nonsmok At the same time, FTC 1966). Congressional Record - ers) believed that “heavy smoking of cigarettes may lead revised its advertising guidelines to permit manufacturers to lung cancer” (Snegireff and Lombard 1954, p. 1042). to include in advertisements “a factual statement of the tar and nicotine content (expressed in milligrams) of the Some had switched to smoking only a pipe or cigars, and mainstream smoke from a cigarette” (Shea 1966, Bates many who continued to smoke had reduced the number - Eventually, this disclosure became man No. 00065004). of cigarettes they smoked. Ninety-three percent of the the National Clearinghouse for Smoking datory. In 1968, respondents supported antitobacco education efforts for and Health, a government office, began a campaign “If You youth, and those who did not said it was not because they Must Smoke ...” aimed at people who wanted to reduce doubted the harms of smoking, but because they doubted their risk but did not want to quit smoking. The pamphlet the effectiveness of educational efforts to change teenag- provided five suggestions: (1) choose a cigarette with less ers’ behavior (Snegireff and Lombard 1959). tar and nicotine, (2) don’t smoke the cigarette all the way Surveys of physicians during the 1960s continued down (the last few puffs have more tar and nicotine), (3) to show decreasing prevalence of smoking and accep - take fewer draws, (4) reduce inhaling, and (5) smoke fewer tance of the hazards of cigarette smoking (Buechner et cigarettes (USDHEW 1968). In the absence of any author - al. 1986). A 1965 survey of Oregon physicians found that ity to mandate changes in the product, public education more than one-third (36%) had modified their tobacco became the primary tool to reduce risk. consumption in response to the 1964 report. Additionally, However, one initiative that had a measurable although many physicians had quit earlier, those who quit impact on the prevalence of smoking was initiated by John before 1964 were more likely to cite physical symptoms as 1 A n FCC regulation that required broadcasters to allot time to contrasting points of view on controversial topics. Chapter 2 24

54 The Health Consequences of Smoking —50 Years of Progress of statements that a link between smoking and health had the reason while, after 1964, former smokers were more been proven; but among smokers, the relationship was likely to cite scientific evidence of harm as their reason for the opposite, and smokers with a higher education level quitting (Meighan and Weitman 1965). The prevalence of smoking was also dropping rapidly among medical train were more likely to be skeptical of the evidence (Cannell - and MacDonald 1956). In another study, a survey found ees and younger physicians. The average prevalence of that male smokers were relatively optimistic about their smoking among medical students at Johns Hopkins Medi- chances of contracting cancer, while female smokers were cal School was 65% for the years 1948 through 1951, but - not (Toch et al. 1961). And a 1963 study found that aware by 1965 the prevalence had dropped below 40% (National NCI ness of science reporting had little impact on smoking ] 1994). Younger physicians were Cancer Institute [ also more likely to report concern over the health effects behavior, as many smokers were prone to doubt the scien- tific claims or exhibit fatalistic attitudes about health risks of smoking on patients, to ask or advise patients about (Robinson 1960). their smoking, and to agree that physicians should set an It was not until the 1970s that a majority of Americans said smoking was a cause of lung cancer. But example by not smoking (Coe and Brehm 1971). By the - early 1980s, surveys suggested that only 5–10% of physi the proportion with this view climbed steadily from about 70% during the 1970s to about 80% in the 1980s. By the cians were smoking (Sachs 1983; Buechner et al. 1986). - 1990s, Gallup polls consistently showed 95% of Ameri In 2006–2007, the prevalence of current smoking among physicians had reached about 2% (Sarna et al. 2010). cans claiming to believe cigarette smoking to be harmful to health and 90% believing it to be a cause of lung cancer Appreciation of the health risks, and subsequent - (Saad 1998; Moore 1999). behavior change, was slow to follow among the gen Cigarette consumption was similarly slow to change. eral population. Gallup polls have surveyed Americans Per capita consumption figures increased every year from about their beliefs on the health effects of smoking since 1950 to 1963, with the exception of 1953 and 1954, when the 1950s (Gallup Organization 1964). In 1954, 70% of - there was the first widespread publicity on early labora respondents believed that smoking was harmful to health. - However, the question—“Do you think cigarette smok tory animal and human cohort study findings (Figure 2.1). Consumption decreased in 1964 and during all of the ing is harmful, or not?”—was phrased in such a general Fairness Doctrine years of 1967–1970. Since 1973, every way as to encompass a wide range of possible effects. year for which data are available has seen declines in per - Respondents were also specifically asked about lung can capita adult cigarette consumption (U.S. Department of cer. Although 83% of respondents answered ‘yes’ to the Agriculture 2007; U.S. Census Bureau 2013; U.S. Depart question “Have you heard or read anything recently that - cigarette smoking may be a cause of cancer of the lung,” ment of the Treasury 2013). only 41% answered ‘yes’ to the next question “What is Although antismoking publicity and news reports did have an impact on beliefs and behavior over time, your opinion -- do you think cigarette smoking is one of there were also forces working against this trend. In the causes of lung cancer.” When respondents were asked about specific health effects from smoking, only 7% men- particular, the tobacco industry’s marketing efforts and tioned cancer of any kind. Instead, most cited a variety organized campaign to promote doubt around smoking and health surely slowed the pace of change. A 1966 PHS of non-life-threatening problems such as coughing, sinus irritation, nervousness, and fatigue (Saad 2002). survey found that more than 60% of smokers agreed that Even after the 1964 report, there was not a dramatic the cancer link was “not yet proved” because it was “only based on statistics” (National Clearinghouse for Smoking - change in public beliefs about smoking. In a 1966 Har and Health 1969, p. 743). Additionally, well over one-half ris poll, only 40% recognized smoking as a major cause of all smokers believed that most people would not be con of lung cancer, 27% considered it a minor cause, and - vinced smoking was harmful until “the tobacco industry one-third were uncertain, saying that “science had not USDHEW yet determined the relation between smoking and lung - 1969). Even as pub itself” admitted the fact ( cancer” (Saad 1998, p. 3). In general, although there was lic knowledge about the link between smoking and lung widespread awareness of reports of findings on smoking cancer became widespread during the 1970s and 1980s, and health, including lung cancer, people were unsure a 1981 FTC review concluded that many Americans still whether to believe the results were conclusive. This had very limited knowledge of the nature and extent of uncertainty may have reflected, at least in part, the doubt- the health risks or how those risks applied to their own 1981). creating strategies of the tobacco industry (Proctor 2011). FTC behavior ( The nature of cigarette advertising also changed, Some early studies hinted at the complexity of apparently in response to adverse publicity, to obscure beliefs about health risks and the factors determining the extent of the danger. During the 1970s, there was an those beliefs. For example, having a higher education level among nonsmokers was associated with acceptance increased emphasis on ads that featured claims about tar 25 Fifty Years of Change 1964–2014

55 Surgeon General’s Report million in revenue for the magazine (Warner 1985a). An - and nicotine content, implying reduced exposures to can cer-causing agents ( analysis of magazine coverage over a 22-year period found 2001). Key words such as “light,” NCI that a sample of major magazines reduced their coverage “smooth,” and “mild,” were used to convey health-related messages (Kozlowski 2010). In the 1980s, these health of smoking and health issues by 65% in the years after messages became more subtle, relying on imagery of the broadcast advertising ban went into effect (Warner and active, healthy models (Warner 1985b). Goldenhar 1989), and another study found that magazines Additionally, the tobacco industry’s power as a source which accepted an average amount of cigarette advertis - ing were 38% less likely to carry stories on smoking and of revenue for many print publications influenced the health than magazines that did not accept cigarette adver - content of smoking and health media coverage (USDHHS tising (Warner et al. 1992). 1989; NCI 2008). After the broadcast advertising ban, ciga- Although many individual physicians rapidly rette advertising and marketing continued to grow, but shifted to print publications, outdoor billboards, sponsor - accepted the smoking and health findings, AMA, the lead - ship of sports, placement of brand implants in movies, and ing professional medical organization, took more than , two decades to take a clear stand on the issue. In 1964, a number of other methods. According to Advertising Age the five major tobacco companies spent $62 million on after the release of the report of the Surgeon General, AMA magazine advertising in 1970, the year before the ban, but published a 7-page brochure for the general public titled “Smoking: Facts You Should Know,” which described a by 1976 they were spending $152 million (Smith 1978). - range of “suspected health hazards” but portrayed experts Some publications became highly dependent on this rev noted Columbia Journalism Review enue. An article in the as divided on the issue (AMA 1964). At the time, AMA a trend: “In magazines that accept cigarette advertising,” officials also opposed federal efforts to mandate warning Smith (1978) wrote, “I was unable to find a single article, labels, advertising restrictions, or other public education - efforts around smoking (Haseltine 1964). Historians have in seven years of publication, that would have given read ers any clear notion” of the nature and extent of the health noted that AMA’s position on smoking during the 1960s and 1970s was influenced by its need for support from effects of cigarette smoking, including news magazines and Newsweek . As late as 1983, a Newsweek Time like congressional allies, particularly in southern tobacco- 16-page special supplement on “personal health care” growing states, as well as its opposing Medicare and pro - prepared with AMA failed to explicitly identify cigarette posed national health insurance legislation during those years (Kluger 1996; Rothstein 2003; Proctor 2011 - smoking as a major health hazard. The same issue car ). ried 12 pages of cigarette advertisements worth about $1 Passive Smoking and Environmental Change Bates No. 91018247/8260, p. 14). He then went on to pro- Surgeon General Jesse L. Steinfeld, appointed by President Richard M. Nixon in December 1969, helped to pose “It is high time to ban smoking from all confined public spaces such as restaurants, theaters, airplanes, bring public attention to the effects of smoking on non- smokers. Although he had more limited authority com- trains, and buses. It is time that we interpret the Bill of pared with his predecessors due to a reorganization within - Rights for the Non-smokers as well as the Smoker” (Stein USDHEW, he made use of the public platform of the Office - feld 1971, Bates No. 91018247/8260, p. 14). The subse quent 1972 report was the first in the series to identify of the Surgeon General to advance public health. He rein - the exposure of nonsmokers to cigarette smoke as a health vigorated the regular reports of the Surgeon General on smoking and health, involving dozens of outside experts as hazard (USDHEW 1972). Dr. Steinfeld bluntly affirmed in authors and peer reviewers to produce a 458-page report - his remarks when releasing the report “There is no dis agreement – cigarette smoking is deadly” (Steinfeld 1972, in 1971 and the first report to address passive smoking in Bates No. TITX0004900/4909, p. 2). In a chapter titled 1972 (see Chapter 3). In a 1971 address to the Interagency Council on “Public Exposure to Air Pollution from Tobacco Smoke,” the report summarized information on the contamination Smoking and Health, Steinfeld asserted that “Nonsmok- of indoor environments by tobacco smoke. The review ers have as much right to clean air and wholesome air as smokers have to their so-called right to smoke, which showed that levels of carbon monoxide in a smoke-filled I would redefine as a ‘right to pollute’ ” (Steinfeld 1971, room could reach concentrations equal to and even above Chapter 2 26

56 —50 Years of Progress The Health Consequences of Smoking “Public Health Enemy Number One” and “slow motion standards for ambient air. The report also concluded the tobacco smoke was a source of discomfort for many - suicide” and declared: “The first and most important ele ment of this new program on smoking and health will be people, but characterized the health risks of tobacco Steinfeld continued to be a major public information and education effort against smoke in the air as unknown. smoking” (Califano 1978, p. 10). Califano’s actions did outspoken and an advocate for smoking bans and, unlike not develop in a vacuum, however. They reflected a grow- his predecessors, he refused to meet with tobacco industry ing national agenda of public health advocacy against representatives (Kluger 1996). smoking (National Commission on Smoking and Public A grassroots movement emerged in the early 1970s Policy 1978). to promote the interests of nonsmokers. Influential early The 1979 Surgeon General’s report, - Smoking and organizations included Group Against Smoking Pollu , tion, with chapters in several states and Californians for Health released under Califano, marked the 15-year anni - versary of the original 1964 report. The report included Non-Smokers Rights (now known as Americans for Non- smokers Rights) based in Berkeley, California. They drew more than 1,100 pages and presented an enormous amount of data from now decades-long epidemiologic explicitly on the rhetoric and discourse of the civil rights “the innocent and environmental movements, referring to cohort studies, studies of mechanisms of disease, studies of behavioral and psychosocial influences on tobacco use, victims of tobacco smoke” and a need to give the “right and the effectiveness of education programs and interven- to breathe clean air” precedence over “the right of the tions. It included a chapter titled “Involuntary Smoking” smoker to enjoy a harmful habit” (Nathanson 1999). At that summarized the data on contamination of indoor the time, there was little data on the harms of exposure environments by tobacco smoke. The report also reviewed to secondhand smoke. However, an increasing number of the initial evidence on the health consequences of invol - nonsmokers viewed it as an annoyance in shared spaces, untary smoking, but called for more research without such as restaurants and airplane cabins. And the existence reaching any conclusions as to risks ( of a potential risk, however uncertain or small, was viewed USDHEW 1979). in a fundamentally different way when it affected involun In the Secretary’s Foreword to the volume, Califano - some of whom might be suscep- wrote: “But why, the reader may nevertheless ask, should tarily exposed bystanders, tible to the effects (Bayer and Colgrove 2002). government involve itself in an effort to broadcast these facts and to discourage cigarette smoking? ... Why, indeed? A wave of new rules and legislation limiting smoking For one reason, because the consequences are not simply followed (USDHHS 2006). Several were at the federal level. personal and private. Those consequences, economic and In 1973, the Civil Aeronautics Board, which had jurisdic - tion, ordered domestic airlines to provide separate seat - medical, affect not only the smoker, but every taxpayer” USDHEW ( 1979, p. ii). That is, smoking went beyond ing for smokers and nonsmokers. In 1974, the Interstate Commerce Commission ruled that smoking be restricted being a private medical concern to being a major public to the rear 20% of seats in interstate buses. Pioneering health problem that affected smokers and nonsmokers. actions on indoor spaces were also taken at the local and In particular, Califano cited two health policy challenges USDHHS 2006). In 1973, Arizona state levels in the 1970s ( then facing the nation—the spiraling costs of health care, became the first state to restrict smoking in some pub with a substantial portion borne by the federal govern - - ment, and the fact that the health care system “overem - lic spaces. In 1974, Connecticut enacted the first statute to restrict smoking in restaurants. Minnesota followed in phasizes expensive medical technology and institutional 1975, requiring no-smoking zones in buildings open to care, while it largely neglects preventive medicine and USDHEW the public. In 1977, Berkeley, California, became the first health promotion” ( 1979, p. ii). Smoking is, he city to pass an ordinance limiting smoking in restaurants. noted, “the largest cause of preventable death in America” - ( At the same time, antismoking efforts in the United States 1979, p. ii). At the same time, Califano acknowl USDHEW edged limits to government’s role in regulating cigarette began to develop into a more diverse movement, involving smoking in a free society and suggested that intervention a broad constituency of volunteer health organizations, would have to focus primarily on research, education, and professional organizations, and newly created advocacy persuasion. The report also brought a renewed focus to groups, such as Doctors Ought to Care created in 1977 the need for understanding smoking behavior and how USDHHS ( 2006). to help people who want to quit. Thirty million Ameri - When lawyer Joseph A. Califano, Jr., became Secre - cans, the report stated, had become former smokers since tary of the Department of Health, Education, and Welfare 1964, and this figure gave encouragement that persuasion under the incoming Carter Administration, he made a and education could have population-level impacts ( USD- strong antismoking campaign one of his first priorities. HEW 1979). The report also highlighted the effects of On January 11, 1978, Califano outlined his battle plan smoking for specific vulnerable or high-risk populations, in a public speech in which he called cigarette smoking 27 Fifty Years of Change 1964–2014

57 Surgeon General’s Report including women, youth, minorities, the developing fetus, tobacco smoke, dosimetry and toxicology, exposures, and and certain occupational groups. In this way, too, gov- the findings of epidemiologic studies ( USDHHS 1986). It ernment intervention was seen as justified by the need interpreted that evidence within the context of what was to protect those who are most vulnerable or at increased already known about active smoking, treating exposure to secondhand smoke as resulting in a lower dose of tobacco risk. In his preface, Surgeon General Julius B. Richmond smoke, compared with active smoking, but to the same similarly highlighted the difficulty of seeing smoking as toxic mixture from a health perspective. The report had simply a personal choice, given the hundreds of millions of dollars spent each year in marketing and promotion of three overall conclusions, including its powerful first cigarettes and the possibility that “nicotine is a powerful conclusion: “Involuntary smoking is a cause of disease, 1979, p. xv). USDHHS including lung cancer, in healthy nonsmokers” ( addictive drug” ( USDHEW At this time, the scientific evidence on the health 1986, p. 7). Its second conclusion described the adverse effects of exposure to secondhand smoke was limited. Stud effects of smoking by parents on the respiratory health of - ies starting in the late 1960s had shown adverse effects of their children. Its third—“Simple separation of smokers maternal smoking on the developing fetus and on children and nonsmokers within the same air space may reduce, exposed to secondhand smoke in smoking households - but does not eliminate, exposure of nonsmokers to envi ronmental tobacco smoke” ( USDHHS (Comstock and Lundin 1967; Colley et al. 1974). How- 1986, p. 7)—carried ever, it was not until the following decade that a critical - implications for controlling exposure to an agent identi mass of scientific evidence emerged linking exposure to fied as carcinogenic in the first conclusion. secondhand smoke with cancer and other chronic health Surgeon General C. Everett Koop, appointed by effects among nonsmoking adults. In 1980 and 1981, sci- President Ronald W. Reagan in 1981, used the visibility - entific journals published epidemiologic research from of the position to a greater degree than any of his pre decessors and used the findings of the report to call for Greece, Japan, and the United States finding that those smoke-free public places. He was an outspoken public foe who breathed “environmental tobacco smoke” suffered of tobacco, advocating a smoke-free environment by the from decreased lung function (White and Froeb 1980) and - year 2000. Although he was aware of the controversy sur - increased risk of lung cancer (Hirayama 1981; Trichopou los et al. 1981). Because the lung cancer investigations rounding the scientific evidence on secondhand smoke, involved people who had experienced heavy exposure to further fueled by the tobacco industry’s efforts to focus attention on the limitations of the data, he insisted that smoke in the home over long periods of time, there were questions about whether, and to what extent, the data the data were sufficient for public health intervention. Koop declared in his Preface to the 1986 report “Critics could be extrapolated to other enclosed public spaces. But often express that more research is required, that certain over the next several years, additional studies gave weight to the argument that adult nonsmokers suffered harm by studies are flawed, or that we should delay action until breathing the cigarette smoke of others and that smok - more conclusive proof is produced” ( USDHHS 1986, p. xi). ing by parents adversely affected the respiratory health of He went on to argue, based on the report’s third overall conclusion, that many of the measures that had been put their children. In 1986, two major scientific reviews were released in the United States—the U.S. Surgeon General’s - into place in many states and communities were inad report, The Health Consequences of Involuntary Smoking equate, such as creating separate nonsmoking sections (USDHHS 1986), and the National Academy of Science’s with a common ventilation system did not eliminate expo- sure for nonsmokers. Koop also asserted that “[t]he right report, Environmental Tobacco Smoke: Measuring Expo- of smokers to smoke ends where their behavior affects the (National Research sures and Assessing Health Effects Council 1986)—both concluding that secondhand smoke health and well-being of others (USDHHS 1986, p. xii). This report, along with the complementary find could cause lung cancer in healthy adult nonsmokers - ings of the reports from the National Academy of Science and respiratory symptoms in children. In that same year, the World Health Organization’s ( WHO ’s) International and IARC, provided the scientific foundation for policies and actions to protect nonsmokers from inhaling tobacco ) concluded that “... IARC Agency for Research on Cancer ( passive smoking gives rise to some risk of cancer” ( smoke (NRC 1986; USDHHS 1986). By the mid-1980s, IARC almost all states had enacted some restrictions on where 1986, p. 314). people could smoke in public; some 80% of the U.S. popu - - The 1986 report of the Surgeon General on involun lation lived in areas covered by such laws (USDHHS 2006). tary smoking represents another landmark in the series Between 1985–1988, the number of communities around of reports. Following the approach of the 1964 report, it the country that had enacted laws restricting public assembled the full body of evidence on exposure to sec - smoking almost quadrupled, to over 300 (USDHHS 1989). ondhand smoke and health, reviewing the composition of Chapter 2 28

58 —50 Years of Progress The Health Consequences of Smoking In 1987, USDHHS established a smoke-free environment Amendments to the Federal Food, Drug, and Cosmetic (1976), Toxic Substances Control Act , and the in all of its buildings nationwide, extending protection to Act of 1938 - more than 100,000 federal employees ( creating new agencies and greatly expanding the regula USDHHS 2006). In 1988, Congress imposed a smoking ban on all U.S. domes - tory authority of some existing agencies. In 1992, EPA tic flights of 2 hours or less. Two years later, the ban was carried out a risk assessment and classified environmental extended to flights of 6 hours or less, in effect banning tobacco smoke as a human carcinogen, Group A under its USEPA carcinogen assessment guidelines ( 1992). smoking on all domestic flights. The emerging evidence on exposure to secondhand Once these efforts gained momentum, new legisla- - tion spread rapidly. T he recognition of exposure to sec - smoke and disease, particularly lung cancer, sparked a vig ondhand smoke as a health risk to nonsmokers meant that orous response from the tobacco industry that is now well documented (Brandt 2007; Proctor 2011). The tobacco the issue was no longer merely one of individual choice. People responded differently to risks that were imposed industry recognized the policy implications of evidence showing that exposure to secondhand smoke caused on them involuntarily. The existence of victims of ciga - rette smoking fundamentally altered the discussion about adverse effects among nonsmokers and initiated strate- the right to smoke, and state and legal intervention was gies to undermine the research findings, seeking to create seen as entirely appropriate. There was also substantial doubt about the credibility of evidence that would drive policy-making ( . 2006; United States v. Philip Morris Inc public support for enacting restrictions on smoking in - public spaces. As early as 1970 (before any Surgeon Gen Brandt 2007; Proctor 2011). The first major study to link exposure to secondhand smoke to lung cancer, the cohort eral had spoken out about harm to nonsmokers), 58% of men who had never smoked and 72% of women who had study carried out in Japan by Hirayama (1981), was the target of an orchestrated campaign to undermine its find - never smoked responded ‘strongly agree’ or ‘agree’ that smoking should be allowed in fewer public spaces than ings. The tactics included arranging critical letters to the , which published the British Medical Journal it was at the time ( USDHEW 1973a, p. 11). More than editor of the paper, commissioned research with the intent of obtaining three-quarters of those who had never smoked felt that findings that would point to bias in the study, and even it was “annoying to be near” someone who was smoking USDHEW ( newspaper advertisements discrediting the findings. Such 1973a, p. 13). A 1983 Gallup poll found that strategies were directed at the wider body of evidence on 82% of nonsmokers believed that smokers should not secondhand smoke and health; the industry and its con smoke in their presence and that smoking posed a health - sultants raised methodologic problems, such as uncon hazard for them; 64% of smokers concurred (American - trolled confounding and exposure measurement error, in Lung Association 1983). Additionally, t he phenomenon - may have been self-reinforcing, acting as order to sustain doubt about the findings (Kluger 1996; a sort of conta gion effect where actions on one locale influenced other Proctor 2011). These same tactics and others were used to try locales (Asbridge 2004).The attention to secondhand smoke was also aided by the growth in public concern - and diminish the impact of the 1986 Surgeon Gener al’s report. An attempt was made to engage some of the over environmental pollutants during the 1970s. In 1970, report’s authors in a symposium that had undisclosed - under the Nixon Administration, both the U.S. Environ tobacco industry sponsorship. The report was charac - mental Protection Agency (EPA) and the Occupational terized as political rather than scientific, and Surgeon Safety and Health Administration were created, and the General Koop’s motives were questioned. The attack on Clean Air Act Extension of 1970 established comprehen- the scientific foundation of the report intensified as well on. The fol- sive regulatory control on outdoor air polluti (Proctor 2011). Some of these same strategies were used lowing years saw a wide range of new environmental and s risk assessment subsequently in an attempt to derail E PA’ - safety laws aimed at protecting the public from involun of environmental tobacco smoke. Consumer Products tary risks, including, for example, the (1974), Safe Drinking Water Act (1972), the Safety Act 29 Fifty Years of Change 1964–2014

59 Surgeon General’s Report Nicotine and Addiction tions,” he stated, “is tobacco’s social acceptability” (Jaffe An estimated 30 million people quit smoking in the decade following the 1964 report. Organized programs to 1977, p. 627). help people quit smoking, such as the Five-Day Plan, had By the late 1970s, as smoking behavior was increas - - gained popularity, and by 1970 there was a US$50 million ingly recognized as resembling that of other drug addic a year industry of for-profit smoking cessation programs, tions, an organized research effort began (Jarvik et al. including Smoke watchers, Quit Now, SmokeEnders, and - 1977). A substantial portion of the 1979 Surgeon Gen Schick Centers for the Control of Smoking, but there eral’s report was devoted to behavioral aspects of smoking ( USDHEW 1979); indeed, of the 11 Surgeon General’s was little rigorous testing of the effectiveness of these programs (Goodman 2005). Additionally, throughout the smoking and health reports published between 1964– - 1960s and 1970s, the general understanding of smok 1980, it was the first to include any mention of smoking ing behavior and nicotine addiction was very limited. At behavior or dependence. The authors of the report sought the time, to avoid using the term addiction, not because they health scientists viewed smoking as primarily psychological and social, rather than pharmacological believed it to be scientifically inaccurate, but because of or biological. The 1964 report - its loaded meaning related to illicit drug use (Henning concluded that tobacco dependence should be characterized as a form of habitu - field and Zeller 2006). It was not until the 1988 report that USDHEW the Surgeon General declared that cigarettes are addict ation rather than addiction ( 1964), drawing on - ing, similar to heroin and cocaine, and that nicotine is the WHO in 1957. That definition a distinction established by emphasized the physical effects of the drug, the compul - primary agent of addiction ( USDHHS 1988). sion to obtain it at any cost, and the habit’s detrimental The focus on the behavioral and psychological effects on the individual and society (WHO 1957). The aspects of cigarette smoking and addiction marked a WHO Expert Committee on Addiction-Producing Drugs substantial shift from the earlier science of smoking and esearchers studying the health effects of smok- observed that for cigarette smoking, evidence was lacking health. R - at the time for a typical abstinence syndrome. “In contrast ing during the 1960s and 1970s were primarily epidemi - to drugs of addiction, withdrawal from tobacco never con ologists, statisticians, and pathologists without expertise stitutes a threat to life,” they wrote. “These facts indicate in studying addictive behavior. These researchers were focused on the consequences of smoking and not on why clearly the absence of physical dependence” ( USDHEW 1964, p. 352). At the same time, because regular smoking people smoked. During the 1970s, scientists who had studied other drug addictions turned their attention to was so widespread and socially accepted during the 1960s, cigarette smoking, developing methods to measure nico - scientists were reluctant to portray smokers as addicts or as presenting a threat to society. Maurice H. Seevers, tine intake and smoking behavior. A substantial body of evidence resulted. - the only pharmacologist on the Surgeon General’s Advi sory Committee, had served on WHO’s expert committee The 1988 report of the Surgeon General, also that produced the 1957 definition of addiction and was a released by Surgeon General Koop, reviewed this new evi- longtime proponent of the view that an observable physi dence on smoking and addiction, concluding that: “Ciga - - cal abstinence syndrome was a crucial defining feature rettes and other forms of tobacco are addicting” (USDHHS of addiction (Rasmussen and Seevers 2009). It would be 1988, p. 9) and “Nicotine is the drug in tobacco that causes another decade before federal research funders and public addiction” (p. 9). The third overall conclusion compared health scientists created an organized research program nicotine addiction to other addicting drugs, including around smoking dependence and nicotine addiction. In heroin and cocaine. The report changed the view that smoking was just a the mid-1970s, scientists were beginning to compare tobacco smoking with other drug addictions. For example, habit. Cigarettes were now cast as addicting and as equally Jerome H. Jaffe, who had promoted methadone treatment addictive as many illegal drugs. The findings also had for heroin addicts as President Richard M. Nixon’s drug implications for treatment, pointing to the possibility of czar from 1971–1973, began to argue in favor of treating using nicotine replacement therapy to increase successful cigarette smoking as an addiction in the mid-1970s, main quitting of nicotine ( USDHHS 1988). For smoking initia - - taining that it did meet the appropriate criteria, including tion by youth, the finding that nicotine is addicting raised - the presence of a withdrawal syndrome. “The major differ concern that adolescents and young adults might become ence between tobacco dependence and other drug addic- addicted through experimentation; by 1988, the pattern Chapter 2 30

60 —50 Years of Progress The Health Consequences of Smoking received great attention from the industry and its con of initiation had moved to the teen years for both males - USDHHS 1988). The 1994 Surgeon General’s sultants. The tobacco industry had information about and females ( the report when it was in development and was quick to Preventing Tobacco Use Among Young People report on emphasized that tobacco use and addiction almost always criticize its findings after release. The finding that nico- begins before 18 years of age and that most adolescent tine was addicting countered the argument that people became smokers by their own free choice. Efforts to dis- smokers face the same challenges as adults in quitting 1994). USDHHS smoking ( credit the report continued long after its publication, even Like the 1986 report, the 1988 report had profound though the industry’s own documents show that it had long known that nicotine was addicting (Proctor 2011). implications for the tobacco industry, and the report also Denormalization and the Tobacco Industry Beginning in the mid-1970s, per capita cigarette For example, during the 1990s, NCI conducted a large nationwide intervention study – American Stop consumption began to decline more steeply than during the decade following the 1964 report (Figure 2.1). The Smoking Intervention Study, known as ASSIST. With scientific findings on tobacco smoke, summarized and a budget of approximately $117 million over 7 years, ASSIST provided funding to 17 states for the development transmitted to the health community and the popula- tion at large through the Surgeon Generals’ reports and - of coalitions to pursue a range of interventions and poli cies at the state and local levels, including (1) promoting other channels, provided a basis for motivating effective action to control tobacco use. Underlying the decline was smoke-free environments; (2) countering tobacco adver - tising and promotion; (3) limiting youth access to tobacco increasing public understanding of the dangers of ciga - products; and (4) raising excise taxes to increase the price rette smoking and increasing unacceptability of being a of tobacco products ( smoker; that is, the social norm around smoking changed NCI 2005). The project was unique at the time for its scale and focus on studying the effec from being completely acceptable and woven into day-to- - day activities and interactions among people to becom- - tiveness of broad strategies for policy change. The inter vention led to a greater reduction in the prevalence of ing an increasingly unacceptable behavior. Many factors smoking in states participating in the ASSIST program contributed to this change, including the evidence on the than in non-ASSIST states, although the effect was mod dangers of exposure to secondhand smoke and the ever- - est, likely because of the general trend of declining per increasing reluctance of nonsmokers to inhale tobacco capita cigarette consumption over the years of the study smoke in their workplaces, public places, and eventually their homes ( USDHHS 2005). 1986). (Figure 2.1) ( NCI State tobacco control programs also took a more Additionally, the tobacco control “toolbox” expanded with an increasing number of strategies: smoking bans, aggressive approach during the 1990s, moving beyond a focus on the harms of exposure to secondhand smoke - which both protected nonsmokers and encouraged cessa to directly countering cigarette advertising efforts. As tion; educating youth and limiting their access to tobacco products with enforced laws; raising taxes to force the cigarette advertising linked smoking to glamour, vitality, - and social success, some state programs, such as those in price of cigarettes upward; encouraging smoking cessa tion and using treatments that were shown to be effec - California, Florida, and Massachusetts, turned to explicit denormalization strategies ( USDHHS 2000). They aimed tive; and using the media to counter the marketing of the tobacco industry (Kluger 1996; Proctor 2011). Advocacy “to push tobacco use out of the charmed circle of normal, at the local grassroots level played a critical role as non- desirable practice to being an abnormal practice” (Cali- smokers demanded smoke-free environments. The need fornia Department of Health Services 1998, p. 3). In the for using a battery of tobacco control measures was recog - late 1990s, the states received substantial funding from ) between the 1998 Master Settlement Agreement ( MSA nized and trials were carried out at the community level to assess the efficacy of combined approaches and their the tobacco companies and the attorneys general of 46 USDHHS 2000, 2012). Initially, some of the funds states ( effectiveness in practice. 31 Fifty Years of Change 1964–2014

61 Surgeon General’s Report The tobacco industry was further discredited by from the MSA were directed to tobacco control, but the funding declined as states used the revenues for other pur - congressional hearings and the litigation brought by the poses and only a few states ever reached the ) against the industry, CDC DOJ U.S. Department of Justice ( ’s recom- mended funding levels (Sloan et al. 2005; 2012). , under the Racketeer Influ- CDC United States v. Philip Morris enced and Corrupt Organizations Act (RICO 1970). FDA Additionally, after decades of failed personal injury lawsuits against the tobacco industry for smoking-related launched a large-scale investigation into the manipulation of nicotine levels in cigarettes and marketing to youth and, harms, the climate for tobacco industry litigation trans- - formed during the 1990s. There was one major develop for the first time, asserted jurisdiction over cigarettes as drug delivery devices (see Chapter 14). At a 1994 hearing, ., a personal Cipollone v. Liggett Group, Inc ment with injury case filed in 1983 on behalf of a New Jersey smoker seven tobacco company CEO s insisted that they believed - Cippollone v. Liggett Group nicotine was not addictive and not a cause of disease. Pho and lung cancer victim ( 1988). The plaintiffs gained access to some internal tographs of the group holding up their right hands and tobacco company documents supporting claims that the being sworn in at the hearing, while denying what most members of the public knew to be true about cigarettes, industry had conspired to withhold information about turned them into objects of ridicule and further dimin - harm from the public. But, it was during the 1990s that ished the public’s view of the tobacco industry (Brandt far more complete access was gained to the industry’s litigation, the industry was found guilty 2007). In the internal documents. Two major events made this possible. DOJ of violating civil racketeering laws and lying to the public First, an employee of a law firm that represented tobacco companies released documents to the public that exposed about the dangers of tobacco and its marketing to chil- the tobacco companies’ misconduct. Second, class-action dren. The opinion by Judge Gladys Kessler focused on the litigation and litigation on behalf of state governments representation of cigarettes with reduced machine yields of tar and nicotine as conveying lower risks and the indus- allowed plaintiffs to combine their resources and expertise - on a scale not before realized (Miura et al. 2006). The liti try’s denial of the health effects of exposure to secondhand smoke ( United States v. Philip Morris et al. gation by the State of Minnesota and Blue Cross and Blue 2006). Momentum from the states’ lawsuits also turned - Shield of Minnesota resulted in the release of the indus try’s documents and their maintenance in two reposito - the political tide against the tobacco industry in the ries, one in Minnesota for the U.S. industry and the other mid-1990s, and their influence in Congress weakened (Sack 1997). Additionally, the characteristics of legisla in Guildford, England, for British American Tobacco’s - documents. Under the MSA, the industry is required to tive debates on tobacco control measures at the state level continue to place its documents into a depository until changed from its prior focus (on the sufficiency of scien- 2021. The Legacy Tobacco Documents Library at the Uni - tific evidence of health effects during the 1970s and early 1980s) to the impact of tobacco industry activities and versity of California at San Francisco (2013) was created to house these documents. marketing on children (Jacobsen and Wasserman 1997). Evidence compiled by FTC and researchers demonstrated The MSA was the result of suits by state governments that the RJ Reynolds’ Joe Camel marketing campaign had against tobacco companies to recover Medicaid expenses a measurable impact on smokers below the legal age and they had paid to care for sick smokers ( USDHHS 2000). was accompanied by an increase in smoking initiation From 1993–1998, almost every state filed an action against among youth (DiFranza et al. 1991; Pierce et al. 1998). the tobacco companies. The process ended with individual During this period, tobacco companies lost credibility in - settlements with the states of Florida, Minnesota, Missis the eyes of the public. A Harris poll taken in March 1997 MSA with the remaining 46 states sippi, and Texas, and the found that 92% of the respondents believed “tobacco com- and the District of Columbia. The MSA required tobacco panies know it causes cancer even if they do not admit it” companies to pay $206 billion over the initial 25 years of and 80% believed that “some tobacco companies market the agreement. The MSA did not just provide monetary their products deliberately to young people” (Sack 1997). relief to the states, but also placed restrictions on the Attitudes around the engagement of scientists and - tobacco companies that included ending cigarette bill physicians with the tobacco industry were also changing board advertising, banning the use of merchandise with during the 1990s. The tobacco industry had long funded cigarette brand names, and limiting sponsorships. Addi - researchers through the Council for Tobacco Research and tionally, as a result of the Minnesota Settlement and the later through the Center for Indoor Air Research (Proctor MSA, tens of millions of pages of internal memoranda, 2011). Such funding became increasingly unacceptable, reports, and other tobacco company documents initially - and universities began to implement policies that pro acquired through litigation were made available to the hibited receipt of funding from the tobacco industry. It USDHHS public ( 2000). Chapter 2 32

62 —50 Years of Progress The Health Consequences of Smoking medical research community was divided over the issue had also recruited researchers as consultants, who were at the time, as some academic medical leaders objected key in its doubt-creating initiatives. Engagement with the industry became increasingly unacceptable for research- - that restrictions on funding from any particular indus try would amount to a restriction on academic freedom. ers whose reputations were tarnished by their industry However, tobacco control advocates countered that the activities. At the same time, concerns about potential tobacco industry’s well-documented record of manipu- conflicts of interest among scientists increased, and dis - lating scientific information and the extent of the harms closure of consulting activities to universities became the from cigarette smoking distinguished them from other norm, making it more difficult for researchers to main- tain secret ties to the tobacco industry. industries (Proctor 2011). By contrast, when Under Commissioner David A. Kessler, who held the 1964 report was released, there was little concern that the office from 1990–1997, FDA had attempted to regu- scientists’ results would be influenced by their funding late tobacco products (USDHHS 2000). This effort was source. During the 1990s, a number of tobacco control researchers and organizations began to speak out against ended by the Supreme Court, which found that Congress tobacco industry funding of research at academic institu- had not intended that FDA should regulate tobacco when tions. Some academic medical journals instituted policies it passed the Food, Drug, and Cosmetic Act (Bayer et al. refusing to accept papers for review if the research had 2013; Orentlicher 2013) . With the passage of the Family - Tobacco been funded by the tobacco industry. In 1994, a num Smoking Prevention and Tobacco Control Act ( FDA received authority to regulate ) in 2009, ber of academic medical centers, including Brigham and Control Act FDA’ Women’s Hospital, Massachusetts General Hospital, MD s Center for Tobacco Products is tobacco products. Anderson Cancer Center, Roswell Park Cancer Institute, now proceeding with implementation of the provisions of the Act (see Chapter 14). and others, adopted policies barring their faculty and - staff from accepting tobacco industry support. The bio Summary Over the “cigarette century,” cigarette smoking - with deception, with a single-minded focus on their finan prevalence has risen and fallen and moved from being cial success, and without regard for the human tragedy or widely accepted to socially unacceptable. In 1964, almost United States v. Phillip social costs that success exacted” ( - Morris one-half of U.S. adults were cigarette smokers and smok 2006, p. 28). The tobacco industry is the only legal ing was ubiquitous in many public places, including res - industry to have been pursued and convicted under fed - eral racketeering statutes. taurants, theaters, and airplane cabins. Today, the overall prevalence of U.S. adult smoking is around 20%, less than The epidemic of smoking-caused disease in the one-half of what it was in 1964 (see Chapter 13); as of twentieth century ranks among the greatest public health April 2013, 81% of the U.S. population lives in munici - catastrophes of the century, while the decline of smok - ing consequent to tobacco control is surely one of public palities covered by a smoke-free workplace law at the state health’s greatest successes. Many premature deaths have - or local level that includes at least nonhospitality work places (American Nonsmokers’ Rights Foundation 2013). been avoided because of tobacco control programs, but - many more could have been avoided if smoking preva Twenty-four states and the District of Columbia have 100% smoke-free workplace laws that also cover bars and lence had dropped more rapidly when the early warnings of lung cancer risk were widely reported in 1950. The 1964 restaurants. In July 2011, a Gallup poll reported that for Surgeon General’s report gave momentum to tobacco the first time, a majority of Americans (59%) supported control; the authority of the Surgeon General, and the a ban on smoking in all public places (Newport 2011). approach of the Advisory Committee to developing the Opinions of the tobacco industry have fallen so low that it is now consistently ranked among the most distrusted report, gave unimpeachable credibility to the conclusion of industries (Harris Poll 2012). The industry has been that smoking caused lung cancer (in men). That same authority has empowered the conclusions of subsequent found guilty in the courts as well. Most notably, in 2006, U.S. District Judge Kessler ruled in the decade-long DOJ’s reports that have covered involuntary smoking, addiction - lawsuit against the tobacco industry, finding “the indus to nicotine, tobacco control interventions, smoking by adolescents and young adults, and other topics. try had marketed and sold their lethal products with zeal, 33 Fifty Years of Change 1964–2014

63 Surgeon General’s Report Tobacco control programs proved more challenging concerns to support their cause, the 1964 report rein - forced the central role of science as the primary authority than simply disseminating knowledge to the population “Smoking to inform public health policy. Subsequent reports have of the dangers of smoking. Brandt notes that is a complex behavior which has reflected deep social, maintained that position. cultural, and economic forces, as well as a powerful bio- Because of the complexity of the factors involved, logical process of addiction. Simply identifying individual it is difficult to measure the degree to which particular behavior as the primary vehicle of risk negates the fact interventions, following the 1964 report, influenced pat - terns of tobacco use. However, it is clear that tobacco con- - that behavior itself is, at times, beyond the scope of indi trol policies and actions need to draw on the full suite of vidual agency” (Brandt 1990, p. 172). This complexity, the addicting nature of nicotine, and the dynamic efforts of interventions of proven efficacy. Grassroots activities and coalitions have played a critical role, as they supported the industry to maintain its market, challenged initial smoking bans and had substantial impact in changing the efforts to curb tobacco use. Over time, the need for broad social norm around smoking. interventions with multiple components was recognized, and cigarette consumption began to decline at a faster The past half-century of public health experience with cigarette smoking, since the 1964 report, holds many pace (Figure 2.1). Several factors were particularly cru - cial in altering social norms around cigarette smoking in important lessons for the future and for the actions that will follow from this report. Overall, this ongoing story the United States, making it increasingly less acceptable: illustrates the complexity of the factors involved and the (1) the emergence of a nonsmokers’ rights movement and evidence linking exposure to secondhand smoke to dis - need to consider cigarette smoking, not simply as an indi- vidual decision about behavior, but as a large-scale social ease; (2) an understanding of regular cigarette smoking as an addictive behavior and one that begins in adoles - and cultural phenomenon. Despite the conclusive evidence of the harms of cigarette smoking presented in the 1964 cence; and (3) a focus on the tobacco industry itself as a report, as evaluated by an objective group of experts, the key influence on smoking behavior and the importance of countering its actions. - process of changing public beliefs, attitudes, and behav Other factors played a role in iors took decades, and the implementation of effective shaping attitudes and policies around cigarette smoking, policies involved a lengthy process of intervention, evalu- including changes in political administrations, the devel- ation, and surveillance. The tobacco industry’s extensive opment of a grassroots advocacy movement, the changing campaign to counteract these forces through marketing, climate for litigation, and developments in the organiza - public relations, political influence, and creation of doubt tion of public health research. about the scientific evidence on tobacco is now well docu - The production of the 1964 Surgeon General’s mented through the industry’s internal documents. The report itself was a significant public health action, even if direct and immediate policy action seemed slow to follow. industry used its influence to thwart public health action Additionally, the 1964 report was a pioneering step toward at all levels and fraudulently misled the public on many issues, including whether lower-yield cigarettes conveyed - anticipating a much larger role for government, in col laboration with scientists, to use science to inform regula less risk to health and whether exposure to secondhand - tory and other policies. This approach is embodied in the smoke harmed nonsmokers. Undoubtedly, these actions . Although early twentieth cen Tobacco Control Act 2009 - slowed progress in tobacco control. tury antitobacco reformers appealed to moral and social Chapter 2 34

64 The Health Consequences of Smoking —50 Years of Progress References . ‘AMA Journal’ stops taking cigarette ads. Advertising Age British American Tobacco Collection. Bates No. Advertising Age 1953;24:1, 39. 400566440-400566490. . - Allen G. [Letter from George Allen to Clarence Cook Lit Buechner JS, Perry DK, Scott HD, Freedman BE, Tier - tle]. 1962. Council for Tobacco Research Collection. Bates No. 11308700/8700. . Rhode Island physicians, 1963–83. American Lung Association. Summary of the results of the 1986;76(3):285–6. of Public Health April 1983 survey by the Gallup Organization Survey Burney L. False and misleading advertising statement of Dr. John R. Heller, Director, National Cancer Institute: of Attitudes toward Smoking [news release]. New York: discussion of statistical data linking lung cancer with American Lung Association, September 20, 1983. American Medical Association. Smoking: Facts You smoking, July 23, 1957, < http://tobaccodocuments. Should Know org/ti/TIMN0105109-5135.html?zoom=750&ocr_ . 1964. Research Collection. Bates No. > ; position=above_foramatted&start_page=24 502817600-502817603. . Americans for Nonsmokers’ Rights. Tobacco Industry Burney LE. Excessive cigarette smoking. Public Health Reports Sound Bites and Responses . Berkeley (CA): Americans 1957;72(9):786. for Nonsmokers’ Rights, 2004, < Burney LE. Smoking and lung cancer: a statement of the http://no-smoke.org/ pdf/industrysoundbites.pdf>; accessed: May 5, 2013. Journal of the American Medi- Public Health Service. cal Association American Nonsmokers’ Rights Association. Summary of 1959;171:1829–37. 100% Smokefree State Laws and Population Protected Burnham JC. American physicians and tobacco use: two Surgeons General, 1929 and 1964. by 100% U.S. Smokefree Laws (April 5, 2013), 2013; Bulletin of the His - tory of Medicine 1989;63(1):1–31. ; accessed: July 3, 2013. Califano JA Jr. Address by Joseph A. Califano, Jr. Secretary Asbridge M. Public place restrictions on smoking in Can of Health Education and Welfare Before the National - Interagency Council on Smoking and Health, Shore - ada: assessing the role of the state, media, science and public health advocacy. - ham Hotel, Washington, D.C. 1978. Mangini Collec Social Science and Medicine tion. Bates No. 500798779/8805. < 2004;58(1):13–24. http://legacy.library. Banzhaf v. FCC, 405 F.2d 1082 (D.C. Cir. 1968), cert. ucsf.edu/tid/ulx62d00>. 396 U.S. 842, 90 S. Ct. 50 (1969). denied, California Department of Health Services. A Model for Bass A. Re: Notes from Tobacco Institute P.R. meeting of Change: The California Experience in Tobacco Con - September 19, 1963. [memorandum]. 1963. Research . Sacramento (CA): California Department of Health trol Collection. Bates No. 84409557-84409558. . Cannell CF, Macdonald JC, Delchamps WF. The impact of Bayer R, Colgrove J. Science, politics, and ideology in Journalism & health news on attitudes and behavior. the campaign against environmental tobacco smoke. 1956;33(3):315–23. Mass Communication Quarterly 2002;92(6):949–54. American Journal of Public Health Centers for Disease Control and Prevention. State tobacco Bayer R, Johns D, Colgrove J. The FDA and graphic cig - - revenues compared with tobacco control appropria arette-pack warnings—thwarted by the courts. New Morbidity and Mor- tions—United States, 1998–2010. 2013;369(3):206–8. England Journal of Medicine 2012;61(20):370–4. tality Weekly Report - Best J. Economic interests and the vindication of devi - The Visible Hand: The Managerial Revo Chandler AD Jr. ance: tobacco in seventeenth century Europe. Socio- lution in American Business . Cambridge (MA): Harvard logical Quarterly 1979;20(Spring):171–82. University Press, 1977. Brandt AM. The Cigarette Century: The Rise, Fall, and , 505 U.S. 504, 112 S. Ct. Cipollone v. Liggett Group, Inc. Deadly Persistence of the Product That Defined Amer - 2608 (1992). ica . New York: Basic Books, 2007. 683 F. Supp. 1487 (D.N.J. Cipollone v. Liggett Group, Inc., Brandt AM. The cigarette, risk, and American culture. 1988). Daedalus 1990;119:155–76. Clean Air Act Extension of 1970, Public Law 91-604, U.S. British American Tobacco. Unique “Kool” campaign 84 (1970):1676. Statutes at Large Information . July 1948. has warm reception. In: 35 Fifty Years of Change 1964–2014

65 Surgeon General’s Report Doll R, Hill AB. The mortality of doctors in relation to Coe RM, Brehm HP. Smoking habits of physicians and British preventive care practices. HSMHA Health Reports their smoking habits; a preliminary report. Medical Journal 1954;1(4877):1451–5. 1971;86(3):217–21. Colley JR, Holland WW, Corkhill RT. Influence of passive Dorn HF. The mortality of smokers and nonsmokers. In: Proceedings of the Social Statistics Section of the - smoking and parental phlegm on pneumonia and bron Lancet chitis in early childhood. . Washington: Ameri- American Statistical Association 1974;2(7888):1031–4. can Statistical Association, 1958:34–71. Comstock GW, Lundin FE Jr. Parental smoking and peri - American Journal of Obstetrics and Dorn HF. Tobacco consumption and mortality from can - natal mortality. 1959; Public Health Reports cer and other diseases. Gynecology 1967;98(5):708–18. Congressional Record . Strengthening the Cigarette 74(7):581–93. . Drew EB. The quiet victory of the cigarette lobby: Congressional Record, July 27, 1966 Labeling Act. In: how it found the best filter yet—Congress. 1966:16468–75. Atlantic . 1965. Brown and Williamson Collection. Consumer Products Safety Act, Monthly U.S. Public Law 92-573, Bates No. 690002862/2866. < http://legacy.library.ucsf. 86 (1972):1207, codified at 15 U.S. Statutes at Large edu/tid/hwp01c00>. Code §§ 2501–2089. Evans AS. Causation and disease: a chronological journey. Cornfield J, Haenszel W, Hammond EC, Lilienfeld AM, Shimkin MB, Wynder EL. Smoking and lung can American Journal of Epi- - The Thomas Parran lecture. demiology 1978;108(4):249–58. cer: recent evidence and a discussion of some ques- Evans AS. Causation and disease: the Henle-Koch postu- tions. 1959; Journal of the National Cancer Institute 22(1):173–203. Yale Journal of Biology and Medicine lates revisited. 1976;49(2):175–95. Council for Tobacco Research. Summary of Dr. Hockett’s Family Smoking Prevention and Tobacco Control Act, . 1964. Council for Tobacco Comments, March 5, 1964 U.S. Statutes at Large Public Law 111-31, 123 1776 Research Collection. Bates No. HK0998002/8008. (2009). . Federal Cigarette Labeling and Advertising Act of 1965, Creek L, Capehart T, Grise V. U.S. Tobacco Statistics, 79 (1965): U.S. Statutes at Large Public Law 89–92, 1935–1992 . Statistical Bulletin No. 869. Washington: Sec. 4. U.S. Department of Agriculture, Economic Research Federal Food, Drug, and Cosmetic Act, Public Law 75-717, Service, Commodity Economics Division, 1994. U.S. Statutes at Large U.S. 52 (1938):1040, codified at Science Culliton BJ. The academic-industrial complex. 21. Code 1982;216(4549):960–2. Federal Trade Commission. A report on cigarette adver - Cullman HS. [Letter to Benjamin Sonenberg about filter - tising and output: Staff report of the Bureau of Eco type cigarettes]. 1964. Philip Morris Collection. Bates nomics. Federal Trade Commission. 1964. RJ Reynolds No. 1005038518. < http://legacy.library.ucsf.edu/tid/ http://legacy. Collection. Bates No. 515698078/8370. < doo94e00>. library.ucsf.edu/tid/ywm92d00>. Cullman HS. [Letter to Leslie Gould about three sug- Federal Trade Commission. FTC renews cigarette war. gestions]. 1962. Philip Morris Collection. Bates No. Tide. April 14, 1950. American Tobacco Collection. 1005038456/8457. . edu/tid/mzi21a00>. Cutler SJ. A review of the statistical evidence on the asso- Federal Trade Commission. Staff report on the cigarette Journal of ciation between smoking and lung cancer. advertising investigation, May 1981; ; accessed: DATTA. Notes on minutes of the Tobacco Industry Research September 27, 2013. Committee meeting, December 28, 1953. 1953. DATTA , 47 Federal Trade Commission v. American Tobacco Co. Collection. Bates No. JH000381-JH000384. < http:// FTC 1393 (1951), order modified, 48 FTC 1161 (1952). legacy.library.ucsf.edu/tid/cbl36b00>. 46 FTC 735 Federal Trade Commission v. P. Lorillard Co., The New York Public Library’s Diefendorf E, editor. In: (1950) order modified, 46 FTC 853 (1950b), aff’d, 186 New York: Oxford University Books of the Century. F.2d 52 (4th Cir. 1950). Press, 1997. Federal Trade Commission v. Philip Morris & Co., Ltd., DiFranza JR, Richards JW, Paulman PM, Wolf-Gillespie N, Inc. , 49 FTC 703 (1952) vacated and remanded on Fletcher C, Jaffe RD, Murray D. RJR Nabisco’s cartoon - motion of Commission, complaint dismissed on affida JAMA camel promotes camel cigarettes to children. vit of abandonment, 51 FTC 857 (1955). 1991;266(22):3149–53. Chapter 2 36

66 —50 Years of Progress The Health Consequences of Smoking %20Poll%2068%20-%20Industry%20Regula , 46 FTC Federal Trade Commission v. R.J. Reynolds Co. - 706 (1950), modified, 192 F.2d 535 (7th Cir 1951), on tion_12.18.12.pdf>; accessed: July 3, 2013. remand, 48 FTC 682 (1952). Haseltine N. AMA supports tobacco industry in fight on Federal Trade Commission. [Press release re: advertis- cigarette hazard labels. Danger held “common knowl - The Washington Post . March 18, 1964. edge.” ing claims of superiority among cigarette brands]. - April 5, 1950. American Tobacco Collection. Bates Henningfield JE, Zeller M. Nicotine psychopharmacol No. 980140356/0361. . tobacco regulation: a look back and a look forward. 2006;184(3-4):286–91. Feinleib M, Havlik RJ, Thom T. Conference on the decline in Psychopharmacology - Hill and Knowlton Inc. Increased stories on Surgeon Gen coronary heart disease mortality. In: Proceedings of the Conference on the Decline in Coronary Heart Disease eral’s Committee Report [memorandum]. 1963. Ameri - can Tobacco Collection. Bates No. 950181555/1557. Mortality, National Heart, Lung, and Blood Institute, . National Institutes of Health; October 24–25, 1978 . Bethesda (MD): U.S. Department of Health, Education, Hirayama T. Non-smoking wives of heavy smokers have a higher risk of lung cancer: a study from Japan. British and Welfare, Public Health Service, National Institutes 1981;282(6259):183–5. of Health, 1979. NIH Publication No. 79–1610. Medical Journal Hockett RC, Thompson C. [Cable re: memo from Hund Ford ES, Capewell S. Proportion of the decline in cardio - - vascular mortality disease due to prevention versus ley]. 1962. Council for Tobacco Research Collection. Annual treatment: public health versus clinical care. Bates No. 11308817/8817. . Review of Public Health Fritschler AL. IARC Smoking and Politics: Policymaking and International Agency for Research on Cancer. Monographs on the Evaluation of Carcinogenic Risk the Federal Bureaucracy . New York: Appleton-Cen - Vol. 38. of Chemicals to Humans: Tobacco Smoking. tury-Crofts, 1969. - Gallup Organization. Trends in public attitudes on the pos Lyon (France): International Agency for Research on sibility of a health hazard in cigarette smoking. 1964. Cancer, 1986. Tobacco Control Laws: Imple- Jacobson PD, Wasserman J. RJ Reynolds Collection. Bates No. 500006198/6249. . Washington: RAND Cor - . mentation and Enforcement poration, 1997. - Garfinkel L, Stellman SD. Cigarette smoking among phy Jaffe JH. Cigarette smoking as an addiction. In: Smoking CA: sicians and other health professionals, 1959–1972. and Health II: Health Consequences, Education, Ces- A Cancer Journal for Clinicians 1976;26(6):373–75. . DHEW sation Activities, and Governmental Action Goodman J, editor. Quitting. In: Tobacco in History and Publication No. 77-1413. Bethesda (MD): National Can- Culture: An Encyclopedia. Framington Hills (MI): cer Institute, 1977:627. Thomson Gale, 2005. Research Jarvik ME, Cullen JW, Gritz ER, Vogt TM, West LJ. Gover M. Cancer Mortality in the United States. 1. Trend NIDA Research Monograph 17. on Smoking Behavior. - of Recorded Cancer Mortality in the Death Registra Washington: U.S. Government Printing Office, 1977. tion States of 1900 from 1900 to 1935. Public Health - Jemal A, Ward E, Hao Y, Thun M. Trends in the lead Bulletin No. 248. Washington: Government Printing ing causes of death in the United States, 1970–2002. Office, 1939. JAMA: the Journal of the American Medical Association Gwynne P. The chemicals around us. August Newsweek 2005;294(10):1255–9. 21, 1978;National Affairs:25. Ashes to Ashes: America’s Hundred-Year Ciga- Kluger R. Haas FP. Re: a meeting of the Tobacco Institute Executive - rette War, the Public Health, and the Unabashed Tri Committee on 1/12/64 [memorandum]. 1964. Research umph of Philip Morris . New York: Alfred A. Knopf, 1996. http:// Collection. Bates No. LG2008203-LG2008210. < Kozlowski LT, O’Connor RJ. Dealing with health fears: cig - legacy.library.ucsf.edu/tid/skj66b00>. arette advertising in the United States in the twentieth Hammond EC, Horn D. Smoking and death rates; report century. In: Boyle P, Gray N, Henningfield J, Seffrin J, on forty-four months of follow-up of 187,783 men: I. Tobacco and Public Health: Sci Zatonski W, editors. - JAMA: the Journal of the American Total mortality. ence and Policy . 2nd ed. New York: Oxford University 1958;166(10):1159–72. Medical Association Press, 2010:31–43. Harris Poll. Oil, Pharmaceutical, Health Insurance, Langmuir AD. The surveillance of communicable diseases - Tobacco, Banking and Utilities Top the List of Indus New England Journal of Medi of national importance. - tries that People Would Like to See More Regulated, 1963;268(4):182–92. cine 2012;

67 Surgeon General’s Report - Lewit EM, Coate D, Grossman M. The effects of govern U.S. Department of Health and Human Services, ment regulation on teenage smoking. National Institutes of Health, National Cancer Insti - Journal of Law and Economics 1981;24(3):545–69. tute, 2008. NIH Publication No. 07-6242. Use of Lilienfeld DE. Definitions of epidemiology. American National Clearinghouse for Smoking and Health. 1978;107(2):87–90. Tobacco. Practices, Attitudes, Knowledge, and Beliefs: Journal of Epidemiology . Washington: United States, Fall 1964 and Spring 1966 Little CC. Some phases of the problem of smoking and New England Journal of Medicine lung cancer. U.S. Department of Health, Education and Welfare, Public Health Service, 1969. 1961;264:1241–5. National Commission on Smoking and Public Policy. A Macklin MT. Has a real increase in lung cancer been National Dilemma: Cigarette Smoking or the Health 1942;17(2): Annals of Internal Medicine proved? 308–24. of Americans: Report of the National Commission on Epidemiologic Methods . Smoking and Public Policy to the Board of Directors, MacMahon B, Pugh TF, Ipsen J. American Cancer Society, Inc. Atlanta (GA): American Boston: Little, Brown and Company, 1960. Meighan SS, Weitman M. Smoking: habits and beliefs of Cancer Society, 1978. Oregon physicians. Journal of the National Cancer Environmental Tobacco National Research Council. 1965;35(5):893–8. Smoke: Measuring Exposures and Assessing Health Institute Effects . Washington: National Academies Press, 1986. Miura M, Daynard RA, Samet JM. The role of litiga- Salud Publica de México tion in tobacco control. Neuberger MB. [Copy of letter to Honorable Paul Rand 2006;48(Suppl 1):S121–S136. Dixon, chairman, Federal Trade Commission on for - - Moore DW. Nine of ten Americans view smoking as harm mulating a response to the Surgeon General’s report on smoking and health]. 1964. Tobacco Institute Col - - ful: public blames smokers more than tobacco compa nies for smoking-related problems, October 7, 1999; lection. Bates No. TI55731588-TI55731591. . ; accessed: May 10, Neuberger MB. [Letter to Paul Rand Dixon, Federal Trade 2013. Commission, January 6, 1964]. 1964. American Tobacco http://legacy. Collection. Bates No. 946144914/4917. < Morris JN. Uses of Epidemiology . Edinburgh (Scotland): library.ucsf.edu/tid/vhh41a00>. E. & S. Livingston, 1957. Nathanson CA. Social movements as catalysts for policy New York Times . U.S. links cancer with cigarettes. New http://legacy.library.ucsf. York Times July 13, 1957:1, < change: the case of smoking and guns. Journal of edu/tid/fjt88h00/pdf>. Health Politics, Policy and Law 1999;24(3):421–88. Newport F. For first time, majority in U.S. supports public National Cancer Institute. Tobacco and the Clinician: smoking ban: little support for making smoking ille - Interventions for Medical and Dental Practice. Smok- http://www.gallup.com/ gal, however, July 15, 2011; < ing and Tobacco Control Monograph No. 5. Bethesda - poll/148514/first-time-majority-supports-public-smok (MD): U.S. Department of Health and Human Services, ing-ban.aspx>; accessed: December 19, 2012. Public Health Service, National Institutes of Health, Ochsner A. Corner of history. My first recognition of the National Cancer Institute, 1994. relationship of smoking and lung cancer. Preventive National Cancer Institute. Risks Associated With Smok- Medicine 1973;2(4):611–4. ing Cigarettes With Low Machine-Measured Yields Ochsner N, DeBakey M. Symposium on Cancer. Primary Smoking and Tobacco Control of Tar and Nicotine. pulmonary malignancy. Treatment by total pneumo - Monograph No. 13. Bethesda (MD): U.S. Department - nectomy; analyses of 79 collected cases and presen of Health and Human Services, Public Health Service, tation of 7 personal cases. Surgery, Gynecology and National Institutes of Health, National Cancer Insti - Obstetrics 1939;638:435–51. tute, 2001. NIH Publication No. 02-5047. Orentlicher D. The FDA’s graphic tobacco warnings and ASSIST: Shaping the Future National Cancer Institute. - New England Journal of Medi the first amendment. of Tobacco Prevention and Control. Tobacco Control 2013;369(3):204–6. cine Monograph No. 16. Bethesda (MD): U.S. Department Pacey J. [Press release re: response to inquiries on the Sur - of Health and Human Services, National Institutes of geon General’s report]. 1964. Philip Morris Collection. Health, National Cancer Institute, 2005. NIH Publica - Bates No. 1005038555/8556. < http://legacy.library.ucsf. tion No. 05-5645. edu/tid/eno94e00>. National Cancer Institute. The Role of the Media in Pro - Parascandola J. The Surgeons General and smoking. Pub- Smoking and moting and Reducing Tobacco Use. lic Health Reports 1997;112(5):440–2. Tobacco Control Monograph No. 19. Bethesda (MD): Chapter 2 38

68 —50 Years of Progress The Health Consequences of Smoking Parascandola M, Weed DL, Dasgupta A. Two Surgeon Gen - Saad L. Tobacco and smoking, August 15, 2002; ; eral’s reports on smoking and cancer: a historical inves www.gallup.com/poll/9910/tobacco-smoking.aspx#4 - Emerging tigation of the practice of causal inference. accessed: July 2, 2013. Themes in Epidemiology 2006;3:1. Sachs DP. Smoking habits of pulmonary physicians. New England Journal of Medicine 1983;309(13):799. Parascandola M. Causes, risks, and probabilities: probabi- listic concepts of causation in chronic disease epidemi- Sack K. For the Nation’s Politicians, Big Tobacco No Longer Bites. ology. Preventive Medicine 2011;53(4-5):232–4. April 22, 1997; New York Times ; accessed: May 10, cal debate. 2004;47(2):244–61. 2013. Patterson JT. The Dread Disease: Cancer and Modern U.S. Stat- Public Law 93-523, Safe Drinking Water Act, 88 (1974):1660, codified at American Culture - utes at Large U.S. Code . Cambridge (MA): Harvard Univer 42, § 300F. sity Press, 1987. Sarna L, Bialous SA, Sinha K, Yang Q, Wewers ME. Are Science Pearl R. Tobacco smoking and longevity. health care providers still smoking? Data from the 1938;87(2253):216–7. 2003 and 2006/2007 Tobacco Use Supplement-Current Pierce JP, Choi WS, Gilpin EA, Farkas AJ, Berry CC. - Nicotine & Tobacco Research Population Surveys. Tobacco industry promotion of cigarettes and adoles 2010;12(11):1167–71. 1998;279(7):511–5. JAMA cent smoking. Shea JW. [Letter to P. Lorillard Company]. 1966. Lorillard Pollay Advertising. A Frank Statement to Cigarette Collection. Bates No. 00065004. . No. MISC0301. . - CA. States’ allocations of funds from the Tobacco Mas Proctor RN. Angel H Roffo: the forgotten father of experi - 2005;24(1): Health Affairs ter Settlement Agreement. Bulletin of the World mental tobacco carcinogenesis. 220–7. 2006;84(6):494–6. Health Organization Smith RC. The magazines’ smoking habit. Columbia Jour- Proctor RN. Golden Holocaust: Origins of the Cigarette 1978;16:29–31. nalism Review . Berkeley (CA): Catastrophe and the Case for Abolition Snegireff LS, Lombard OM. Smoking habits of Massachu - University of California Press, 2012. setts physicians; five-year follow-up study (1954–1959). Public Law Public Health Cigarette Smoking Act of 1969, 1959;261:603–4. New England Journal of Medicine U.S. Statutes at Large 87 (1970), codified at 91-222, 84 Snegireff LS, Lombard OM. Survey of smoking habits of 15 § 1331. U.S. Code New England Journal of Massachusetts physicians. Racketeer Influenced and Corrupt Organizations (“RICO”) 1954;250(24):1042–5. Medicine Public Law 91-452, 84 Stat. 922 (1970), codified at Act, Steinfeld JL. Remarks by Jesse L. Steinfeld, M.D. Surgeon 18 §§ 1961–68, (1994). U.S. Code General, U.S. Public Health Service. 1971. Lorillard Rasmussen N. Maurice Seevers, the stimulants and the http://legacy. Collection. Bates No. 91018247/8260. < political economy of addiction in American biomedi- library.ucsf.edu/tid/fhu60e00>. cine. BioSocieties 2009;5:105–23. Steinfeld JL. [Remarks at press briefing on 1972 report, - Robinson EJ. Analyzing the impact of science report “The Health Consequences of Smoking”]. 1972. Tobacco ing. Journalism & Mass Communication Quarterly Institute Collection. Bates No. TITX0004900/4909. 1963;40(3):306–14. . Public Health and the Risk Factor: A His - Rothstein WG. Stern MP. The recent decline in ischemic heart disease tory of an Uneven Medical Revolution . Rochester (NY): Annals of Internal Medicine 1979;91(4): mortality. University of Rochester Press, 2003. 630–40. Royal College of Physicians. Smoking and Health: Sum- Study Group on Smoking and Health. Smoking and mary and Report of the Royal College of Physicians of health: joint report of the Study Group on Smoking London on Smoking in Relation to Cancer of the Lung 1957;125:1129–33. and Health. Science . London: Pitman Medical Publish and Other Diseases - Susser M. Invited commentary on “causation and disease: ing Co., 1962. a chronological journey. The Thomas Parran lecture”. Saad L. A half-century of polling on tobacco: most 1995;142(11):1125. American Journal of Epidemiology Public Perspective don’t like smoking but tolerate it. 1998;9(5):1–4. 39 Fifty Years of Change 1964–2014

69 Surgeon General’s Report U.S. Department of Health and Human Services. Reduc- Tate C. Cigarette Wars: The Triumph of “The Little White Slaver.” New York: Oxford University Press, 1999. ing the Health Consequences of Smoking: 25 Years of Principles of Epidemiology Taylor I, Knowelden J. . Lon- Progress. A Report of the Surgeon General . Rockville don (England): J. & A. Churchill, 1957. (MD): U.S. Department of Health and Human Services, Terry LL. The Surgeon General’s first report on smoking Public Health Service, Centers for Disease Control, National Center for Chronic Disease Prevention and New and health. A challenge to the medical profession. 1983;83(13):1254–5. Health Promotion, Office on Smoking and Health, York State Journal of Medicine 1989. DHHS Publication No. (CDC) 89-8411. Toch H. Long term effects of the cancer scare: the residue of news impact. Journalism Quarterly Prevent- U.S. Department of Health and Human Services. 1961;38:25–34. ing Tobacco Use Among Young People: A Report of the New Toth R. Celebrezze wary on smoking issue. York Times Surgeon General . Atlanta (GA): U.S. Department of http://legacy.library.ucsf.edu/tid/ . 1962. < mtr23a00>. Health and Human Services, Public Health Service, Public Law 94-469, U.S. Toxic Substances Control Act, Centers for Disease Control and Prevention, National - Center for Chronic Disease Prevention and Health Pro Statutes at Large 90 (1976), codified at 15, U.S. Code motion, Office on Smoking and Health, 1994. §§ 2601–2629. Trichopoulos D, Kalandidi A, Sparros L, MacMahon B. U.S. Department of Health and Human Services. For International Jour- a Healthy Nation: Returns on Investment in Public Lung cancer and passive smoking. nal of Cancer 1981;27(1):1–4. . Atlanta (GA): U.S. Department of Health and Health Human Services, Public Health Service, Office of Dis - U.S. Census Bureau. Population estimates, 2013, < http:// ease Prevention and Health Promotion, and Centers for www.census.gov/popest/>; accessed: July 30, 2013. Disease Control and Prevention, 1995. U.S. Congress. False and Misleading Advertising (Filter- U.S. Department of Health and Human Services. Tip Cigarettes): Hearings Before a Subcommittee of the Reducing Tobacco Use. A Report of the Surgeon General Committee on Government Operations, House of Rep . Atlanta - (GA): U.S. Department of Health and Human Services, resentatives, 85th Congress, 1st Session 1957, p. 147. Centers for Disease Control and Prevention, National - U.S. Congress. Cigarette Labeling and Advertising: Hear ings Before the House Committee on Interstate and Center for Chronic Disease Prevention and Health Pro - motion, Office on Smoking and Health, 2000. Foreign Commerce, 88th Congress, 2nd Session, 1964, U.S. Department of Health and Human Services. p. 18. The Health Consequences of Involuntary Exposure to U.S. Congress. Cigarette Labeling and Advertising—1965: Tobacco Smoke: A Report of the Surgeon General . Hearings on H.R. 2248 Before the House Committee on Atlanta (GA): U.S. Department of Health and Human Interstate and Foreign Commerce, 89th Congress, 1st Services, Centers for Disease Control and Prevention, Session, 1965, p. 193. Coordinating Center for Health Promotion, National . TBS- Tobacco Outlook U.S. Department of Agriculture. Center for Chronic Disease Prevention and Health Pro - 263. Washington: U.S. Department of Agriculture, Eco - motion, Office on Smoking and Health, 2006. nomics Research Service, 2007; < http://usda.mannlib. Pre- U.S. Department of Health and Human Services. cornell.edu/usda/ers/TBS//2000s/2007/TBS-10-24- venting Tobacco Use Among Youth and Young Adults: 2007.pdf>; accessed: July 15, 2013. A Report of the Surgeon General . Atlanta (GA): U.S. U.S. Department of Health and Human Services. The Department of Health and Human Services, Centers Health Consequences of Involuntary Smoking: A for Disease Control and Prevention, National Center Report of the Surgeon General . Rockville (MD): U.S. for Chronic Disease Prevention and Health Promotion, Department of Health and Human Services, Public Office on Smoking and Health, 2012. Health Service, Centers for Disease Control, Center Smok- U.S. Department of Health, Education, and Welfare. - for Health Promotion and Education, Office on Smok ing and Health: Report of the Advisory Committee to ing and Health, 1986. DHHS Publication No. (CDC) . the Surgeon General of the Public Health Service 87-8398. Washington: U.S. Department of Health, Education, The U.S. Department of Health and Human Services. and Welfare, Public Health Service, Center for Disease Health Consequences of Smoking: Nicotine Addic - Control, 1964. PHS Publication No. 1103. . Atlanta (GA): tion. A Report of the Surgeon General If U.S. Department of Health, Education, and Welfare. U.S. Department of Health and Human Services, Public You Must Smoke ... 5 Ways to Reduce the Risks of Health Service, Centers for Disease Control, National . Public Health Service Publication No. 1786. Smoking Center for Chronic Disease Prevention and Health Pro - Arlington (VA): National Clearinghouse for Smoking motion, Office on Smoking and Health, 1988. DHHS and Health, 1968. Publication No. (CDC) 88-8406. Chapter 2 40

70 The Health Consequences of Smoking —50 Years of Progress Wakeham H. [Letter re: article by Dr. Joseph Berkson]. Use of U.S. Department of Health, Education, and Welfare. 1964. Philip Morris Collection. Bates No. 1005038550. Tobacco: Practices, Attitudes, Knowledge, and Beliefs. . . Public United States—Fall 1964 and Spring 1966 Warner KE. Clearing the airwaves: the cigarette ad ban Health Service Publication No. 1786. Washington: revisited. 1979;5(4):435–50. Policy Analysis National Clearinghouse for Smoking and Health, 1969. Warner KE. Cigarette advertising and media coverage of U.S. Department of Health, Education, and Welfare. The smoking and health. Health Consequences of Smoking. A Report of the Sur New England Journal of Medicine - 1985a;312(6):384–8. . Washington: U.S. Department of geon General: 1972 Warner KE. Tobacco industry response to public health Health, Education, and Welfare, Public Health Service, Health Services and Mental Health Administration, Health concern: a content analysis of cigarette ads. Education Quarterly 1985b;12(2):115–27. 1972. DHEW Publication No. (HSM) 72-7516. Warner KE, Goldenhar LM. The cigarette advertising U.S. Department of Health, Education, and Welfare. Adult broadcast ban and magazine coverage of smoking and Use of Tobacco 1970 . Washington: U.S. Department of Journal of Public Health Policy 1989;10(1): Health, Education, and Welfare, Public Health Service, health. 32–42. Center for Disease Control, National Clearinghouse for http://legacy.library.ucsf. Warner KE, Goldenhar LM, McLaughlin CG. Cigarette Smoking and Health, 1973a; < advertising and magazine coverage of the hazards of edu/action/document/page?tid=ryp52f00>. New England Journal of smoking. A statistical analysis. U.S. Department of Health, Education, and Welfare. If you - must smoke ... 1992;326(5):305–9. Medicine Bethesda (MD): National Clearing house for Smoking and Health, 1973b. Weissman G. [Letter re: thoughts on the 1964 Surgeon General’s Report]. 1964. Research Collection. Bates No. U.S. Department of Health, Education, and Welfare. Smoking and Health. A Report of the Surgeon Gen 1005038559-1005038561. . . Washington: U.S. Department of Health, Educa White JR, Froeb HF. Small-airways dysfunction in non- tion, and Welfare, Office of the Assistant Secretary for New Health, Office on Smoking and Health, 1979. DHEW smokers chronically exposed to tobacco smoke. Publication No. (PHS) 79-50066. 1980;302(13):720–3. England Journal of Medicine Witschi H. A short history of lung cancer. Toxicological U.S. Department of the Treasury. Tobacco Industry Statis- 2001;64(1):4–6. tics, 2013; ; Sciences World Health Organization. WHO Expert Committee on accessed: July 30, 2013. Respiratory . World U.S. Environmental Protection Agency. Addiction-Producing Drugs. Seventh Report Health Organization Technical Report Series No. 116. Health Effects of Passive Smoking: Lung Cancer and Geneva (Switzerland): World Health Organization, Other Disorders . Washington: U.S. Environmental Pro- tection Agency, Office of Research and Development, 1957. Wynder EL, Graham EA, Croninger AB. Experimental Office of Health and Environmental Assessment, 1992. United States v. Philip Morris , 449 F. Supp. 2d 1 (D.D.C. production of carcinoma with cigarette tar. Cancer 2006). Research 1953;13(12):855–64. University of California at San Francisco. Legacy Tobacco - Yerushalmy J, Palmer CE. On the methodology of investi Documents Library, 2013; < http://legacy.library.ucsf. Journal gations of etiologic factors in chronic diseases. of Chronic Diseases 1959;10(1):27–40. edu/>; accessed: May 9, 2013. 41 Fifty Years of Change 1964–2014

71

72 Chapter 3 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose Introduction 45 Development of a Scientific Consensus 45 Development of the Criteria for Causation Used in the 1964 Report 51 Evolution of the Application of the Criteria for Disease Causation in Subsequent Reports 52 Methods for Reviewing the Evidence and Developing Conclusions 53 Process of Ensuring Consensus and Strength of the Peer Review 55 Separation of Scientific Conclusions and the Formation of Policy 56 References 57 43

73

74 —50 Years of Progress The Health Consequences of Smoking Introduction In 1964, U.S. Surgeon General Luther L. Terry The reports of the Surgeon General have developed a formal framework for assessing evidence on disease causa - appointed an expert committee to submit a report to tion, and the formats of the reports have provided detailed review and evaluate the current data on smoking and presentations of the scientific evidence underlying each health. The publication of the committee’s report, Smok- of their conclusions (USDHEW 1964; U.S. Department of ing and Health: Report of the Advisory Committee to the (U.S. Department of Health and Human Services [USDHHS] 2004). The reports Surgeon General of the United States Health, Education, and Welfare [USDHEW] 1964), marked have been produced using a balanced and comprehensive review and editorial process to ensure that the evidence, the first of a long series of reports issued by the Office of the Surgeon General to the American people on smoking - rather than the authors’ opinions, defines the conclu sions. Across the five decades of reports, the emphasis has and health (Table 3.1). This series of reports, released over been on the evidence base and the scientific validity of the a 50-year period, comprises a remarkable set of scientific documents that have provided internationally accepted conclusions, and scientific conclusions have been clearly consensus judgments on the scientific evidence implicat - separated from any policy decisions that may result from the findings. ing smoking in disease causation. In addition, the reports The result of the work undertaken over this 50-year have built a scientific foundation to support tobacco con - trol programs and interventions intended to control the period has been a series of reports that have maintained - their utility and credibility despite marked shifts in gov epidemic of tobacco-caused disease (see Chapter 2, “Fifty Years of Change—1964–2014”). The reports have also been ernmental policies toward tobacco, powerful opposi- invaluable to the scientific community by highlighting tion from tobacco industry interests, and the sometimes what is known in this area and identifying the critical evi heated debates on science and policy that have taken place - dence gaps to be addressed with further research. Finally, within the tobacco control community. This chapter cov - the methods for reviewing evidence and causal inference ers the production and evolution of the reports during the have been widely applied in other contexts (Rothman and past 50 years, emphasizing the processes that have sus - tained their utility. Greenland 1998). Development of a Scientific Consensus scientists was asked to review all of the available evidence Often considered the first report of the Surgeon General on the health consequences of smoking, the 1964 on possible links between cigarette smoking and disease - and to form a scientific judgment on this issue. In addi report on smoking and health was actually (as noted in tion, the scientists were expected to report back to the the introduction above) a report of an expert Advisory Committee to the Surgeon General (USDHEW 1964). Surgeon General with a solid evidence-based foundation Although this report is widely viewed as pivotal in estab- for appropriate remedial action by the U.S. Public Health lishing with certainty that cigarette smoking causes lung Service (PHS) responding to the emerging epidemic of lung cancer being caused by a highly profitable consumer cancer, a similar conclusion with regard to causation had been reached earlier by several scientific reviews and by product, the cigarette. Surgeon General Leroy E. Burney (see Chapter 2). To ensure transparency, the committee codified the criteria used to reach the conclusion that smoking causes Given the rising evidence and to once again criti- cally review the cumulative evidence, the Surgeon Gen- lung cancer. Both the resulting systematic, transparent review and the synthesis of evidence using those criteria eral in 1962 convened an independent group of scientists were pioneering for the time. who had not up to that time publicly expressed an opinion on whether smoking caused lung cancer. This group of 45 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose

75 Surgeon General’s Report Surgeon General’s reports on smoking and health, 1964–2012 Table 3.1 Number Title Surgeon General Subject/highlights Year of pages 387 First official report of the federal government on Luther L. Terry, 1964 Smoking and Health: Report of M.D. smoking and health. Concluded that “Cigarette smoking the Advisory Committee to the is a health hazard of sufficient importance in the United Surgeon General of the Public States to warrant appropriate remedial action” (p. 33). Health Service Also concluded that cigarette smoking is a cause of lung cancer in men and a suspected cause of lung cancer in women. The report was also responsible for the passage of the Cigarette Labeling and Advertising Act of 1965, which among other things, mandated the familiar Surgeon General’s health warnings on cigarette packages (USDHEW 1964). 199 The Health Consequences of 1967 Confirmed and strengthened conclusions of the 1964 William H. Smoking: A Public Health Stewart, M.D. report. Stated that “the case for cigarette smoking as the Service Review principal cause of lung cancer is overwhelming” (p. 16). Found that evidence “strongly suggests that cigarette smoking can cause death from coronary heart disease” (p. 26), which was upgraded from the 1964 conclusion of an “association.” Also concluded that “Cigarette smoking is the most important of the causes of chronic non- neoplastic bronchopulmonary diseases in the United States” (p. 31) (USDHEW 1967). 117 Updated information that was presented in the 1967 William H. The Health Consequences of 1968 report. Estimated that smoking-related loss of life Stewart, M.D. Smoking: 1968 Supplement to expectancy among young men as 8 years for “heavy” the 1967 Public Health Service smokers (more than 2 packs/day) and 4 years for “light” Review smokers (less than ½ pack/day) (USDHEW 1968). 98 Also supplemented 1967 report. Confirmed association William H. 1969 The Health Consequences of Stewart, M.D. between maternal smoking and infant low birthweight. Smoking: 1969 Supplement to Identified evidence of increased incidence of the 1967 Public Health Service prematurity, spontaneous abortion, stillbirth, and Review neonatal death (USDHEW 1969). Jesse L. Steinfeld, Reviewed entire field of smoking and health with 1971 The Health Consequences of 458 Smoking M.D. emphasis on most recent literature. Discussed new data indicating associations between smoking and peripheral vascular disease, atherosclerosis of the aorta and coronary arteries, increased incidence and severity of respiratory infections, and increased mortality from cerebrovascular disease and nonsyphilitic aortic aneurysm. Concluded that smoking is associated with cancers of the oral cavity and esophagus. Found that “Maternal smoking during pregnancy exerts a retarding influence on fetal growth” (p. 13) (USDHEW 1971). 46 Chapter 3

76 The Health Consequences of Smoking —50 Years of Progress Continued Table 3.1 Number Title Surgeon General of pages Year Subject/highlights 158 Examined evidence on immunologic effects of tobacco The Health Consequences of 1972 Jesse L. Steinfeld, Smoking M.D. and tobacco smoke, harmful constituents of tobacco smoke, and “public exposure to air pollution from tobacco smoke” (p. 121). Found tobacco and tobacco smoke antigenic in humans and animals; tobacco may impair protective mechanisms of immune system; nonsmokers’ exposure to tobacco smoke may exacerbate allergic symptoms; carbon monoxide in smoke-filled rooms may harm health of persons with chronic lung or heart disease; tobacco smoke contains hundreds of compounds, several of which have been shown to act as carcinogens, tumor initiators, and tumor promoters. Identified carbon monoxide, nicotine, and tar as smoke constituents most likely to produce health hazards of smoking (USDHEW 1972). 1973 The Health Consequences of Merlin K. DuVal, 249 Presented evidence on health effects of smoking pipes, a M.D. Smoking cigars, and “little cigars.” Found mortality rates of pipe and cigar smokers higher than those of nonsmokers but lower than those of cigarette smokers. Found that cigarette smoking impairs exercise performance in healthy young men. Presented additional evidence on smoking as a risk factor in peripheral vascular disease and problems of pregnancy (USDHEW 1973). 1974 The Health Consequences of Charles C. 124 Tenth anniversary report. Reviewed and strengthened a Smoking evidence on major hazards of smoking. Reviewed Edwards evidence on association between smoking and atherosclerotic brain infarction and on synergistic effect of smoking and asbestos exposure in causing lung cancer (USDHEW 1974). The Health Consequences of 1975 Theodore Cooper, 235 Updated information on health effects of involuntary a Smoking M.D. (passive) smoking. Noted evidence linking parental smoking to bronchitis and pneumonia in children during the first year of life (USDHEW 1975). 1976 The Health Consequences of Theodore Cooper, Compiled selected chapters from 1971–1975 reports 657 a M.D. Smoking: Selected Chapters (USDHEW 1976). from 1971 through 1975 60 The Health Consequences of Julius B. 1979 Combined 2-year report focused on smoking-related Richmond, M.D. health problems unique to women. Cited studies Smoking, 1977–1978 showing that use of oral contraceptives potentiates harmful effects of smoking on the cardiovascular system (USDHEW 1979b). 47 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose

77 Surgeon General’s Report Continued Table 3.1 Number Year Subject/highlights Title of pages Surgeon General 1,194 Smoking and Health Julius B. Fifteenth anniversary report. Presented most 1979 Richmond, M.D. comprehensive review of health effects of smoking ever published, and first Surgeon General’s report to carefully examine behavioral, pharmacologic, and social factors influencing smoking; to consider role of adult and youth education in promoting nonsmoking; and to review health consequences of smokeless tobacco. Many new sections, including one identifying smoking as “one of the primary sources of drug interactions in man” (p. 12-22) (USDHEW 1979a). 359 Devoted to health consequences of smoking for women. Julius B. 1980 The Health Consequences of Richmond, M.D. Reviewed evidence that strengthened previous findings Smoking for Women and permitted new ones. Noted projections that lung cancer would surpass breast cancer as leading cause of cancer mortality in women. Identified trend toward increased smoking by adolescent females (USDHHS 1980). 252 The Health Consequences Julius B. Examined health consequences of “the changing 1981 Richmond, M.D. cigarette” (i.e., lower tar and nicotine cigarettes). of Smoking–The Changing Concluded that lower yield cigarettes reduced risk of Cigarette lung cancer, but found no conclusive evidence that they reduced risk of cardiovascular disease, COPD, and fetal damage. Noted possible risks from additives and their products of combustion. Discussed compensatory smoking behaviors that might reduce potential risk of lower yield cigarettes. Emphasized that there is no safe cigarette and that any risk reduction associated with lower yield cigarettes would be small compared with benefits of quitting smoking (USDHHS 1981). 322 C. Everett Koop, 1982 The Health Consequences of Reviewed and extended understanding of the health consequences of smoking as a cause or contributory Smoking–Cancer M.D. factor of numerous cancers. Included first Surgeon General’s report consideration of emerging epidemiologic evidence of increased lung cancer risk in nonsmoking wives of smoking husbands. Did not find evidence at that time sufficient to conclude that relationship was causal, but labeled it “a possible serious public health problem” (p. 9). Discussed potential for low-cost smoking cessation interventions (USDHHS 1982). 384 Examined health consequences of smoking for C. Everett Koop, The Health Consequences 1983 cardiovascular disease. Concluded that cigarette M.D. of Smoking–Cardiovascular smoking is 1 of 3 major independent causes of CHD and, Disease given its prevalence, “should be considered the most important of the known modifiable risk factors for CHD” (p. 6). Discussed relationships between smoking and other forms of cardiovascular disease (USDHHS 1983). 48 Chapter 3

78 The Health Consequences of Smoking —50 Years of Progress Continued Table 3.1 Number Subject/highlights of pages Surgeon General Title Year 545 1984 The Health Consequences of C. Everett Koop, Reviewed evidence on smoking and COLD. Concluded Smoking–Chronic Obstructive M.D. that smoking is the major cause of COLD, accounting Lung Disease for 80–90% of COLD deaths in the United States. Noted that COLD morbidity has greater social impact than COLD mortality because of extended disability periods of COLD victims (USDHHS 1984). 542 1985 The Health Consequences of C. Everett Koop, Examined relationship between smoking and hazardous Smoking–Cancer and Chronic substances in the workplace. Found that for the majority M.D. Lung Disease in the Workplace of smokers, smoking is a greater cause of death and disability than their workplace environment. Risk of lung cancer from asbestos exposure characterized as multiplicative with smoking exposure. Observed special importance of smoking prevention among blue-collar workers because of their greater exposure to workplace hazards and their higher prevalence of smoking (USDHHS 1985). 359 The Health Consequences of C. Everett Koop, Focused on involuntary smoking, concluding that 1986 Involuntary Smoking M.D. “Involuntary smoking is a cause of disease, including lung cancer, in healthy nonsmokers” (p. 7). Also found that, compared with children of nonsmokers, children of smokers have higher incidence of respiratory symptoms and infections and reduced rates of increase in lung function. Presented detailed examination of growth in restrictions on smoking in public places and workplaces. Concluded that simple separation of smokers and nonsmokers within same airspace reduces but does not eliminate exposure to environmental tobacco smoke (USDHHS 1986a). 195 C. Everett Koop, 1986 Special report of advisory committee appointed by the The Health Consequences of Surgeon General to study the health consequences of Using Smokeless Tobacco M.D. smokeless tobacco. Concluded that use of smokeless tobacco can cause cancer in humans and can lead to nicotine addiction (USDHHS 1986b). 639 The Health Consequences of C. Everett Koop, 1988 Established nicotine as a highly addictive substance, M.D. Smoking–Nicotine Addiction comparable in its physiological and psychological properties to other addictive substances of abuse (USDHHS 1988). 703 Twenty-fifth anniversary report highlighted the C. Everett Koop, Reducing the Health 1989 M.D. dramatic progress that was achieved since the first Consequences of Smoking–25 report was issued in 1964. Highlighted important gains Years of Progress in preventing smoking and smoking-related disease, reviewed changes in programs and policies designed to reduce smoking, and emphasized remaining challenges (USDHHS 1989). 49 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose

79 Surgeon General’s Report Continued Table 3.1 Number Year Subject/highlights Title of pages Surgeon General The Health Benefits of Smoking 1990 Antonia C. Examined how an individual’s risk of smoking-related 628 Cessation Novello, M.D., diseases declines after quitting smoking (USDHHS M.P.H. 1990). Smoking and Health in the Antonia C. 1992 Reviewed broad issues surrounding production and 213 Novello, M.D., Americas consumption of tobacco in the Americas (USDHHS M.P.H. 1992). 314 1994 M. Joycelyn Preventing Tobacco Use Among Addressed the crucial problems of adolescent tobacco use by providing a detailed look at adolescence, the time Elders, M.D. Young People of life when most tobacco users begin, develop, and establish their smoking behavior (USDHHS 1994). 332 Tobacco Use Among U.S. Racial/ David Satcher, Described the 4 major U.S. racial/ethnic minority 1998 Ethnic Minority Groups M.D., Ph.D. groups—African Americans, American Indians and Alaska Natives, Asian Americans and Pacific Islanders, and Hispanics—patterns of tobacco use, adverse health effects, and the effectiveness of interventions in terms of tobacco’s cultural and socioeconomic effects on the members of these groups. This report described the complex factors that play a part in the growing epidemic of diseases caused by tobacco use in these 4 groups (USDHHS 1998). 462 Reducing Tobacco Use David Satcher, 2000 First report to offer a composite review of the various M.D., Ph.D. methods used to reduce and prevent tobacco use. This report evaluated each of the 5 major approaches to reducing tobacco use: educational, clinical, regulatory, economic, and comprehensive (USDHHS 2000). 675 Concluded that the increased likelihood of lung Women and Smoking 2001 David Satcher, M.D., Ph.D. cancer, cardiovascular disease, and reproductive health problems among female smokers make tobacco use a serious women’s health issue (USDHHS 2001). 941 The Health Consequences of Richard 2004 Concluded that smoking causes diseases in nearly every Carmona, M.D., organ of the body. Also concluded that cigarette smoking Smoking is causally linked to leukemia, cataracts, pneumonia, M.P.H. and cancers of the cervix, kidney, pancreas, and stomach (USDHHS 2004). 709 Concluded that there is no risk-free level of exposure to Richard 2006 The Health Consequences secondhand smoke. Found that even brief secondhand Carmona, M.D., of Involuntary Exposure to smoke exposure can cause immediate harm. The M.P.H. Tobacco Smoke report said the only way to protect nonsmokers from the dangerous chemicals in secondhand smoke is to eliminate smoking indoors (USDHHS 2006). 704 Regina Benjamin, Described in detail the specific pathways by which How Tobacco Smoke Causes 2010 tobacco smoke damages the human body (USDHHS M.D., M.B.A. Disease—The Biologic and 2010). Behavioral Basis for Smoking- Attributable Disease 50 Chapter 3

80 The Health Consequences of Smoking —50 Years of Progress Continued Table 3.1 Number of pages Subject/highlights Year Surgeon General Title 899 Preventing Tobacco Use Among Updated the 1994 report on youth and described the Regina Benjamin, 2012 Youth and Young Adults M.D., M.B.A. epidemic of tobacco use among youth 12–17 years of age and young adults 18–25 years of age, including the epidemiology, causes, and health effects of this tobacco use and interventions proven to prevent it (USDHHS 2012). COPD coronary heart disease; COLD = chronic obstructive lung disease; = chronic obstructive pulmonary disease; CHD Note: = USDHHS = U.S. Department Health and Human Services. USDHEW = U.S. Department of Health, Education and W elfare; a Assistant Secretary for Health issued the report. Development of the Criteria for Causation Used t in the 1964 Repor The 2004 Surgeon General’s report, The meaning of the word “cause” has a long and The Health , described the subsequent Consequences of Smoking rich philosophical history; the term has been applied vari- ably in different scientific contexts. Among these contexts refinement of the 1964 language for application in public have been the demonstration of causation experimen- health and epidemiologic considerations as “The qualita - tally in the laboratory, the causal attribution of a specific tive judgment that an exposure causes a particular disease infectious disease to a specific microbiologic agent, and the signifies that in the absence of exposure some fraction of understanding of the mechanism(s) leading to a disease. cases or deaths would not occur or would occur at a later In the years before the creation of the Advisory Commit - age” (USDHHS 2004, p. 10). The 1964 report of the Advisory Committee clearly tee, the groups which considered the question of cigarette stated that the decision that cigarette smoking was a cause smoking as a cause of lung cancer recognized that these of lung cancer in men resulted from a judgment based scientific contexts, and the resulting context-specific con - structs of causation, could not be readily applied to the on a synthesis of all of the available evidence, rather than question of whether cigarette smoking caused human the outcome of a single scientific study or a single line of disease (Blackburn and Labarthe 2012; Glass et al. 2013). evidence. Specifically, the report (USDHEW 1964, p. 20) noted: Obtaining direct experimental evidence in humans is an ethical impossibility and understanding the mechanisms Statistical methods cannot establish proof of a of disease occurrence involves considering an ongoing, causal relationship in an association. The causal open-ended set of scientific questions. Furthermore, nei- significance of an association is a matter of ther of these scientific contexts for defining causation is well suited to examining the effects of human behav - judgment which goes beyond any statement of statistical probability. To judge or evaluate the iors and exposures on subsequent disease occurrence causal significance of the association between in populations. the attribute or agent and the disease, or effect The early scientific reviews that examined whether smoking causes human disease used the common, gen- upon health, a number of criteria must be uti - erally understood meaning of the term “cause”: that the lized, no one of which is an all-sufficient basis disease occurs as a result of exposure to the agent. This for judgment. meaning was expressed in the 1964 report of the Surgeon cause General as follows: “The word Included in the evidence base for the 1964 report is the one in general usage in connection with matters considered in this study, were observational data from epidemiologic studies of human populations. During the 1940s and 1950s, and it is capable of conveying the notion of a significant, - effectual, relationship between an agent and an associated epidemiology was rapidly developing as a scientific disci disorder or disease in the host” (USDHEW 1964, p. 21). pline, but the observational, as opposed to experimental, 51 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose

81 Surgeon General’s Report nature of epidemiologic approaches led some scientists be used to define the causation of human disease: (1) the to question whether such approaches could be used to consistency of the association (replication of findings determine causation scientifically. Others confused epi - across different studies and populations), (2) the strength of the association (magnitude of the increased risk associ - demiologic analyses with the statistical methods used to - ated with exposure), (3) the specificity of the association describe the data (Shimkin 1979). Cigarette manufactur ers and their spokespersons capitalized on this confusion (presence of a unique exposure-disease association), (4) by claiming that only experimental approaches could lead the temporal relationship of the association (exposure to evidence establishing causation: the evidence used by - comes before effect), and (5) the coherence of the associa - tion (support for the association from other lines of evi public health authorities to conclude that smoking caused dence) (USDHEW 1964). lung cancer was only “statistical” and therefore not scien- tific (Brandt 2007; Proctor 2011). These criteria were included in the widely recog- nized criteria for interpreting epidemiologic evidence in Given the ethical impossibility of conducting human public health presented by Sir Austin Bradford Hill in experiments to establish causation and recognizing the 1965 (Hill 1965). The Bradford Hill criteria added four validity of epidemiologic methods, the various groups (before the Advisory Committee’s report) that examined additional criteria, most notably the presence of a biologic gradient (dose-response relationship) in the evidence. the question of whether cigarette smoking caused lung cancer had relied heavily on epidemiologic studies as The other three included plausibility (subsumed under coherence in the Surgeon General’s criteria), experiment, a key part of the evidence base establishing causation. Each review described how the epidemiologic data were and analogy. examined and considered. The reviews acknowledged that Detailed discussions of these criteria, how they epidemiologic studies lacked the methodologic reassur - evolved, and how they are applied in reviewing epi - demiologic evidence are presented in the 1964 report ance and needed careful attention to identify potential methodologic flaws, various biases, and both measured (USDHEW 1964) and the 2004 report (USDHHS 2004); and unrecognized confounding (e.g., lifestyle differences that discussion will not be repeated here. Rather, the between never smokers and smokers) that might have public health significance of formally expressed criteria for the use of epidemiologic evidence in defining causal resulted in the demonstrated association. Each of these - reports explained how these factors were considered in ity is the focus of the present discussion. Historically, the articulation of these criteria marked a turning point in the assessing the evidence, but the Advisory Committee went further and defined the criteria by which epidemiologic utilization and acceptance of epidemiologic evidence. It evidence could be examined and synthesized to reach a - laid the foundation for the current widespread use of epi causal judgment. - demiologic evidence to define disease causation and iden The Committee’s process for using epidemiologic tify methods for disease prevention and education of the public. These criteria, and their use by the Advisory Com- data in assessing causation included multiple steps. The mittee in reaching a judgment that smoking caused lung process involved: (1) establishing that cigarette smoking was associated with lung cancer; (2) examining whether cancer in men, established an approach that remains in use for causal inference based around epidemiological and the association could be explained by other factors such as other evidence. methodologic flaws, bias, or confounding; (3) examining whether there were plausible alternative explanations for the observed association; (4) considering the main points of criticisms raised about the association and its poten- Evolution of the Application of the tial causal nature; and (5) ensuring all of the lines of evi- Criteria for Disease Causation in dence were generally consistent with a causal hypothesis (USDHEW 1964). A similar careful and extensive process Subsequent Reports for considering evidence of causality had been imple- As the evidence on smoking as a cause of disease mented earlier by Cornfield and colleagues (1959) in their expanded to include numerous disorders or problems review of smoking and lung cancer: their considerations (various cancers, multiple manifestations of atheroscle- provided guidance for the methodologic approach adopted - rotic vascular disease, chronic obstructive pulmonary dis by the Advisory Committee in 1964. Subsequent reports ease [COPD], complications of pregnancy, and a myriad of of the Surgeon General have used the same approach other diseases and conditions [USDHHS 2004]), a variety for examining questions of causality for smoking and of terms were used to describe the established causal asso - specific diseases. ciations, including “cause,” “causal factor,” “risk factor,” In its report, the Advisory Committee formally pre- “contributing factor,” and “causal association.” Some of sented a set of criteria by which epidemiologic data could 52 Chapter 3

82 —50 Years of Progress The Health Consequences of Smoking made uniform to ensure clarity across the divergent dis- these descriptor choices were stylistic, reflecting the pref- erences of authors and editors; others reflected differences ease processes, as illustrated by the following statement - in how causal associations were described for different dis from the report: - ease processes, notably the use of risk factor in the litera ture on cardiovascular disease, where there are multiple The first step in introducing this revised approach causal factors. However, some uses of these terms were is to outline the language that will be used for intended to convey different levels of certainty about the summary conclusions regarding causality, which strength of the evidence establishing causation. follows hierarchical language used by Institute of Medicine committees (Institute of Medicine This use of multiple terms led to some ambiguity 1999) to couch causal conclusions, and by IARC and confusion as to what was actually being said. Eventu- [International Agency for Research on Cancer] ally, terms modifying the descriptors of causality were also introduced. These terms described the impact of smoking to classify carcinogenic substances (IARC 1986). on the population in relation to either other causes of dis - These entities use a four-level hierarchy for clas- ease or the contribution of smoking for a specific disease. sifying the strength of causal inferences based on suf- - For example, the 1989 Surgeon General’s report on smok available evidence as follows: (a) Evidence is to infer a causal relationship; (b) Evidence ficient ing and health stated that “Smoking remains the single suggestive but not sufficient to infer a causal is most important preventable cause of death in our soci- to infer relationship; (c) Evidence is ety” (USDHHS 1989, p. 11). This modifier was intended inadequate to describe the magnitude of the effect of smoking on the the presence or absence of a causal relationship population in contrast to other causes of premature death. (which encompasses evidence that is sparse, of Similarly, the relationship of cigarette smoking and lung poor quality, or conflicting); and (d) Evidence is suggestive of no causal relationship (USDHHS cancer was described as “Cigarette smoking is the major 2004, pp. 17–18). cause of lung cancer in the United States” (USDHHS 1982, p. 5), which qualitatively characterized the fraction of lung - The evidence on disease causation for each spe cancer deaths in the population caused by smoking. This cific disease is synthesized, and a judgment on causation mixing of terms, which quantified the population disease burden with terms describing the strength of the evidence is made and expressed using the standardized language presented above. This format clearly defines both the evi- establishing disease causation, had the potential to create ambiguity about what was being concluded, particularly dence on which the judgment is based and the strength with which that conclusion can be expressed. As for the - when the modifier was used for some diseases but not oth public health impact of smoking-caused disease for the ers, in the same report. population and the fraction of the disease caused by smok Importantly, the 2004 Surgeon General’s report - on smoking and health (USDHHS 2004) standardized ing, both are presented in these reports under a separate heading named “Implications” following the Conclusions the forms in which judgments on disease causation and - statements about the population consequences of dis section. It is in that section that the population-level impact of smoking and the fraction of the disease caused eases caused by smoking were presented. For causation, the language, which defined the strength of the evidence by smoking are examined. establishing that smoking caused a specific disease, was Methods for Reviewing the Evidence and Developing Conclusions The reports of the Surgeon General have continued processes used for reviewing and presenting the evidence and for the development of the conclusions. The processes to play a role in defining the science that underlies efforts used for subsequent reports evolved from the process used in tobacco control by certifying the causation of various - in the 1964 report of the Advisory Committee. diseases and expressing the state of the science on the effec The 1964 report, at 387 pages, was substantively lon- tiveness of tobacco control interventions, approaches, and policies. The success of the series of reports reflects the ger than the independent reviews that had preceded it. As 53 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose

83 Surgeon General’s Report a consequence, it was able to offer a much more detailed combined estimate of the risk for calculating the associ- presentation of the evidence in the text rather than simply ated burden of disease and, potentially, for exploring why providing references to the individual studies in support results vary from study to study. The practice of presenting the relevant evidence of the conclusions. needed to support the conclusions also has helped to The depth of the evidence presentation in the Advi - ensure the validity of the conclusions as has the tiered sory Committee’s report in 1964 can be seen readily in approach and peer review process of the chapters. In the subsequent reports, and this comprehensive approach has - been one reason for the reports’ continuing credibility. An Surgeon General’s reports, the initial author of an indi editorial standard evolved that required the conclusions of vidual section is tasked with reviewing and assembling all individual sections of the report to be based on discussions of the relevant evidence available and presenting it in the of the literature presented in the text that were coupled text and related tables and figures with a level of detail with relevant study results presented in the text, tables, - sufficient to support the conclusions. Based on that pre and figures of those sections. This approach, of present sentation, the author then considers and discusses what - conclusions the evidence supports. This comprehen ing the totality of evidence in sufficient detail to allow the - sive review process helps reduce inaccuracies that may reader to evaluate it, contrasted with the general approach of the time for written reviews, which relied heavily on occur when authors synthesize the evidence and reach conclusions based on their recall of what the literature syntheses of evidence by authors with literature citations for the publications reviewed. In the Surgeon General’s shows, rather than on the evidence actually contained in that literature. reports, presentation of the critical findings from the relevant studies, coupled with discussion of the methods Passing the section on to the editors allows a differ - used to generate the evidence in the text of the report, has ent group of people to consider the evidence presented to evaluate the basis for the conclusions and to revise them, allowed readers to assess the validity of the conclusions directly rather than requiring them to conduct a time- - if appropriate. Similarly, as the chapters and reports pro ceed through the various review stages, the reviewers - consuming search of the cited publications. This trans can independently consider the evidence presented as parency has strengthened the reports’ findings in the face of the inevitable criticisms. they consider the accuracy, completeness, balance, tone, - and language of the conclusions. In providing their com In synthesizing the evidence on exposure to sec- ments, the reviewers can focus on the evidence presented, ondhand smoke and disease, meta-analysis has been used, consider whether the review of that evidence is complete, - both in the Surgeon General’s reports and in other evalu ations. Generally, the term “meta-analysis” refers to the and judge whether the conclusions are supported by the evidence. systematic analysis and quantitative summarization of the findings of multiple studies containing evidence to The intense criticisms of the reports by the cigarette address the same question (Greenland 1987; Egger and manufacturers and their representatives prior to the late Davey Smith 1997; Institute of Medicine 2011). In a meta- 1990s (see Chapter 2) helped to strengthen the process of developing conclusions for the reports. The anticipation of analysis, the data are the summary findings of the studies criticism motivated the development of conclusions that identified through a systematic review and not the data at - were firmly based on evidence without speculation. Before the individual level. Meta-analysis has been used to sum its elimination as a result of the 1998 Master Settlement marize the evidence on exposure to secondhand smoke, - Agreement, the Tobacco Institute (a representative of cig primarily because the associations are generally much arette manufacturers) conducted a well-funded and highly weaker than they are for active smoking. Meta-analysis visible public relations campaign to denigrate the quality was not used in the 1986 report, but it was applied to of the science in each Surgeon General’s report and ques - The Health Conse- multiple outcomes in the 2006 report, , and tion the validity of their conclusions (Kluger 1996; Brandt quences of Involuntary Exposure to Tobacco Smoke is used in this report. 2007; Proctor 2011). Based on the historical pattern of challenges to the Surgeon General’s reports (see Chapter - Although meta-analysis has proven useful for sum 14, “Current Status of Tobacco Control”) the authors, edi - - marizing the evidence and quantifying the risks of expo - sure to secondhand smoke as precisely as possible, the tors, and reviewers of the reports assumed that every con clusion might be challenged and, therefore, each had to findings of meta-analyses and, particularly, information be solidly and fully supported by sufficient evidence. The on whether an association found in the meta-analysis was statistically significant, have not figured directly in the result was that, as conclusions were drafted and reviewed, there was an intense focus on the quality and robustness causal inferences presented in the reports of the Surgeon of the evidence. Conclusions were structured to be unas- General. The results are most useful for providing a single, 54 Chapter 3

84 —50 Years of Progress The Health Consequences of Smoking disease causation, the effectiveness of efforts in tobacco sailably grounded in a foundation of evidence and the control, and the consequences of changes in public policy. language of the conclusions was “conservative” such In addition, it has provided solid support for evidence- that the strength of evidence was not overstated. As - based public policy decisions on tobacco issues, has iden the evidence foundation advanced, conclusions were tified the areas where scientific certainty exists as separate strengthened. from those areas where uncertainty remains, and has This effort to achieve scientific transparency by lay - been a principal reason for the enduring credibility of this ing out the evidence foundation for the conclusions has series of reports. defined with clarity the state of the scientific evidence on Process of Ensuring Consensus and Strength of the Peer Review In a series of governmental reports, such as those - the chapters were assembled into a draft report. That ver sion of the report was sent to a larger group of experts, of the Surgeon General which have both great visibility and a substantial impact on public policy, protections are broadly knowledgeable in smoking and health, who were needed to resist influences that could distort the process asked to comment on the balance, tone, and accuracy of of forming a consensus and affect the conclusions. the volume and its conclusions. The draft report was also submitted for review to those agencies within PHS that As a report is in development, a myriad of factors may come into play: political pressures; pressures from a were involved with tobacco issues. Revisions were made in response to these comments, and the volume was then variety of individuals and groups to have the conclusions conform to their preexisting policy positions; the recog submitted for formal clearance and release as the official - nition that some conclusions can influence decisions on position of PHS on the science of tobacco and health. As required by law, it was also transmitted to the U.S. Con - research funding; and even the well-intentioned belief of gress. This complex, multilayered peer review helped to authors of sections of the report that the final conclusions should substantiate positions they have adopted based ensure not only that the science in the volume was accu - rate but also that the positions expressed on the science on their own research. Without a process to insulate were the prevailing view of the scientific community at the report’s conclusions from such influences, the con - clusions might be perceived as based on the politics and the time and represented concurrence without being pressures of the moment rather than on a consensus of unduly influenced by any one individual or group. Beginning with the 1979 Surgeon General’s report scientific opinion. The National Clearinghouse for Smoking and Health on smoking and health and continuing to the present, an additional layer of insulation was added by selecting a prepared the initial series of reports (1967–1976) which set of editors for each volume who were drawn from the followed the 1964 Surgeon General’s report. The scientific academic and scientific communities and, when selected, and technical staff of the clearinghouse, a forerunner of - were not employees of the federal government. These edi the current Centers for Disease Control and Prevention’s Office on Smoking and Health, was responsible for both tors have been tasked with ensuring the accuracy of the scientific content of the reports and providing additional The drafting and editing the volume. The 1971 report, Health Consequences of Smoking , was a comprehensive independent oversight for the process of incorporating reviewers’ comments. These independent editors, rather review of all of the available evidence, but the other reports than the authors, have been responsible for making the in the 1967–1976 period were intended to review the evi - final decisions on incorporating reviewers’ comments into dence on the relationship of smoking to cancer, cardiovas - the text, thereby creating a layer of objectivity regarding cular disease, and COPD that had been published since the previous report, with additional chapters focusing in more reviewers’ comments as they are considered and pre - depth on specific topics. The “in-house” preparation of the venting the views of any single author from controlling - volume was counterbalanced by a multilevel review pro the conclusions. cess. Each draft chapter was reviewed by experts, external The evolution of this production process demon - strates that it is possible for a governmental review of a to the clearinghouse, from the academic community and select PHS agencies who were asked to evaluate the accu- scientific topic of high societal interest and relevance to racy and completeness of the chapter. After the review - - be conducted in a way that ensures independence and sci ers’ comments were incorporated into the draft chapters, entific accuracy for the resulting scientific conclusions. 55 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose

85 Surgeon General’s Report Separation of Scientific Conclusions and the Formation of Policy The findings of the reports of the Surgeons General concluded scientifically from the evidence and to lay out have been the basis for a wide-ranging set of policy deci - the implications of those conclusions for the population. The separation of scientific conclusions from policy sions and consequently some may consider the reports as recommendations, initially adopted because policy deci- offering policy recommendations. The overall intent of the reports, however, has been to provide a clear evidence sions and implementation occurred at organizational foundation for scientific judgments on the diseases caused levels well above that of the National Clearinghouse for by smoking, the factors influencing smoking initiation Smoking and Health, has helped to ensure the ongoing credibility of this series of reports. Public policy decisions and cessation, the effectiveness of smoking and tobacco - control interventions, and the results of tobacco control are, and often must be, made before the evidence support - ing them is complete. These reports have been the bench - programs and changes in public policy. The characteriza tion of the state of the science on these issues remains mark on the status of the evidence for decision-making. the mission of the reports of the Surgeon General and is By preserving its exclusive focus on the scientific foundation and avoiding the inclusion of policy recom- their principal enduring value. Although it is hoped that these scientific judgments will be used in the formation mendations by the scientists involved with the report, the of public policy, and the reports have often examined the reports of the Surgeon General have preserved their cred - evidence on the effects of public policy decisions, the con - ibility and somewhat insulated the report development tent of the reports has been limited to the state of the sci process from the need for scientific certainty among those - ence on these issues. The reports have avoided defining responsible for forming public policy. Correspondingly, or recommending specific public policies, leaving those the recognized independence of the reports’ conclusions decisions to the entities responsible for policy formation, has resulted in a solid and enduring foundation that sup - - including the Secretary of HHS and the various compo - ports those who are tasked with defining and implement ing public policy. nents of that department. The conclusions of the report have been intentionally framed to state what could be 56 Chapter 3

86 —50 Years of Progress The Health Consequences of Smoking References Blackburn H, Labarthe D. Stories from the evolution of Office of the Assistant Secretary for Health, Office on Smoking and Health, 1980. - guidelines for causal inference in epidemiologic asso American Journal of Epidemiol ciations: 1953–1965. - The U.S. Department of Health and Human Services. 2012;176(12):1071–7. ogy Health Consequences of Smoking: The Changing Ciga- Brandt AM. The Cigarette Century: The Rise, Fall, and rette. A Report of the Surgeon General . Rockville (MD): - Deadly Persistence of the Product That Defined Amer U.S. Department of Health and Human Services, Public ica . New York: Basic Books, 2007. Health Service, Office on Smoking and Health, 1981. Cornfield J, Haenszel W, Hammond EC, Lilienfeld AM, DHHS Publication No. (PHS) 81-50156. Shimkin MB, Wynder EL. Smoking and lung can The - U.S. Department of Health and Human Services. Health Consequences of Smoking: Cancer. A Report of cer: recent evidence and a discussion of some ques- Journal of the National Cancer Institute . Rockville (MD): U.S. Department the Surgeon General tions. 1959; of Health and Human Services, Public Health Service, 22(1):173–203. - Egger M, Smith GD. Meta-analysis. Potentials and prom Office on Smoking and Health, 1982. DHHS Publica- ise. British Medical Journal 1997;315(7119):1371–4. tion No. (PHS) 82-50179. Glass TA, Goodman SN, Hernan MA, Samet JM. Causal The U.S. Department of Health and Human Services. inference in public health. Annual Review of Public - Health Consequences of Smoking: Cardiovascular Dis 2013;34:61–75. Health ease. A Report of the Surgeon General . Rockville (MD): U.S. Department of Health and Human Services, Public Greenland S. Quantitative methods in the review of epi - Health Service, Office on Smoking and Health, 1983. Epidemiologic Reviews 1987; demiologic literature. 9:1–30. DHHS Publication No. (PHS) 84-50204. The - Hill AB. The environment and disease: association or cau U.S. Department of Health and Human Services. sation? Proceedings of the Royal Society of Medicine Health Consequences of Smoking: Chronic Obstruc - 1965;58:295–300. tive Lung Disease. A Report of the Surgeon General . Institute of Medicine. Rockville (MD): U.S. Department of Health and Human Veterans and Agent Orange: Update 1998 . Washington: National Academy Press, 1999. Services, Public Health Service, Office on Smoking and Finding What Works in Health Health, 1984. DHHS Publication No. (PHS) 84-50205. Institute of Medicine. . Washington: U.S. Department of Health and Human Services. Care: Standards for Systematic Reviews The Health Consequences of Smoking: Cancer and Chronic The National Academies Press, 2011. Lung Disease in the Workplace. A Report of the Sur IARC - International Agency for Research on Cancer. Monographs on the Evaluation of Carcinogenic Risk geon General . Rockville (MD): U.S. Department of Vol. 38. Health and Human Services, Public Health Service, of Chemicals to Humans: Tobacco Smoking. Lyon (France): International Agency for Research on Office on Smoking and Health, 1985. DHHS Publica- Cancer, 1986. tion No. (PHS) 85-50207. Ashes to Ashes: America’s Hundred-Year Ciga- The U.S. Department of Health and Human Services. Kluger R. rette War, the Public Health, and the Unabashed Tri Health Consequences of Involuntary Smoking: A - Report of the Surgeon General umph of Philip Morris . New York: Alfred A. Knopf, 1996. . Rockville (MD): U.S. Proctor RN. Golden Holocaust: Origins of the Cigarette Department of Health and Human Services, Public Catastrophe and the Case for Abolition Health Service, Centers for Disease Control, Center . Berkeley (CA): University of California Press, 2012. for Health Promotion and Education, Office on Smok - . 2nd ed. ing and Health, 1986a. DHHS Publication No. (CDC) Modern Epidemiology Rothman KJ, Greenland S. Philadelphia: Lippincott-Raven, 1998. 87-8398. U.S. Department of Health and Human Services. The Shimkin MB. In the middle: 1954–63—historical note. Health Consequences of Using Smokeless Tobacco: Journal of the National Cancer Institute 1979;62(5): - Report of the Advisory Committee to the Surgeon Gen 1295–317. eral of the Public Health Service . Bethesda (MD): U.S. The U.S. Department of Health and Human Services. Department of Health and Human Services, Public Health Consequences of Smoking for Women. A Report of the Surgeon General . Washington: U.S. Department Health Service, 1986b. NIH Publication No. 86-2874. of Health and Human Services, Public Health Service, 57 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose

87 Surgeon General’s Report U.S. Department of Health and Human Services. U.S. Department of Health and Human Services. The Reducing - Tobacco Use. A report of the Surgeon General Health Consequences of Smoking: Nicotine Addic . Atlanta . Atlanta (GA): (GA): U.S. Department of Health and Human Services, tion. A Report of the Surgeon General U.S. Department of Health and Human Services, Public Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Pro Health Service, Centers for Disease Control, National - - Center for Chronic Disease Prevention and Health Pro motion, Office on Smoking and Health, 2000. U.S. Department of Health and Human Services. Women motion, Office on Smoking and Health, 1988. DHHS and Smoking. A Report of the Surgeon General . Rock- Publication No. (CDC) 88-8406. ville (MD): U.S. Department of Health and Human Reduc- U.S. Department of Health and Human Services. Services, Public Health Service, Office of the Surgeon ing the Health Consequences of Smoking: 25 Years of Progress. A Report of the Surgeon General General, 2001. . Rockville (MD): U.S. Department of Health and Human Services, The U.S. Department of Health and Human Services. Public Health Service, Centers for Disease Control, - Health Consequences of Smoking: A Report of the Sur geon General . Atlanta (GA): U.S. Department of Health National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, and Human Services, Centers for Disease Control and - Prevention, National Center for Chronic Disease Pre 1989. DHHS Publication No. (CDC) 89-8411. The vention and Health Promotion, Office on Smoking and U.S. Department of Health and Human Services. Health, 2004. Health Benefits of Smoking Cessation: A Report of the The U.S. Department of Health and Human Services. . Rockville (MD): U.S. Department of Surgeon General Health and Human Services, Centers for Disease Con- Health Consequences of Involuntary Exposure to trol, National Center for Chronic Disease Prevention Tobacco Smoke: A Report of the Surgeon General . and Health Promotion, Office on Smoking and Health, Atlanta (GA): U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 1990. DHHS Publication No. (CDC) 90-8416. Smok- Coordinating Center for Health Promotion, National U.S. Department of Health and Human Services. - Center for Chronic Disease Prevention and Health Pro ing and Health in the Americas. A 1992 Report of the motion, Office on Smoking and Health, 2006. Surgeon General, in Collaboration with the Pan Amer - U.S. Department of Health and Human Services. . Atlanta (GA): U.S. Depart ican Health Organization How - ment of Health and Human Services, Public Health Tobacco Smoke Causes Disease—The Biology and Service, Centers for Disease Control, National Center Behavioral Basis for Smoking-Attributable Disease: for Chronic Disease Prevention and Health Promotion, A Report of the Surgeon General . Atlanta (GA): U.S. Office on Smoking and Health, 1992. DHHS Publica- Department of Health and Human Services, Centers tion No. (CDC) 92–8419. for Disease Control and Prevention, National Center Prevent- for Chronic Disease Prevention and Health Promotion, U.S. Department of Health and Human Services. ing Tobacco Use Among Young People: A Report of the Office on Smoking and Health, 2010. Pre- U.S. Department of Health and Human Services. . Atlanta (GA): U.S. Department of Surgeon General Health and Human Services, Public Health Service, venting Tobacco Use Among Youth and Young Adults: A Report of the Surgeon General Centers for Disease Control and Prevention, National . Atlanta (GA): U.S. - Department of Health and Human Services, Centers Center for Chronic Disease Prevention and Health Pro motion, Office on Smoking and Health, 1994. for Disease Control and Prevention, National Center Tobacco U.S. Department of Health and Human Services. for Chronic Disease Prevention and Health Promotion, Use Among U.S. Racial/Ethnic Minority Groups-Afri - Office on Smoking and Health, 2012. can Americans, American Indians and Alaska Natives, U.S. Department of Health, Education, and Welfare. Smok- Asian Americans and Pacific Islanders, and Hispanics. ing and Health: Report of the Advisory Committee to . Atlanta (GA): U.S. A Report of the Surgeon General the Surgeon General of the Public Health Service . Department of Health and Human Services, Centers Washington: U.S. Department of Health, Education, for Disease Control and Prevention, National Center and Welfare, Public Health Service, Center for Disease for Chronic Disease Prevention and Health Promotion, Control, 1964. PHS Publication No. 1103. Office on Smoking and Health, 1998. 58 Chapter 3

88 —50 Years of Progress The Health Consequences of Smoking Health, Education, and Welfare, Public Health Service, The U.S. Department of Health, Education, and Welfare. Health Consequences of Smoking. A Public Health Health Services and Mental Health Administration, 1973. DHEW Publication No. (HSM) 73-8704. . Washington: U.S. Department of Service Review: 1967 The U.S. Department of Health, Education, and Welfare. Health, Education, and Welfare, Public Health Service, - Health Consequences of Smoking. A Report of the Sur Health Services and Mental Health Administration, geon General, 1974 1967. PHS Publication No. 1696. . Washington: U.S. Department of Health, Education, and Welfare, Public Health Service, The U.S. Department of Health, Education, and Welfare. Health Consequences of Smoking. 1968 Supplement Center for Disease Control, 1974. DHEW Publication to the 1967 Public Health Service Review No. (HSM) 74-8704. . Washington: The U.S. Department of Health, Education, and Welfare. U.S. Department of Health, Education, and Welfare, - Public Health Service, 1968. DHEW Publication No. Health Consequences of Smoking. A Report of the Sur . Washington: U.S. Department of geon General, 1975 1696 (Supplement). The U.S. Department of Health, Education, and Welfare. Health, Education, and Welfare, Public Health Service, Center for Disease Control, 1975. DHEW Publication Health Consequences of Smoking. 1969 Supplement to the 1967 Public Health Service Review . Washington: No. (HSM) 76-8704. U.S. Department of Health, Education, and Welfare, U.S. Department of Health, Education, and Welfare. The Public Health Service, 1969. DHEW Publication No. Health Consequences of Smoking. Selected Chapters 1696-2. From 1971 Through 1975 Reports . Atlanta (GA): U.S. The U.S. Department of Health, Education, and Welfare. Department of Health, Education, and Welfare, Cen - Health Consequences of Smoking. A Report of the Sur - ters for Disease Control, 1976. DHEW Publication No. geon General: 1971 . Washington: U.S. Department of 78-8357. Health, Education, and Welfare, Public Health Service, U.S. Department of Health, Education, and Welfare. Health Services and Mental Health Administration, Smoking and Health. A Report of the Surgeon Gen - 1971. DHEW Publication No. (HSM) 71-7513. . Washington: U.S. Department of Health, Educa eral - U.S. Department of Health, Education, and Welfare. tion, and Welfare, Office of the Assistant Secretary for The Health Consequences of Smoking. A Report of the Sur - Health, Office on Smoking and Health, 1979a. DHEW . Washington: U.S. Department of geon General: 1972 Publication No. (PHS) 79-50066. Health, Education, and Welfare, Public Health Service, The U.S. Department of Health, Education, and Welfare. Health Services and Mental Health Administration, - Health Consequences of Smoking. A Report of the Sur - . Washington: U.S. Depart geon General, 1977–1978 1972. DHEW Publication No. (HSM) 72-7516. U.S. Department of Health, Education, and Welfare. ment of Health, Education, and Welfare, Office of the The Health Consequences of Smoking. A Report of the Sur Assistant Secretary for Health, Office on Smoking and - geon General, 1973 . Washington: U.S. Department of Health, 1979b. DHEW Publication No. (CDC) 79-50065. 59 Producing the Surgeon General’s Report From 1964–2014: Process and Purpose

89

90 Chapter 4 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014 Introduction 63 Evolution of Conclusions on Cigarette Smoking and Exposure to Secondhand Smoke as a Cause of Disease 67 Active Smoking 67 Exposure to Secondhand Smoke 68 Summary 69 104 References 61

91

92 The Health Consequences of Smoking —50 Years of Progress Introduction existence. Chronic obstructive pulmonary disease (COPD) The 50-year span beginning in 1964 and ending in - was recognized, but it was referred to as “chronic bronchi - 2014 covers an era of remarkable advances in the under tis” or “emphysema,” and the prevalence of what we now - standing of disease etiology and opportunities for the pre vention, diagnosis, and treatment of disease. There have call COPD was far below the present level (Petty 2006; Kim and Criner 2013). Antibiotics were available for most bac - also been striking changes seen in the incidence of dis - - terial infections, but not all infections could be cured with ease, in mortality rates, and substantial gains in life expec tancy. For example, in 1964 cancer was widely regarded these drugs; antiviral agents, other than vaccines, were lacking altogether. as incurable and few causal agents had been identified, although tobacco smoke was already of concern because During the last half-century, major changes in it had been identified as carcinogenic (Mukherjee 2010). - disease occurrence have taken place that provide a criti - Physicians and public health officials lacked today’s pre cal context for the tobacco epidemic (Figure 4.1). The infectious diseases, particularly tuberculosis, declined ventive strategies for coronary heart disease and widely as leading contributors to mortality to be replaced by used drugs, such as statins, had not yet been developed. the noncommunicable diseases: cardiovascular dis Coronary care units for managing acute myocardial - infarctions and heart rhythm disturbances were not in eases, COPD, and cancer. Studies on the causes of these Mortality rates for major diseases in the United States, 1900–2005 Figure 4.1 300 Infectious disease (Men and women) CVD (Men and women) 250 200 150 All cancers (Men) All cancers (Women) 100 Age-adjusted rate per 100,000 population Lung and bronchus cancer (Men) COPD (Men) 50 Stomach cancer (Men) Lung and bronchus Stomach cancer (Women) cancer (Women) COPD (Women) J 0 1910 1915 1905 1900 1920 1960 1955 2000 1985 1950 1935 1945 1965 1995 2005 1940 1930 1990 1925 1975 1970 1980 Years Source: Infectious disease and CVD rates from Cutler et al. 2006. Age-adjusted rates for stomach, lung, and bronchus cancer from American Cancer Society 2009. Age-standardized rate for all cancers from World Health Organization Mortality Database 2012. Age-adjusted rates for COPD from National Center for Health Statistics 2012. COPD = chronic obstructive pulmonary disease; CVD = cardiovascular disease. Note: Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014 63

93 Surgeon General’s Report noncommunicable diseases were motivated by their ris - and in subsequent reports as follow-up of Health Service, participants continued and risks were tracked over time. ing frequency. Observational epidemiologic studies had a critical role in the search for causes, while complemen - During the 50 years since the first Surgeon Gen - - eral’s report on smoking and health in 1964 to this anni tary laboratory research expanded the understanding of versary report, the observational evidence on the causes the biological mechanisms by which risk factors caused of noncommunicable diseases has continued to advance these diseases. But even before 1964, advances had been as numerous case-control and cohort studies were carried made in characterizing the etiology of noncommunicable out and our understanding of the mechanistic processes diseases. These advances relied on case-control and land- - mark cohort studies started in the late 1940s and 1950s, leading to these diseases was greatly enhanced. Numer ous risk factors were identified that have been classified by such as the Framingham Heart Study (which identified the Global Burden of Disease project into broad groups, multiple risk factors for noncommunicable diseases, and explored blood pressure, lipids, and smoking in relation - including air pollution, tobacco smoking including expo - sure to secondhand smoke, alcohol and drug use, dietary ship to risk for incident coronary heart disease) (Kannel risk factors and physical inactivity, physiological risk fac et al. 1961), the British Doctors Study in the United King - - dom (Doll and Hill 1954), and studies carried out by the tors, and occupational risk factors (Lim et al. 2012). Many of these risk factors, such as physical inactivity, unhealthy American Cancer Society in the United States (Hammond diet, and smoking, could be avoided, making primary and Horn 1954) linking cigarette smoking to multiple diseases. Findings from these studies figured prominently prevention possible. Pharmacological therapies provided control for some risk factors, such as treatment of lipid Smoking and Health: Report of the in the 1964 report, Advisory Committee of the Surgeon General of the Public abnormalities with statins and other medications. a Figure 4.2 Age-adjusted mortality rates for all causes , United States, selected years, 1900–2010 Hoyert et al. 2001; National Center for Health Statistics 2013. Source: a All causes of deaths combined. 64 Chapter 4

94 —50 Years of Progress The Health Consequences of Smoking a Figure 4.3 Mortality rates from selected cancers among men in the United States, 1930–2008 100 80 Lung and bronchus 60 40 Prostate Rate per 100,000 male population 20 Colon and rectum Pancreas Stomach Liver Leukemia 0 2008 1970 2000 1990 1940 1950 1975 1980 2005 1935 1960 1945 1995 1985 1930 1955 1965 Years Source: Surveillance, Epidemiology, and End Results Program 2013. Due to changes in coding, numerator information has changed over time. Rates for Note: International Classification of Diseases cancer of the liver, lung and bronchus, and colon and rectum are affected by these coding changes. a Per 100,000, age-adjusted to the 2000 U.S. standard population. During the 50-year period reviewed in this report, Another important advance over the last several decades has been the incorporation of genetics into there have been substantial changes in disease patterns in the United States. Figure 4.1 shows the rates for mortality research on the etiology of noncommunicable diseases, for selected major diseases across the twentieth century, especially in the use of genetics to identify those men and women who are particularly susceptible to certain and Figure 4.2 shows the rates for all-cause mortality. Although the time spans covered differ for the various extrinsic exposures, such as cigarette smoking. For the diseases caused by smoking, emphasis has been placed causes of death because of changes in coding used in the and in the avail- International Classification of Diseases on understanding why some people who are exposed to ability of data, major patterns are evident. These include tobacco smoke develop disease while others do not. Also the substantial decline in all-cause mortality (Figure 4.2) in the last few decades, the approaches used to explore the and the sharp drop in infectious disease mortality (Figure - genetic basis of disease have evolved from family and link age studies to genome-wide association studies (GWAS) 4.1), both long antedating the general availability of mod - ern antibiotics at mid-century. The rising mortality from (Wellcome Trust Case Control Consortium 2007). The lung cancer and cardiovascular disease that triggered GWAS approach involves comparing the distribution of numerous epidemiologic inquiries is also evident in Fig- markers (single nucleotide polymorphisms) across the ure 4.1. In the later decades of the time period, rates for genome between (a) people affected by the disease of interest and (b) a control population. To date, however, coronary heart disease mortality declined sharply, while even though hundreds of thousands of markers across lung cancer mortality in men reached a plateau and then began to decline around 1990. In contrast, lung cancer the genome have been examined, few promising associa - - mortality in women rose, reaching a plateau by the cen tions have been found (Visscher et al. 2012), but work is in tury’s end. Mortality from COPD, variably described across progress to further explore the GWAS-identified markers in greater depth (U.S. Department of Health and Human the century with labels including chronic bronchitis and Services [USDHHS] 2010). Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014 65

95 Surgeon General’s Report a Mortality rates from selected cancers among women in the United States, 1930–2008 Figure 4.4 100 80 60 40 Lung and bronchus b Uterus Pancreas 20 Rate per 100,000 female population Colon and rectum Breast Ovary Stomach 0 1970 2000 1990 1940 1950 1975 1980 2005 2008 1935 1960 1945 1995 1985 1930 1955 1965 Years Source: Surveillance, Epidemiology, and End Results Program 2013. Due to changes in Note: International Classification of Diseases coding, numerator information has changed over time. Rates for cancer of the liver, lung and bronchus, and colon and rectum are affected by these coding changes. a Per 100,000, age-adjusted to the 2000 U.S. standard population. b Uterus refers to uterine cervix and uterine corpus combined. emphysema, has risen progressively, even as death rates Collaborators 2013). Tobacco control measures, driven by the emerging findings on the health consequences of for other major diseases, such as cardiovascular disease and lung cancer caused by smoking have declined (Petty tobacco smoking, have been a key determinant of changes in these rates. 2006; Kim and Criner 2013). - Figures 4.3 and 4.4, spanning 1930–2008, pro This chapter reviews the evolution of the conclu- vide further detail on mortality rates for cancer in men sions in the Surgeon General’s reports with regard to the health consequences of smoking. The chapters following and women. For both genders, the rise of lung cancer this one review the evidence for diseases and other adverse to become the leading cause of cancer death is evident. Stomach cancer, once the leading cause of cancer death in effects for which the evidence was previously found to be suggestive, including macular degeneration, colorectal men and second among women in 1930, dropped so far as cancer, breast cancer, prostate cancer, and male sexual dys - to eventually rank last among the seven cancers portrayed in Figures 4.3 and 4.4. Also during the 1930−2008 period, function. Additionally, the chapters cover several health outcomes that have not been comprehensively addressed the uterine cancer mortality rate for women declined in previous Surgeon General’s reports, including general steeply. In addition, among women the mortality rate for lung cancer surpassed that for breast cancer in the 1980s effects on the immune system and the development of and continued to rise to a plateau as breast cancer mor several diseases in which the immune system plays a key - tality declined. The mortality rate for pancreatic cancer role, such as tuberculosis, diabetes, rheumatoid arthritis, rose slowly between 1930–2008 for both men and women. and systemic lupus erythematosus. The reviews extend to active smoking and exposure to secondhand smoke, as Although many factors have driven these changing pat- terns of disease, the patterns reflect, in part, the rise and appropriate. New reviews in Chapter 8 cover the relation- fall of the prevalence of cigarette smoking across the ship between exposure to secondhand smoke and stroke twentieth century (USDHHS 2004; U.S. Burden of Disease and the potential that smokefree policies will reduce 66 Chapter 4

96 The Health Consequences of Smoking —50 Years of Progress the incidence of cardiovascular events. For lung cancer, the first time, including the impact of smoking on treat- COPD, and cardiovascular diseases—well-established ment outcomes for cancer sites that have not been caus - major consequences of cigarette smoking—perspectives ally related to smoking. Chapter 11 addresses general morbidity and all-cause mortality, and updated estimates - are provided on the most critical issues relative to smok ing in the etiology of these diseases. Several chapters of the burden of smoking-attributable mortality and mor - bidity and of the direct and indirect costs of smoking are address general and nonspecific consequences of smoking provided in Chapter 12. tobacco. The clinically significant topic of smoking and outcomes following the diagnosis of cancer is covered for Evolution of Conclusions on Cigarette Smoking and Exposure to Secondhand Smoke as a Cause of Disease 1964 report. The language ranges widely in describing the During the past 50 years, both the number and findings, from the clear conclusion that smoking causes - strength of the conclusions on active smoking and expo lung cancer in men to the characterizations of the uncer - sure to secondhand smoke as a cause of disease and other tainty and limitations of the evidence for some diseases. In adverse health effects have increased markedly, moving most cases, the conclusions provide summary descriptions from the two specific causal conclusions on lung cancer of the state of the evidence as well. The lack of knowledge in males and on chronic bronchitis that were drawn in the 1964 report to numerous other conclusions that span - of the mechanism(s) underlying the association of smok ing with birth weight is mentioned. most organs and now include exposure to secondhand - In Table 4.1, which deals with active smoking smoke. Tables 4.1–4.5 address the evolution of the con and cancer, there has been consistency over time in the clusions on active smoking, listing the report in which a nomenclature so that interpretation of the changes in particular health consequence was first mentioned; the conclusions is not complicated by shifting terminology. strongest conclusion(s) reached before the 2004 report, With the exception of stomach cancer, causal conclu- The Health Consequences of Smoking (in which the clas- sification of the strength of evidence was standardized); sions were reached within the next two decades for can - cer sites other than the lung that were mentioned in the the conclusion(s) of the 2004 report; and any subsequent conclusions. The changes in the conclusions over time 1964 report (i.e., oral cancer, laryngeal cancer, esophageal cancer, stomach cancer, and cancer of the urinary blad - are characterized in this fashion because of the variable der) (U.S. Department of Health, Education, and Welfare terminology used before the 2004 report (USDHHS 2004). Tables 4.6–4.10 provide a similar listing for exposure to Health Consequences [USDHEW] 1964). The 1982 report, , which focused on cancer, identified of Smoking – Cancer secondhand smoke. smoking as a contributory factor for pancreatic cancer and - Although these conclusions relate primarily to spe cific diseases and other adverse health effects, the Surgeon kidney cancer (USDHHS 1982). The list of cancers causally linked to active smoking lengthened with the 2004 report, General’s reports have also tracked the evolution of the which added cervical cancer and acute myeloid leukemia understanding of the pathogenesis and adverse health effects of these diseases and conditions. This deepening (USDHHS 2004). That report found the evidence on cau- sation to be suggestive for breast cancer, colorectal cancer, understanding has supported reaching stronger conclu- and liver cancer. For prostate cancer, the evidence overall How Tobacco Smoke sions on causation. The 2010 report, Causes Disease: The Biology and Behavioral Basis for was not suggestive of a causal relationship. - For cardiovascular diseases (Table 4.2), the trends , provides conclusions spe Smoking-Attributable Disease in the conclusions reflect the advancing understand- cific to this topic (USDHHS 2010). ing of the pathogenesis of these diseases and their com- mon mechanistic basis (see Chapter 8 “Cardiovascular Diseases”). The 1964 report commented on the higher Active Cigarette Smoking death rates from coronary artery disease among smokers compared with nonsmokers, but it expressed uncertainty Table 4.11 provides the conclusions formally adopted with regard to the causal significance of the association by the Advisory Committee to the Surgeon General in the 67 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

97 Surgeon General’s Report (USDHEW 1964). The conclusions on cardiovascular dis- and male sexual dysfunction has now been reached in this eases strengthened throughout the next several decades. 2014 report. Smoking and Health , offered a causal The 1979 report, Numerous other diseases and adverse consequences conclusion on coronary heart disease, but one that was of smoking have been addressed in the reports of the Surgeon General (Table 4.5). These have included dental introduced by the phrase “In summary, for the purposes diseases, cataract and macular degeneration, peptic ulcer of preventive medicine ...” (USDHEW 1979, p. 1-15). This apparently cautious phrasing may have been reflective disease, fractures and osteoporosis, and diabetes. Nonspe - cific consequences of smoking have also been considered. of the preventive implications of the causal conclusion, however, and not an indication that there was some doubt All-cause mortality was covered in the 1964 report, but about the statement. Later, the 2004 report found the a specific conclusion was not offered. Several subsequent evidence to be sufficient to infer causation for abdominal - reports identified smoking as the leading cause of avoid aortic aneurysm, atherosclerosis and peripheral vascular able premature mortality (Table 4.12). The 2004 report - disease, cerebrovascular disease, and coronary heart dis- assembled a wide range of evidence on nonspecific con sequences of smoking, such as absenteeism and postop - ease (Table 4.2) (USDHHS 2004). erative complications, with the report concluding that The conclusions on respiratory diseases over the smoking caused “diminished health status” (Table 4.5), years (Table 4.3) have addressed COPD, variably desig- based on a review of a wide range of evidence (USDHHS nated, as well as the respiratory symptoms caused by smoking and its reduction of lung function which, if sus 2004). The report’s conclusion stated that diminished - tained, leads to COPD. The 1964 report concluded that health status may manifest as “... increased absenteeism “Cigarette smoking is the most important of the causes of from work and increased use of medical care services” (USDHHS 2004, p. 29). chronic bronchitis in the United States, and increases the risk of dying from chronic bronchitis” (USDHEW 1964, p. 302). Although chronic bronchitis is the term long used for chronic cough and sputum production, at the time it Exposure to Secondhand Smoke was also used to refer to what is now called COPD. The Health Consequences of Smoking: Chronic 1984 report, - The topic of secondhand smoke was first consid Obstructive Lung Disease , which focused on the respira - ered in the 1972 Surgeon General’s report, Health Conse- tory consequences of smoking, classified cigarette smok - quences of Smoking , in a chapter titled “Public Exposure ing as “... the major cause of COLD [chronic obstructive to Air Pollution from Tobacco Smoke” (USDHEW 1972). lung disease] morbidity in the United States...” (USDHHS - The involuntary inhalation of tobacco smoke by nonsmok 1984, p. 9). The 2004 report used the term COPD, finding ers has been referred to in the Surgeon General’s reports the evidence to be sufficient to infer a causal relationship as involuntary smoking or passive smoking. The smoke between smoking and both COPD morbidity and mortality inhaled has been called secondhand smoke or environ - (USDHHS 2004). The Surgeon General’s reports have also mental tobacco smoke. This chapter in the 1972 report addressed asthma, influenza, and pneumonia. reviewed the accumulating evidence on levels of air pol- The effects of smoking on reproductive health (Table - lutants, such as carbon monoxide, in indoor environ 4.4) have been addressed since the 1964 report, covering ments where people were smoking. The report concluded an increasing number and diversity of topics as the multi- that “An atmosphere contaminated with tobacco smoke ple adverse effects of smoking on reproductive health were can contribute to the discomfort of many individuals” identified. In fact, the 1964 report considered only birth (USDHEW 1972, p. 7). The 1982 report, which had a chap - weight and devoted just one page to the topic, citing just ter on the relationship between exposure to secondhand five retrospective and two prospective studies (USDHEW smoke and lung cancer (USDHHS 1982), reviewed the 1964). Over time, the effects of smoking have been found findings of three epidemiologic studies, but it did not offer to extend from fertility to pregnancy and its outcome as a conclusion, while noting the limited evidence available. well as the subsequent development of the child. There - The Health Consequences of Involun The 1986 report, - has also been substantial advancement in the understand tary Smoking , was the first to have involuntary smoking ing of how smoking affects reproductive health, the health as its topic, and the 2006 report followed suit, as it was of the fetus, and neurodevelopment as summarized in the titled The Health Consequences of Involuntary Exposure 2010 report (USDHHS 2010). Male sexual functioning, to Tobacco Smoke (USDHHS 1986, 2006). not directly mentioned in the 1964 report, was covered The 1984 Surgeon General’s report addressed COPD, extensively in the 2004 report (USDHHS 2004), and a and the report’s chapter on passive smoking addressed the causal conclusion on the relationship between smoking Chapter 4 68

98 —50 Years of Progress The Health Consequences of Smoking nancies considered. respiratory consequences, other than cancer, of exposure Women and Smoking, The 2001 report, had con- to secondhand smoke. By that time, a substantial body - of literature had accumulated on the respiratory conse sidered the relationship between exposure to secondhand quences of exposure to secondhand smoke in children, smoke and breast cancer, and that topic was discussed in the 2006 report as well. Other cancers considered in and there was a more limited body of evidence related to relation to exposure to secondhand smoke included nasal adults. Notably, the conclusions in the 1984 report were sinus cavity and nasopharyngeal carcinoma (2006), and overall summaries of the evidence and not statements as cervical cancer (2006); the conclusions drawn were that to the strength of the evidence for causation. the evidence was either suggestive (breast cancer and Exposure to secondhand smoke and its effects was the sole topic of the 1986 report. With regard to the effects nasal sinus cavity) or inadequate (nasopharyngeal carci - noma and cervical cancer). Reports after 2006 expanded of parental smoking on child respiratory health, that the topics related to exposure to secondhand smoke and report addressed the range of outcomes considered in the childhood health to include adverse effects on reproduc- 1984 report, comprehensively reviewed the evidence, and - tion, risk for sudden infant death syndrome, and neurode - offered summary conclusions, but it did not provide state ments on the strength of evidence for causation. The 1986 velopment. The 1986 report did not cover exposure to second- - report did, however, comprehensively cover the relation hand smoke and cardiovascular diseases because only a ship of lung cancer to exposure to secondhand smoke and few studies on that topic had been reported at that time. concluded that involuntary smoking caused lung cancer in never smokers. This causal conclusion was repeated in The 2001 report was the first to consider the topic, and found that the evidence did indicate a causal relationship. - the 2006 report, which also addressed exposure to second hand smoke. That report also found sufficient evidence to Finally, the 2006 report found that the evidence for a link between exposure to secondhand smoke and coronary infer causation for the principal adverse effects considered heart disease was sufficient to infer a causal relationship, in the earlier reports. The 2006 report covered childhood but it designated as suggestive the evidence for a similar cancers as well, but the evidence was not judged to be suf - link with atherosclerosis and cerebrovascular disease. - ficient to infer a causal relationship for any of the malig Summary , Preventing Tobacco Use Among Youth and Young Adults Over the 50 years that began with the seminal 1964 report, the conclusions of the Surgeon Generals’ reports provided additional coverage of the effects of smoking on smoking and health have evolved greatly, moving from on the health of children, adolescents, and young adults, highlighting the linkages between early life events and the few causal associations set forth in the 1964 report to the inference of causal relationships between not only subsequent risk for disease (USDHHS 2012). active smoking but also exposure to secondhand smoke Notably, this 2014 review extends the list of diseases and other adverse health effects caused by smoking and and a wide range of diseases and other adverse health - reaffirms the widespread consequences of smoking. In the effects. The 2004 and 2006 reports provided compre hensive coverage of the evidence on active smoking and 2004 report, it was noted that smoking affects nearly every exposure to secondhand smoke, respectively, and the 2010 organ of the body; the evidence in this report provides report addressed the mechanisms underlying the causal additional support for that finding. relationships described in these reports. The 2012 report, 69 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

99 Surgeon General’s Report Conclusions from Surgeon General’s report on active cigarette smoking and cancer Table 4.1 Additional or updated Highest level conclusion(s) Conclusion(s) from the conclusion(s) from First mention and finding(s) from subsequent Surgeon 2004 Surgeon General’s the 2014 Surgeon in a Surgeon General’s report General’s reports before 2004 report Disease General’s report (year) (year) “The evidence is sufficient to “Smoking is a cause of bladder “Available data suggest an — Bladder — infer a causal relationship cancer; cessation reduces risk by association between cigarette between smoking and ... about 50 percent after only a few smoking and urinary bladder cancer bladder cancer.” (p. 26) years, in comparison with in the male but are not sufficient to continued smoking.” (1990, p. 10) support a judgment on the causal significance of this association.” (1964, p. 225) “The evidence is suggestive of — — Brain (adult) —— — — no causal relationship between smoking cigarettes and brain cancer in men and women.” (p. 26) “Thus, active smoking does not appear “The evidence is suggestive of no “The evidence is — Breast — to appreciably affect breast cancer causal relationship between sufficient to identify risk overall. However, several issues active smoking and breast mechanisms by which were not entirely resolved, including cancer.” (p. 26) cigarette smoking may whether starting to smoke at an early cause breast cancer.” age increases risk, whether certain subgroups defined by genetic polymorphisms are differentially affected by smoking, and whether ETS exposure affects risk.” (2001, p. 217) “The evidence is suggestive but not sufficient to infer a causal “The totality of the evidence does not relationship between support an association between active smoking and breast smoking and risk for breast a cancer.” (Chapter 6) cancer.” (2001, p. 224) “There are conflicting results in studies Cervical “Smoking has been consistently — “The evidence is sufficient to published to date on the existencof a associated with an increased risk for — infer a causal relationship relationship between smoking cervical cancer. The extent to which between smoking and and cervical cancer; further research this association is independent of cervical cancer.” is necessary to define whether an human papillomavirus infection is (p. 26) association exists and, if so, whether that uncertain.” (2001, p. 224) association is direct or indirect.” (1982, p. 8) “Smoking may be associated with an increased risk for vulvar cancer, but the extent to which the association is independent of human papillomavirus infection is uncertain.” (2001, p. 224) Chapter 4 70

100 The Health Consequences of Smoking —50 Years of Progress Table 4.1 Continued Additional or updated Highest level conclusion(s) First mention and finding(s) in conclusion(s) from the Conclusion(s) from the from subsequent Surgeon a Surgeon General’s report 2014 Surgeon 2004 Surgeon General’s General’s reports before 2004 Disease (year) General’s report report (year) “Women who smoke may have Colorectal “The evidence is suggestive but “The evidence is — increased risks for...colorectal not sufficient to infer a causal sufficient to infer a — 231 cancer.” (2001, p. relationship between smoking and causal relationship colorectal adenomatous polyps and between smoking and colorectal cancer.” (p. 26) colorectal adenomatous polyps and colorectal cancer.” (Chapter 6) “Several studies have reported that Endometrial “The evidence is sufficient to —— “Current smoking is associated infer that current smoking endometrial cancer is less frequent with a reduced risk for reduces the risk of endometrial among women who smoke cigarettes endometrial cancer, but the effect cancer in postmenopausal than among nonsmokers (Baron et is probably limited to women.” (p. 26) l. 1986). Cigarette smoking exerts a postmenopausal disease. The risk an antiestrogenic effect that may for this cancer among former explain this inverse association. The smokers generally appears more public health significance of this similar to that of women who have association is limited because of the never smoked.” (2001, p. 224) overall adverse impact of cigarette smoking on morbidity and mortality.” (1989, p. 58) “The evidence is sufficient to “Cigarette smoking is a major “The evidence on the tobacco- —— Esophageal cause of esophageal cancer in the infer a causal relationship esophageal cancer relationship supports United States.” (1982, p. 7) between smoking and cancers of the belief that an association exists. the esophagus.” (p. 26) However, the data are not adequate to decide whether the relationship is causal.” (1964, p. 218) “There is a positive association between “Cigarette smoking is a contributory Kidney —— “The evidence is sufficient to smoking and kidney cancer, with factor in the development of kidney infer a causal relationship relative risks ranging from 1 to more cancer in the United States. The term between smoking and renal than 5. The increased risk of kidney ‘contributory factor’ by no means cell, [and] renal pelvis... cancer due to cigarette smoking is excludes the possibility of a causal role cancers.” (p. 26) found for both males and females, and for smoking in cancers of this there is a dose-response relationship as site.” (1982, p. 7) measured by the number of cigarettes smoked per day.” (1989, p. 56) 71 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

101 Surgeon General’s Report Table 4.1 Continued Additional or updated Highest level conclusion(s) conclusion(s) from First mention and finding(s) Conclusion(s) from the from subsequent Surgeon the 2014 Surgeon in a Surgeon General’s report 2004 Surgeon General’s General’s reports before 2004 General’s report (year) report Disease (year) “The evidence is sufficient to “Cigarette smoking is causally “Evaluation of the evidence leads to — Laryngeal — infer a causal relationship associated with cancer of the the judgment that cigarette smoking between smoking and cancer of lung, larynx, oral cavity, and is a significant factor in the the larynx.” (p. 25) esophagus in women as well as in causation of laryngeal cancer in the p.126) men...” (1980, p.212) male.” (1964, “The evidence is sufficient to — Leukemia “Leukemia has recently been “Smoking may be associated with an (acute) infer a causal relationship — increased risk for acute myeloid implicated as a smoking-related between smoking and acute leukemia among women but does disease ... but this observation has myeloid leukemia.” (p. 26) not been consistent.” (1990, p. 176) not appear to be associated with other lymphoproliferative or hematologic cancers.” (2001, p. 231) “The evidence is “The evidence is suggestive but “Women who smoke may have “Primary hepatocellular cancer has been Liver sufficient to infer a not sufficient to infer a causal associated with smoking in a number increased risks for liver causal relationship relationship between smoking cancer...” (2001, p. 231) ofrecent studies.”(1990,p. 176) between smoking and and liver cancer.” (p. 26) hepatocellular carcinoma.” (Chapter 6) Chapter 4 72

102 —50 Years of Progress The Health Consequences of Smoking Table 4.1 Continued Additional or updated Highest level conclusion(s) conclusion(s) from Conclusion(s) from the First mention and finding(s) from subsequent Surgeon the 2014 Surgeon 2004 Surgeon General’s in a Surgeon General’s report General’s reports before 2004 General’s report Disease report (year) (year) “The evidence is sufficient to “The evidence is “Additional epidemiological, Lung “Cigarette smoking is causally related infer a causal relationship sufficient to conclude pathological, and experimental to lung cancer in men; the magnitude between smoking and lung that the risk of data not only confirm the of the effect of cigarette smoking far developing cancer.” (p. 25) conclusion of the Surgeon outweighs all other factors. The data adenocarcinoma of the General’s 1964 Report regarding for women, though less extensive, lung from cigarette lung cancer in men but point in the same direction.” (1964, p. smoking has increased strengthen the causal relationship 196) since the 1960s.” of smoking to lung cancer in women.” (1967, p. 36) “The evidence is sufficient to conclude “Cigarette smoking is causally that the increased risk of related to lung cancer in both men adenocarcinoma of the and women.” (1979, p. 1-16) lung in smokers results from changes in the “Cigarette smoking is the major design and composition cause of lung cancer in the of cigarettes since the United States.” (1982, p. 5) 1950s.” “Cigarette smoking is the major “The evidence is not cause of lung cancer among sufficient to specify women. About 90 percent of all which design changes lung cancer deaths among U.S. are responsible for women smokers are attributable the increased risk of to smoking.” (2001, p. 13) adenocarcinoma, but there is suggestive evidence that ventilated filters and increased levels of tobacco-specific nitrosamines have played a role.” “The evidence shows that the decline of squamous carcinoma follows the trend of declining smoking prevalence.” (Chapter 6) 73 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

103 Surgeon General’s Report Continued Table 4.1 Highest level conclusion(s) Additional or updated First mention and finding(s) Conclusion(s) from the from subsequent Surgeon conclusion(s) from in a Surgeon General’s report 2004 Surgeon General’s General’s reports before 2004 the 2014 Surgeon (year) report Disease (year) General’s report “The evidence is sufficient to indicate “Epidemiological studies “The causal relationship of the — Oral infer a causal relationship that smoking is a significant causal smoking of pipes to the development — cavity between smoking and cancers factor in the development of oral of cancer of the lip appears to be and p. 204) established.” (1964, of the oral cavity and cancer. The risk increases with the pharyngeal pharynx.” (p. 25) number of cigarettes smoked per p. 1-17) day.” (1979, “Although there are suggestions of relationships between cancer of other “Cigarette smoking is a major specific sites of the oral cavity and the cause of cancers of the oral several forms of tobacco use, their cavity in the United causal implications cannot at present States.” (1982, p. 6) be stated.” (1964, p. 205) “The evidence is inadequate to — — Ovarian “Smoking does not appear to be — — infer the presence or absence of associated with risk for ovarian a causal relationship between cancer.” (2001, p. 224) smoking and ovarian cancer.” (p. 26) Pancreatic “The evidence is sufficient to “Smoking cessation reduces the — “Cigarette smoking is a contributory — infer a causal relationship risk of pancreatic cancer, factor in the development of between smoking and compared with continued pancreatic cancer in the United pancreatic cancer.” (p. 26) smoking, although this States. This relationship is not as reduction in risk may only be strong as that noted for the measurable after 10 years of association between smoking and abstinence.” (1990, p. 10) cancers of the lung, larynx, oral cavity, and esophagus. The term ‘contributory factor’ by no means excludes the possibility of a causal role for smoking in cancers of this site.” (1982, p. 7) Chapter 4 74

104 The Health Consequences of Smoking —50 Years of Progress Table Continued 4.1 First mention and finding(s) Highest level conclusion(s) Conclusion(s) from the Additional or updated Di s ease from subsequent Surgeon in a Surgeon General’s report 2004 Surgeon General’s conclusion(s) from (year) General’s reports before 2004 report the 2014 Surgeon (year) General’s report Prostate cancer “The evidence is “The evidence is suggestive — — — suggestive of no causal of no causal relationship — relationship between between smoking and risk smoking and the risk of for prostate cancer.” (p. 26) incident prostate cancer.” “The evidence is suggestive of a higher risk of death from prostate cancer in smokers than in nonsmokers.” “In men who have prostate cancer, the evidence is suggestive of a higher risk of advanced- stage disease and less well-differentiated cancer in smokers than in nonsmokers, and— independent of stage and histologic grade—a higher risk of disease progression.” (Chapter 6) “The evidence is sufficient to — “No relationship has been established Stomach “Data on smoking and cancer of the infer a causal relationship between tobacco use and stomach — stomach ... are unclear.” (2001, p. 231) cancer.” (1964, p. 229) between smoking and gastric cancers.” (p. 26) Note: ETS = environmental tobacco smoke. a Refers to a general conclusion that was reached for breast cancer. 75 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

105 Surgeon General’s Report Conclusions from Surgeon General’s report on active cigarette smoking and cardiovascular Table 4.2 diseases Additional or updated First mention and finding(s) Conclusion(s) from the Highest level conclusion(s) Disease conclusion(s) from the 2004 Surgeon General’s in a Surgeon General’s report from subsequent Surgeon 2012/2014 Surgeon report (year) General’s reports before General’s report 2004 (year) “The evidence is sufficient to “The evidence is sufficient “Death from rupture of an “Cigarette smoking is a strong Abdominal conclude that there is a causal to infer a causal relationship atherosclerotic abdominal risk factor for atherosclerotic aortic aneurysm relationship between active between smoking and aortic aneurysm.” (1979, p. aneurysm is more smoking in adolescence and abdominal aortic 4-56) common in cigarette young adulthood and early aneurysm.” (p. 27) smokers than in abdominal aortic atherosclerosis nonsmokers.” (1983, p. in young adults.” (2012, p. 111) 195) Atherosclerosis/ “The evidence is suggestive but “The evidence is sufficient “Autopsy studies suggest that “Cigarette smoking is the most not sufficient to conclude that to infer a causal relationship cigarette smoking is powerful risk factor predisposing peripheral there is a causal relationship between smoking and associated with a significant to atherosclerotic peripheral vascular disease between smoking in adolescence subclinical atherosclerosis.” (p. increase in the atherosclerosis vascular disease.” (1983, p. 8) and young adulthood and 26) of the aorta and coronary coronary artery atherosclerosis arteries.” (1969, p. 4) in adulthood.” (2012, p. 111) “The evidence is sufficient to “Cigarette smoking is a major “Additional evidence — Cerebrovascular infer a causal relationship — cause of cerebrovascular strengthens the association disease between smoking and disease (stroke), the third between cigarette smoking and stroke.” (p. 27) leading cause of death in the cerebrovascular disease, and United States.” (1989, p. 12) suggests that some of the pathogenetic [ sic ] considerations pertinent to coronary heart disease may also apply to cerebrovascular disease.” (1967, p. 28) “The evidence is sufficient “In summary, for the “Male cigarette smokers have a Coronary heart — — to infer a causal relationship purposes of preventive higher death rate from coronary disease between smoking and medicine, it can be artery disease than non-smoking coronary heart disease.” (p. concluded that smoking is males, but it is not clear that the 27) causally related to coronary association has causal heart disease for both men significance.” (1964, p. 327) and women in the United States.” (1979, p. 1-15) Chapter 4 76

106 The Health Consequences of Smoking —50 Years of Progress Conclusions from Surgeon General’s report on active cigarette smoking and respiratory diseases Table 4.3 Disease Conclusion(s) from the Highest level conclusion(s) Additional or updated First mention and finding(s) in a 2004 Surgeon General’s from subsequent Surgeon conclusion(s) from the Surgeon General’s report (year) report General’s reports before 2012/2014 Surgeon 2004 (year) General’s report “The evidence is sufficient “The evidence is inadequate to — Asthma “Cigarette smoking does not p . 302) — , appear to cause asthma.” (1964 to conclude that there is a infer the presence or absence causal relationship between of a causal relationship active smoking and between active smoking and wheezing severe enough to asthma in adults.” (p. 28) be diagnosed as asthma in susceptible child and adolescent populations.” (2012, p. 111) “The evidence is suggestive “The evidence is suggestive but not sufficient to infer a but not sufficient to infer a causal relationship causal relationship between between active smoking active smoking and increased and incidence of asthma in nonspecific bronchial adolescents.” hyperresponsiveness.” (p. 28) (2014, Chapter 7) “The evidence is suggestive but not sufficient to infer a causal relationship between active smoking and exacerbation of asthma among children and adolescents.” (2014, Chapter 7) “The evidence is suggestive but not sufficient to infer a causal relationship between active smoking and the incidence of asthma in adults.” (2014, Chapter 7) “The evidence is sufficient to “The evidence is sufficient infer a causal relationship to infer a causal between active smoking and relationship between poor asthma control.” (p. 28) active smoking and exacerbation of asthma in adults.” (2014, Chapter 7) 77 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

107 Surgeon General’s Report Table 4.3 Continued Additional or updated Highest level conclusion(s) conclusion(s) from the Conclusion(s) from the from subsequent Surgeon First mention and finding(s) 2012/2014 Surgeon 2004 Surgeon General’s General’s reports before 2004 in a Surgeon General’s report General’s report report Disease (year) (year) “The evidence is sufficient to “The evidence is sufficient “Cigarette smoking is the “Cigarette smoking is the COPD infer a causal relationship to infer that smoking is the major cause of COLD ... most important of the causes (Formerly designated between active smoking and dominant cause of chronic morbidity in the United of chronic bronchitis in the as chronic bronchitis; chronic obstructive pulmonary obstructive pulmonary States; 80 to 90 percent of United States, and increases emphysema; COLD; disease morbidity and disease (COPD) in men and COLD in the United States is the risk of dying from chronic obstructive chronic mortality.” (p. 28) women in the United States. attributable to cigarette bronchitis.” (1964, p. 302) bronchopulmonary Smoking causes all elements smoking.” (1984, p. 9) disease) of the COPD phenotype, “A relationship exists between “Cigarette smoking is a primary including emphysema and pulmonary emphysema and cause of COPD among women, and damage to the airways of the cigarette smoking but it has the risk increases with the amount lung.” (2014, Chapter 7) not been established that the and duration of smoking. relationship is causal. The “Chronic obstructive Approximately 90 percent of smoking of cigarettes is pulmonary disease mortality mortality from COPD among associated with an increased has increased dramatically women in the United States risk of dying from pulmonary in men and women since c an be attributed to cigarette emphysema.” (1964, p. 302) the 1964 Surgeon General’s smoking.” (2001, p. 14) report. The number of women dying from COPD the number of now surpasses men.” (2014, Chapter 7) “The evidence is suggestive but not sufficient to infer that women are more susceptible to develop severe chronic obstructive pulmonary disease at younger ages.” (2014, Chapter 7) “The evidence is sufficient to infer that severe a -1- antitrypsin deficiency and cutis laxa are genetic causes of chronic obstructive pulmonary disease.” (2014, Chapter 7) Chapter 4 78

108 —50 Years of Progress The Health Consequences of Smoking Table 4.3 Continued Additional or updated Highest level conclusion(s) conclusion(s) from the Conclusion(s) from the from subsequent Surgeon First mention and finding(s) 2012/2014 Surgeon General’s 2004 Surgeon General’s Surgeon General’s report in a General’s reports before 2004 Disease report report (year) (year) “Cough, sputum production, or the “The evidence is sufficient to infer “Cigarette smokers have Chronic respiratory —— two combined are consistently an increased frequency of relationship between active causal a symptoms (cough, more frequent among cigarette respiratory symptoms, and at major respiratory smoking and all phlegm, wheeze, smokers than among non least two of them, cough and adults, including symptoms among dyspnea, etc.) p. smokers.” (1964, sputum production, are coughing, phlegm, wheezing, and 302) dose-related.” (1979, p. 1-18) dyspnea.” (p. 28) —— “The evidence is sufficient to “Although death certification “Smoking cessation reduces Influenza, of respiratory symptoms rates infer a causal relationship shows that cigarette smokers pneumonia, such as cough, sputum between smoking and acute have a moderately increased infections, and production, and wheezing,and respiratory illnesses, including risk of death from influenza acute respiratory respiratory infections such as pneumonia, in persons without and pneumonia, an illnesses bronchitis and pneumonia, underlying smoking-related association of cigarette compared with continued chronic obstructive lung smoking and infectious smoking.” (1990, p. 11) disease.” (p. 27) diseases is not otherwise substantiated.” (1964, p. 302) “The evidence is suggestive but not sufficent to infer a causal relationship between smoking and acute respiratory infections among persons with preexisting chronic obstructie pulmonary disease.” (p. 27) 79 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

109 Surgeon General’s Report Continued Table 4.3 Additional or updated Highest level conclusion(s) conclusion(s) from the from subsequent Surgeon Conclusion(s) from the 2004 First mention and finding(s) in a 2012/2014 Surgeon General’s reports before 2004 Surgeon General’s report Surgeon General’s report (year) General’s report Disease (year) — — Tuberculosis — “The evidence is sufficient to — — — infer a causal relationship between smoking and an increased risk of Mycobacterium tuberculosis disease.” (2014, Chapter 7) “The evidence is sufficient to infer a causal relationship between smoking and mortality due to tuberculosis.” (2014, Chapter 7) “The evidence is suggestive of a causal relationship between smoking and the risk of recurrent tuberculosis disease.” (2014, Chapter 7) “The evidence is inadequate to infer the presence or absence of a causal relationship between active smoking and the risk of tuberculosis infection.” (2014, Chapter 7) “The evidence is sufficient “The evidence is sufficient to “Cigarette smoking “Cigarette smoking is Lung function level to conclude that there is a infer a causal relationship accelerates the age-related associated with a reduction in causal relationship between between active smoking in decline in lung function that ventilatory function. Among active smoking and both adulthood and a premature occurs among never smokers. males, cigarette smokers have reduced lung function and onset of and an accelerated age- With sustained abstinence a greater prevalence of impaired lung growth related decline in lung from smoking, the rate of breathlessness than non- during childhood and function.” (p. 27) decline in pulmonary smokers.” (1964, p. 302) adolescence.” (2012, function among former p. 111) smokers returns to that of “The evidence is sufficient to never smokers.” (1990, p. 11) infer a causal relationship between sustained cessation from smoking and a return of the rate of decline in pulmonary function to that of persons who had never smoked.” (p. 27) Chapter 4 80

110 —50 Years of Progress The Health Consequences of Smoking Continued Table 4.3 Additional or updated Highest level conclusion(s) Conclusion(s) from the conclusion(s) from the from subsequent Surgeon First mention and finding(s) 2004 Surgeon General’s Surgeon 2012/2014 General’s reports before 2004 in a Surgeon General’s report report General’s report Disease (year) (year) Respiratory effects — — “The evidence is sufficient to infer a causal “Cigarette smoking during due to active smoking — — relationship between active smoking and childhood and adolescence during childhood and impaired lung growth during childhood produces significant health adolescence and adolescence.” (p. 27) problems among young people, including cough and phlegm production, an “The evidence is sufficient to increased number and severity infer a causal relationship of respiratory illnesses, between active smoking and decreased physical fitness, an the early onset of lung function unfavorable lipid profile, and decline during late adolescence potential retardation in the and early adulthood.” (p. 27) rate of lung growth and the level of maximum lung function.” (1994, p. 41) “The evidence is sufficient to infer a causal relationship between active smoking and respiratory symptoms in children and adolescents, including coughing, phlegm, wheezing, and dyspnea.” (p. 27) “The evidence is sufficient to infer a causal relationship between active smoking and asthma-related symptoms (i.e., wheezing) in childhood and adolescence.” (p. 27) = chronic obstructive Note: COLD = chronic obstuctive lung disease; COPD pulmonary disease. 81 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

111 Surgeon General’s Report Conclusions from Surgeon General’s report on active cigarette smoking and adverse reproductive outcomes or childhood Table 4.4 neurobehavioral disorders Highest level conclusion(s) from Additional or updated Conclusion(s) from the First mention and subsequent Surgeon General’s conclusion(s) from the 2014 2004 Surgeon General’s finding(s) in a Surgeon reports before 2004 (year) Surgeon General’s report Disease report General’s report (year) “The evidence is suggestive but “The evidence is inadequate “Maternal smoking during “According to studies of physical, Child not sufficient to infer a causal to infer the presence or pregnancy may adversely affect long-term growth and behavioral,and relationship between maternal absence of a causal the child’s long-term growth, development, smoking cognitive prenatal smoking and relationship between intellectual development, and during pregnancy may development disruptive behavioral disorders, maternal smoking and behavioral characteristics.” (1980, affect physical growth, and ADHD in particular, physical growth and p. 11) mental development, and among children.” neurocognitive development behavioral characteristics of children.” (p. 28) of children at least up to the age of 11.” (1979, p. “The evidence is insufficient to 1-21) infer the presence or absence of a causal relationship between maternal prenatal smoking and anxiety and depression in children.” “The evidence is insufficient to infer the presence or absence of a causal relationship between maternal prenatal smoking and Tourette syndrome.” “The evidence is insufficient to infer the presence or absence of a causal relationship between maternal prenatal smoking and schizophrenia in her offspring.” “The evidence is insufficient to infer the presence or absence of a causal relationship between maternal prenatal smoking and intellectual disability.” (Chapter 9) Chapter 4 82

112 The Health Consequences of Smoking —50 Years of Progress Table 4.4 Continued Additional or updated Highest level conclusion(s) from Conclusion(s) from the First mention and conclusion(s) from the 2014 subsequent Surgeon General’s 2004 Surgeon General’s finding(s) in a Surgeon Surgeon General’s report reports before 2004 (year) report Disease General’s report (year) “The evidence is sufficient to “The evidence is inadequate Congenital “...no conclusions can be “The accumulated evidence does infer a causal relationship to infer the presence or not support a conclusion that drawn about any malformations between maternal smoking in absence of a causal maternal smoking increases the relationship between early pregnancy and orofacial relationship between incidence of congenital maternal cigarette smoking 1-22) p. clefts.” maternal smoking and malformations.” (1979, and congenital congenital malformations in malformations at the “The evidence is suggestive but “There are insufficient data to general.” (p. 28) present time.” (1973, p. not sufficient to infer a causal support a judgment on whether 137) relationship between maternal “The evidence is suggestive maternal and/or paternal smoking in early pregnancy and but not sufficient to infer a cigarette smoking increases the clubfoot, gastroschisis, and causal relationship between risk of congenital atrial septal heart maternal smoking and oral malformations.” (1980, p. 11) defects.” (Chapter 9) clefts.” (p. 28) “Smoking does not appear to affect the overall risk for congenital malformations.” (2001, p. 307) Fertility “The evidence is sufficient to — “Studies in women and men “The available information suggests — infer a causal relationship suggest that cigarette smoking that current smoking is related to between smoking and reduced may impair fertility.” (1980, p . low sperm density. However, these fertility in women.” (p. 28) 12) data are limited.” (1990, p. 405) “The evidence is inadequate to infer the presence or absence “Women who smoke have increased of a causal relationship risks for conception delay and for between active smoking and both primary and secondary sperm quality.” infertility.” (p. 28) (2001, p. 307) “The evidence is suggestive but “Cigarette smoking is now — “...it appears that Fetal death, — not sufficient to infer a causal considered to be a probable maternal smoking stillbirths, relationship between maternal cause of ...increased infant during pregnancy may be and infant active smoking and spontaneous mortality.” (1989, p. 20) associated with mortality abortion.” (Chapter 9) an increased incidence “The risk for perinatal mortality of spontaneous abortion, —both stillbirth and neonatal stillbirth, and neonatal deaths—and the risk for sudden death and that this infant death syndrome (SIDS) are relationship may be most increased among the offspring of marked in the presence women who smoke during of other risk pregnancy.” (2001, p. 307) factors.” (1969, p. 5) “Women who smoke may have a modest increase in risks for... 307) spontaneous abortion.”(2001, p. 83 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

113 Surgeon General’s Report Continued Table 4.4 First mention and Highest level conclusion(s) from Additional or updated Conclusion(s) from the finding(s) in a Surgeon subsequent Surgeon General’s conclusion(s) from the 2014 2004 Surgeon General’s General’s report (year) Surgeon General’s report reports before 2004 (year) Disease report Infant birth “Infants born to women who “Women who smoke — “The evidence is sufficient to infer weight — smoke during pregnancy have a cigarettes during a causal relationship between lower average birth weight...than pregnancy tend to have maternal active smoking and fetal infants born to women who d o babies of lower birth growth restriction and low birth not smoke." (2001, p. 307) weight.” (1964, p. 343) weight.” (p. 28) “Infants born to women who smoke during pregnancy ...are more likely to be small for gestational age than are infants born to women who do not smoke.” (2001, p. 307) “The evidence is sufficient to “The evidence is suggestive “In summary, the level of sexual “...element of Male sexual infer a causal relationship but not sufficient to infer a activity does not appear to be masculinity as indicated function between smoking and erectile causal relationship between affected by cigarette smoking. by external morphologic dysfunction.” (Chapter 9) smoking and erectile Cigarette smoking may be features”... “weakness of dysfunction.” (p. 29) associated with impaired male the masculine sexual performance. ...Because of component is limited and uncontrolled data, no significantly more conclusions can be drawn frequent in smokers regarding sexual performance or than in nonsmokers, PBI among former and most frequent in smokers.” (1990, heavy smokers.” (1964, pp. 403–4) pp. 383–4) Chapter 4 84

114 The Health Consequences of Smoking —50 Years of Progress Table 4.4 Continued First mention and finding(s) Additional or updated Highest level conclusion(s) from in a Surgeon General’s report conclusion(s) from the 2014 Conclusion(s) from the 2004 subsequent Surgeon General’s (year) Surgeon General’s report Disease Surgeon General’s report reports before 2004 (year) “Smoking during pregnancy is “Maternal smoking “The evidence is sufficient “The evidence is sufficient to Pregnancy infer a causal relationship to infer a causal relationship increases the risk of fetal associated with increased risks complications between maternal active death through maternal for preterm premature rupture between maternal active smoking and premature of membranes, abrupio smoking and ectopic complications such as pregnancy.” (Chapter 9) abruptio placenta, rupture of the membranes, placentae, and placenta previa, placenta previa, and with a modest increase in placenta previa, and placental abruption.” (p. 28) antepartum hemorrhage, risk for preterm delivery.” (2001, p. 14) and prolonged rupture of membranes.” “The evidence is sufficient “Women who smoke may have a (1979, p. 1-22) to infer a causal relationship modest increase in risks for between maternal active ectopic pregnancy and smoking and preterm spontaneous abortion.” (2001, p. delivery and shortened 14) gestation.” (p. 28) “Women who smoke during “The evidence is sufficient to pregnancy have a decreased risk infer a causal relationship for preeclampsia.” (2001, p. 14) between maternal active smoking and a reduced risk for preeclampsia.” (p. 28) “The evidence is suggestive but not sufficient to infer a causal relationship between maternal active smoking and ectopic pregnancy.” (p. 28) “The evidence is suggestive but not sufficient to infer a causal relationship between maternal active smoking and spontaneous abortion.” (p. 28) 85 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

115 Surgeon General’s Report Table 4.4 Continued First mention and Additional or updated Highest level conclusion(s) from Conclusion(s) from the finding(s) in a Surgeon conclusion(s) from the 2014 subsequent Surgeon General’s 2004 Surgeon General’s General’s report (year) Surgeon General’s report Disease reports before 2004 (year) report “The evidence is sufficient to “In utero exposure to maternal — — Respiratory infer a causal relationship smoking is associated with — — effects in between maternal smoking reduced lung function among during pregnancy and a infants...” (2001, p. 14) infants and reduction of lung function in children due infants.” (p. 27) to maternal “The evidence is suggestive smoking active but not sufficient to infer a causal relationship between maternal smoking during pregnancy and an increase in the frequency of lower respiratory tract illnesses during infancy.” (p. 27) “The evidence is suggestive but not sufficient to infer a causal relationship between maternal smoking during pregnancy and an increased risk for impaired lung function in childhood and adulthood.” (p. 27) “... the risk for sudden infant “Smoking by pregnant “The evidence is sufficient to Sudden — infer a causal relationship death syndrome (SIDS) are women contributes to the infant death — risk of their infants being increased among the offspring of between sudden infant death syndrome victims of the “sudden syndrome and maternal women who smoke during (SIDS) infant death smoking during and after pregnancy.” (2001, p. 307) syndrome.” (1979, p. 1-22) pregnancy.” (p. 28) Note: ADHD = attention deficit hyperactivity disorder; PBI = penile- brachial index. Chapter 4 86

116 The Health Consequences of Smoking —50 Years of Progress Conclusions from Surgeon General’s report on active cigarette smoking and other adverse Table 4.5 health effects Conclusion(s) from the Highest level conclusion(s) from Additional or updated First mention and finding(s) 2004 Surgeon General’s subsequent Surgeon General’s in a Surgeon General’s report conclusion(s) from the 2014 report reports before 2004 (year) (year) Surgeon General’s report Disease “Smoking is associated with — — — Accidents accidental deaths from fires in — — — the home.” (1964, p. 39) “No conclusive information is available on the effects of smoking on traffic accidents.” (1964, p. 39) “The evidence is sufficient to “The evidence is suggestive but “Tobacco use, excessive alcohol Dental — infer a causal relationship not sufficient to infer a causal use, and inappropriate dietary — diseases between smoking and relationship between active practices contribute to many periodontitis.” (p. 29) cigarette smoking and dental diseases and disorders. In caries.” particular, tobacco use is a risk “The evidence is inadequate to factor for oral cavity and infer the presence or absence of pharyngeal cancers, periodontal a causal relationship between “The evidence is suggestive but diseases, candidiasis, and dental smoking and coronal dental not sufficient to infer a causal caries, among other caries.” (p. 29) relationship between cigarette a diseases.” (2000, p. 6) smoking and failure of dental implants.” (Chapter 10) “The evidence is suggestive but not sufficient to infer a causal relationship between smoking and root-surface caries.” (p. 29) “The evidence is sufficient to “Smoking appears to affect — Diabetes — infer that cigarette smoking is glucose regulation and mellitus a cause of diabetes.” related metabolic processes, but conflicting data exist on — “The risk of developing diabetes the relationship of smoking — is 30–40% higher for active and the development of type smokers than nonsmokers.” 2 diabetes mellitus and “There is a positive dose-response gestational diabetes among relationship between the number of women.” (2001, p. 14) cigarettes smoked and the risk of developing diabetes.“ (Chapter 10) 87 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

117 Surgeon General’s Report Table 4.5 Continued Highest level conclusion(s) Additional or updated First mention and finding(s) from subsequent Surgeon conclusion(s) from the 2014 Conclusion(s) from the 2004 in a Surgeon General’s report General’s reports before Disease Surgeon General’s report Surgeon General’s report (year) 2004 (year) “The evidence is sufficient to infer — — Diminished — — — — a causal relationship between health status smoking and diminished health status that may manifest as increased absenteeism from work and increased use of medical care services.” (p. 29) “The evidence is sufficient to infer a causal relationship between smoking and increased risks for adverse surgical outcomes related to wound healing and respiratory complications.” (p. 29) “Tobacco amblyopia had been Eye “The evidence is sufficient to “The evidence is sufficient to “Women who smoke have related to pipe and cigar infer a causal relationship infer a causal relationship an increased risk for diseases smoking by clinical impressions. between smoking and nuclear between cigarette smoking and cataract.” (2001, p. 15) The association has not been cataract.” (p. 29) neovascular and atrophic forms substantiated by epidemiological “Women who smoke may of age-related macular or experimental studies.” (1964, have an increased risk for degeneration.” age-related macular p. ) 342 “The evidence is suggestive but not degeneration.” (2001, p. 15) sufficient to infer that smoking cessation reduces the risk of “Studies show no consistent advanced age-related macular association between smoking degeneration.” (Chapter 10). and open-angle glaucoma.” (2001, p. 15) “The evidence is sufficient to “Women who currently — — Hip — — infer a causal relationship smoke have an increased fractures between smoking and hip risk for hip fracture fractures.” (p. 29) compared with women who p. ) do not smoke.” (2001, 321 “Increased mortality of — — — Liver — — — smokers from cirrhosis of the cirrhosis liver has been shown in the prospective studies. The data are not sufficient to support a direct or causal association.” (1964, p. 342) Chapter 4 88

118 The Health Consequences of Smoking —50 Years of Progress Table 4.5 Continued Highest level conclusion(s) Additional or updated Conclusion(s) from the First mention and from subsequent Surgeon conclusion(s) from the 2014 2004 Surgeon General’s finding(s) in a Surgeon General’s reports before Surgeon General’s report report Disease General’s report (year) 2004 (year) “In postmenopausal women, the “Postmenopausal women — — Low bone — — evidence is sufficient to infer a who currently smoke have density causal relationship between lower bone density than do smoking and low bone women who do not density.” (p. 29) smoke.” (2001, p. 321) Peptic “The evidence is sufficient to “The relationship between “Epidemiological studies — — infer a causal relationship cigarette smoking and death indicate an association ulcer between smoking and peptic rates from peptic ulcer, between cigarette smoking ulcer disease in persons who are especially gastric ulcer, is and peptic ulcer which is positive.” (p. Helicobacter pylori confirmed. In addition, greater for gastric than for 29) morbidity data suggest a duodenal ulcer.” (1964, p. similar relationship exists 340) with the prevalence of reported disease from this cause.” (1967, p. 40) a Oral Health in America: A Report of the U.S. Department of Health and Human Services 2000. Surgeon General. 89 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

119 Surgeon General’s Report Conclusions from Surgeon General’s report on exposure to secondhand smoke Table 4.6 a and cancer Additional or updated Highest level conclusion(s) First mention and finding(s) conclusion(s) from the from subsequent Surgeon in a Surgeon General’s report 2014 Surgeon General’s reports before Conclusion(s) from the 2006 Disease (year) General’s report 2006 (year) Surgeon General’s report “The evidence is suggestive “Several studies suggest that — Breast “The evidence is suggestive but not but not sufficient to infer a exposure to environmental — sufficient to infer a causal relationship causal relationship tobacco smoke is associated with between secondhand smoke and breast between exposure to an increased risk of breast cancer, cancer.” secondhand tobacco smoke but this association remains (p. 15) and breast uncertain.” (2001, p. 13) cancer.” (Chapter 6 ) “The evidence is inadequate to infer — — — Cervical the presence or absence of a causal — — — relationship between secondhand smoke exposure and the risk of cervical cancer among lifetime nonsmokers.” (p. 15) “Although the currently available — “Involuntary smoking can Lung “The evidence is sufficient to infer a evidence is not sufficient — cause lung cancer in causal relationship between to conclude that passive or nonsmokers.” secondhand smoke exposure and lung involuntary smoking causes lung (1986, p. 13) cancer among lifetime nonsmokers. cancer in nonsmokers, the “Exposure to ETS is a cause This conclusion extends to all evidence does raise concern of lung cancer among secondhand smoke exposure, about a possible serious public women who have never regardless of location.” (p. 15) health problem.” (1982, p. 9) smoked.” (2001, “The evidence is suggestive but not sufficient — Nasal sinus — — to infer a causal relationship between — cavity and — p — 350 secondhand smoke exposure and a risk of nasopharyngeal . ) nasal sinus cancer among nonsmokers.” carcinoma (p. 15) “The evidence is inadequate to infer the presence or absence of a causal relationship between secondhand smoke exposure and a risk of nasopharyngeal carcinoma among nonsmokers.” (p. 15) Note: ETS = environmental tobacco smoke. a General conclusion on cancers other than lung: “The associations between cancers, other than cancer of the lung, and involuntary smoking require before a determination can be made about the relationship of involuntary smoking to these cancers.” (1986, p. 14) further investigation Chapter 4 90

120 The Health Consequences of Smoking —50 Years of Progress Conclusions from Surgeon General’s report on exposure to secondhand smoke and a Table 4.7 cardiovascular diseases Highest level conclusion(s) Additional or updated Conclusion(s) from the First mention and finding(s) from subsequent Surgeon conclusion(s) from the 2014 2006 Surgeon General’s in a Surgeon General’s General’s reports before 2006 Surgeon General’s report report Disease report (year) (year) — — “Studies of secondhand smoke Atherosclerosis/ — — — — and subclinical vascular disease, subclinical vascular disease particularly carotid arterial wall thickening, are suggestive but not sufficient to infer a causal relationship between exposure to secondhand smoke and atherosclerosis.” (p. 15) “The evidence is suggestive but “The evidence is sufficient to — — Cerebrovascular not sufficient to infer a causal — — infer a causal relationship disease relationship between exposure to between exposure to secondhand smoke and an secondhand smoke and increased risk of stroke.” (p. 15) increased risk of stroke.” “The estimated increase in risk for stroke from exposure to secondhand smoke is about 20–30%.” (Chapter 8) “The evidence is sufficient to infer a “Epidemiologic and other Coronary “The presence of such levels” as found causal relationship between exposure data support a causal heart in cigarettes “indicates that the effect disease to secondhand smoke and increased relationship between ETS of exposure to carbon monoxide may — risks of coronary heart disease exposure from the spouse on occasion, depending upon the — morbidity and mortality among both and coronary heart disease length of exposure, be sufficient to be men and women.” (p. 15) mortality among women harmful to the health of an exposed nonsmokers.” (2001, person. This would be particularly p. 356) significant for people who are already suffering from...coronary heart disease.” (1972, p. 7) Note: = environmental tobacco smoke. ETS a General conclusion on cardiovascular disease: “Further studies on the relationship between involuntary smoking and cardiovascular disease are needed determine whether involuntary smoking increases the risk of cardiovascular disease.” (1986, p. 14). [“The evidence is sufficient to infer that in order to smoking is the dominant cause of chronic obstructive pulmonary disease (COPD) in men and women in the United States. Smoking causes all elements of the COPD phenotype, including emphysema and damage to the airways of the lung” (2014)]. 91 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

121 Surgeon General’s Report Conclusions from Surgeon General’s report on exposure to secondhand smoke and respiratory effects in children Table 4.8 Additional or updated Highest level conclusion(s) conclusion(s) from from subsequent Surgeon First mention and finding(s) the 2012/2014 Conclusion(s) from the 2006 General’s reports before in a Surgeon General’s Surgeon General’s Surgeon General’s report Disease 2006 (year) report (year) report Asthma “The limited existing data “The evidence is sufficient to infer — — — — a causal relationship between yield conflicting results parental smoking and ever having concerning the relationship asthma among children of school between passive smoke age.” (p. 14) exposure and pulmonary function changes in patients a with asthma.” (1984, p. 13) “The evidence is suggestive but not sufficient to infer a causal relationship between secondhand smoke exposure from parental smoking and the onset of childhood asthma.” (p. 14) “Chronic cough and phlegm are more “The evidence is sufficient to Chronic — — frequent in children whose parents — — infer a causal relationship respiratory smoke compared with children of between parental smoking and symptoms nonsmokers. The implications of cough, phlegm, wheeze, and (cough, phlegm, chronic respiratory symptoms for breathlessness among children wheeze, dyspnea, respiratory health as an adult are of school age.” (p. 14) etc.) unknown and deserve further study.” (1986, p. 13) “The evidence is sufficient to infer a causal relationship between secondhand smoke exposure from parental smoking and the onset of wheeze illnesses in early childhood.” (p. 14) Chapter 4 92

122 The Health Consequences of Smoking —50 Years of Progress Table 4.8 Continued Additional or updated Highest level conclusion(s) conclusion(s) from the First mention and finding(s) from subsequent Surgeon Conclusion(s) from the 2006 2012/2014 Surgeon in a Surgeon General’s General’s reports before Surgeon General’s report General’s report Disease report (year) 2006 (year) “The evidence is sufficient to infer a “The children of parents who — “The children of smoking Influenza, — causal relationship between smoke have an increased parents have an increased pneumonia, secondhand smoke exposure from frequency of a variety of acute prevalence of reported infections, and parental smoking and lower respiratory illnesses and respiratory symptoms, and acute respiratory illnesses in infants and infections, including chest have an increased frequency respiratory children.” (p. 14) illnesses before 2 years of age of bronchitis and pneumonia illnesses and physician-diagnosed early in life.” (1984, p. 13) bronchitis, tracheitis, and laryngitis, when compared with the children of nonsmokers.” (1986, p. 13) “The children of parents who smoke have an increased frequency of hospitalization for bronchitis and pneumonia during the first year of life when compared with the children of nonsmokers.” (1986, p. 13) “The children of smoking “The children of parents who “The evidence is sufficient to — Lung smoke have small differences in parents appear to have infer a causal relationship — growth tests of pulmonary function when measurable between maternal smoking and compared with the children of but small differences in tests during pregnancy and persistent pulmonary of pulmonary function when nonsmokers. Although this adverse effects on lung function function decrement is insufficient to cause compared with children of across childhood.” (p. 14) nonsmoking parents. The symptoms, the possibility that it may increase susceptibility to significance of this finding to “The evidence is sufficient to chronic obstructive pulmonary the future development of infer a causal relationship disease with exposure lung disease is between exposure to secondhand unknown.” (1984, p. 13) to other agents in adult life, e.g., smoke after birth and a lower [ ] active smoking or sic level of lung function during occupational exposures, needs childhood.” (p. 14) investigation.” (1986, p. 13) 93 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

123 Surgeon General’s Report Table 4.8 Continued Additional or updated First mention and finding(s) Highest level conclusion(s) from conclusion(s) from the Conclusion(s) from the 2006 in a Surgeon General’s subsequent Surgeon General’s 2012/2014 Surgeon Disease Surgeon General’s report report (year) reports before 2006 (year) General’s report Middle ear disease “A number of studies report that “The evidence is sufficient to infer — — and chronic middle ear effusions are — — a causal relationship between adenotonsillectomy more common in young children parental smoking and middle ear whose parents smoke than in disease in children, including children of nonsmoking acute and recurrent otitis media parents.” (1986, p. 14) and chronic middle ear effusion.” (p. 14) “The evidence is suggestive but not sufficient to infer a causal relationship between parental smoking and the natural history of middle ear effusion.” (p. 14) “The evidence is inadequate to infer the presence or absence of a causal relationship between parental smoking and an increase in the risk of adenoidectomy or tonsillectomy among children.” (p. 14) “The evidence is inadequate to — — — Atopy — — infer the presence or absence of a causal relationship between parental smoking and the risk of immunoglobulin E-mediated allergy in their children.” (p. 14) Note: TB = tuberculosis. or a General conclusion without specification of outcome in children adults. Chapter 4 94

124 The Health Consequences of Smoking —50 Years of Progress Conclusions from Surgeon General’s report on exposure to secondhand smoke and Table 4.9 respiratory effects in adults Additional or updated conclusion(s) from Highest level conclusion(s) First mention and the 2012/2014 from subsequent Surgeon Conclusion(s) from the 2006 finding(s) in a Surgeon Surgeon General’s General’s reports before 2006 Surgeon General’s report Disease General’s report (year) report (year) “The limited existing data “The evidence is suggestive but — Asthma — — — not sufficient to infer a causal yield conflicting results relationship between secondhand concerning the smoke exposure and adult-onset relationship between asthma.” (p. 16) passive smoke exposure and pulmonary function “The evidence is suggestive but changes in patients with not sufficient to infer a causal a asthma.” (1984, p. 13) relationship between secondhand smoke exposure and a worsening of asthma control.” (p. 16) — “The evidence is suggestive but Chronic respiratory — — — — — not sufficient to infer a causal symptoms (cough, relationship between secondhand phlegm, wheeze, smoke exposure and acute dyspnea, etc.) respiratory symptoms including cough, wheeze, chest tightness, and difficulty breathing among persons with asthma.” (p. 15) “The evidence is suggestive but not sufficient to infer a causal relationship between secondhand smoke exposure and acute respiratory symptoms including cough, wheeze, chest tightness, and difficulty breathing among healthy persons.” (p. 15) “The evidence is suggestive but not sufficient to infer a causal relationship between secondhand smoke exposure and chronic respiratory symptoms.” (p. 15) 95 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

125 Surgeon General’s Report Continued Table 4.9 Additional or updated conclusion(s) from First mention and finding(s) Highest level conclusion(s) from the 2012/2014 in a Surgeon General’s report subsequent Surgeon General’s Surgeon General’s Conclusion(s) from the 2006 Disease (year) reports before 2006 (year) report Surgeon General’s report — “The evidence is suggestive but not — “Healthy adults exposed to Chronic — — sufficient to infer a causal relationship environmental tobacco obstructive between secondhand smoke exposure smoke may have small pulmonary and risk for chronic obstructive changes on pulmonary disease pulmonary disease.” (p. 16) function testing, but are unlikely to experience “The evidence is inadequate to infer clinically significant deficits in the presence or absence of a causal pulmonary function as a result of relationship between secondhand exposure to environmental tobacco smoke exposure and morbidity in pp. 13–14) smoke alone.” (1986, persons with chronic obstructive pulmonary disease.” (p. 16) “The evidence is suggestive but not — “...some studies suggest that “Other components of tobacco sufficient to infer a causal — high levels of involuntary smoke, such as particulate Lung function relationship between short-term [tobacco] smoke exposure might matter and the oxides of (Formerly secondhand smoke exposure and an produce small changes in nitrogen, designated as acute decline in lung function in pulmonary function in normal have been shown in various chronic bronchitis; persons with asthma.” (p. 16) subjects. ... Two studies have concentrations to affect emphysema; reported differences in measures adversely animal “The evidence is inadequate to infer chronic obstructive of lung function in older pulmonary...function. The the presence or absence of a causal lung disease; populations between subjects extent of the contributions of relationship between short-term chronic obstructive chronically exposed to these substances to illness secondhand smoke exposure and an bronchopulmonary involuntary in humans exposed to the acute decline in lung function in disease) concentrations present in an healthy persons.” (p. 16) smoking and those who were atmosphere contaminated with not. This difference was not tobacco smoke is not presently “The evidence is suggestive but not found in a younger and known.” (1972, pp. 7–8) sufficient to infer a causal possibly less exposed relationship between chronic population.” (1984, p. 13) secondhand smoke exposure and a small decrement in lung function in the general population.” (p. 16) “The evidence is inadequate to infer the presence or absence of a causal relationship between chronic secondhand smoke exposure and an accelerated decline in lung function.” (p. 16) Chapter 4 96

126 The Health Consequences of Smoking —50 Years of Progress Table 4.9 Continued Additional or updated conclusion(s) from Highest level conclusion(s) the 2012/2014 First mention and finding(s) from subsequent Surgeon Conclusion(s) from the 2006 Surgeon General’s in a Surgeon General’s report General’s reports before Surgeon General’s report report Disease (year) 2006 (year) — “The evidence is sufficient to infer a “An atmosphere Odor and “The main effects of the — causal relationship between contaminated with tobacco irritants present in ETS occur irritation secondhand smoke exposure and odor smoke can contribute to in the conjunctiva of the eyes annoyance.” (p. 15) the discomfort of many and the mucous membranes of individuals.” (1972, p. 7) the nose, throat, and lower “The evidence is sufficient to infer a respiratory tract. These irritant “Cigarette smoke in the air causal relationship between effects are a frequent cause of can produce an increase in secondhand smoke exposure and complaints about poor air both subjective and objective nasal irritation.” (p. 15) quality due to environmental measures of eye tobacco smoke.” (1986, p. 252) irritation.” (1984, p. 13) “The evidence is suggestive but not sufficient to conclude that persons with nasal allergies or a history of respiratory illnesses are more susceptible to developing nasal irritation from secondhand smoke exposure.” (p. 15) “The evidence Tuberculosis is inadequate to infer the presence or absence of a — — — causal relationship between — — — exposure to secondhand smoke and the risk of tuberculosis infection.” (2014, Chapter 7) “The evidence is inadequate to infer the presence or absence of a causal relationship between exposure to secondhand smoke and the risk of tuberculosis disease.” (2014, Chapter 7) ETS = environmental tobacco smoke. Note: a General conclusion without specification of outcome in children or adults. 97 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

127 Surgeon General’s Report Table 4.10 Conclusions from Surgeon General’s report on exposure to secondhand smoke amd reproductive and developmental effects Highest level conclusion(s) from Additional or updated subsequent Surgeon conclusion(s) from the First mention and finding(s) Conclusion(s) from the 2006 General’s reports before 2012/2014 Surgeon in a Surgeon General’s Surgeon General’s report Disease 2006 (year) General’s report report (year) “The evidence is inadequate to infer the — — — Child physical — — — presence or absence of a causal relationship and cognitive between exposure to secondhand smoke and development cognitive functioning among children.” (p. 13) “The evidence is inadequate to infer the presence or absence of a causal relationship between exposure to secondhand smoke and behavioral problems among children.” (p. 13) “The evidence is inadequate to infer the presence or absence of a causal relationship between exposure to secondhand smoke and children’s height/growth.” (p. 13) “The evidence is inadequate to infer the presence — — Congenital — — — — or absence of a causal relationship between malformations exposure to secondhand smoke and congenital malformations.” (p. 13) Fertility “The evidence is inadequate to infer the presence — — — — — — or absence of a causal relationship between maternal exposure to secondhand smoke and female fertility or fecundability. No data were found on paternal exposure to secondhand smoke and male fertility or fecundability.” (p. 13) “Studies of ETS exposure and “The evidence is inadequate to infer the — — Fetal the risks for delay in conception, — — death, presence or absence of a causal relationship spontaneous abortion, and between exposure to secondhand smoke and stillbirths, perinatal mortality are few, and neonatal mortality.” (p. 13) and the results are inconsistent.” infant (2001, p. 372) mortality “The evidence is sufficient to infer a causal — — — Sudden infant — — — relationship between exposure to secondhand death syndrome smoke and sudden infant death syndrome.” (p. 13) (SIDS) Chapter 4 98

128 The Health Consequences of Smoking —50 Years of Progress Table 4.10 Continued Additional or updated Highest level conclusion(s) from conclusion(s) from the 2012/2014 subsequent Surgeon First mention and finding(s) Surgeon General’s Conclusion(s) from the 2006 Surgeon General’s reports in a Surgeon General’s report General’s report Disease before 2006 (year) report (year) “The evidence is sufficient to infer a causal “...maternal exposure to — — Infant — — birth relationship between maternal exposure to ETS appears to be causally secondhand smoke during pregnancy and a associated with detrimental weight small reduction in birth weight.” (p. 13) effects on fetal growth.” (2001, p. 364) “The evidence is inadequate to infer the — — — Pregnancy presence or absence of a causal — — — complications relationship between maternal exposure to secondhand smoke during pregnancy and spontaneous abortion.” (p. 13) “The evidence is suggestive but not sufficient to infer a causal relationship between maternal exposure to secondhand smoke during pregnancy and preterm delivery.” (p. 13) ETS = environmental tobacco Note: smoke. 99 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

129 Surgeon General’s Report Table 4.11 Conclusions reached by the Advisory Committee to the Surgeon General in 1964 Lung Cancer “Cigarette smoking is causally related to lung cancer in men; the magnitude of the effect of cigarette smoking far outweighs all other factors. The data for women, though less extensive, point in the same direction. The risk of developing lung cancer increases with duration of smoking and the number of cigarettes smoked per day, and is diminished by discontinuing smoking. The risk of developing cancer of the lung for the combined group of pipe smokers, cigar smokers, and pipe and cigar smokers, is greater than for nonsmokers, but much less than for cigarette smokers. The data are insufficient to warrant a conclusion for each group individually.” (p. 196) Oral Cancer “The causal relationship of the smoking of pipes to the development of cancer of the lip appears to be established. Although there are suggestions of relationships between cancer of other specific sites of the oral cavity and the several forms of tobacco use, their causal implications cannot at present be stated.” (pp. 204–5) Cancer of the Larynx “Evaluation of the evidence leads to the judgment that cigarette smoking is a significant factor in the causation of laryngeal cancer in the male.” (p. 212) Cancer of the Esophagus “The evidence on the tobacco-esophageal cancer relationship supports the belief that an association exists. However, the data are not adequate to decide whether the relationship is causal.” (p. 218) Cancer of the Urinary Bladder “Available data suggest an association between cigarette smoking and urinary bladder cancer in the male but are not sufficient to support a judgment on the causal significance of this association.” (p. 225) Stomach Cancer “No relationship has been established between tobacco use and stomach cancer.” (p. 229) Non-Neoplastic Respiratory Diseases, Particularly Chronic Bronchitis and Pulmonary Emphysema “Cigarette smoking is the most important of the causes of chronic bronchitis in the United States, and increases the risk of dying from chronic bronchitis. A relationship exists between pulmonary emphysema and cigarette smoking but it has not been established that the relationship is causal. The smoking of cigarettes is associated with an increased risk of dying from pulmonary emphysema. For the bulk of the population of the United States, the importance of cigarette smoking as a cause of chronic bronchopulmonary disease is much greater than that of atmospheric pollution or occupational exposures. Cough, sputum production, or the two combined are consistently more frequent among cigarette smokers than among non- smokers. Cigarette smoking is associated with a reduction in ventilatory function. Among males, cigarette smokers have a greater prevalence of breathlessness than non-smokers. Cigarette smoking does not appear to cause asthma. Although death certification shows that cigarette smokers have a moderately increased risk of death from influenza and pneumonia, an association of cigarette smoking and infectious diseases is not otherwise substantiated.” (p. 302) Chapter 4 100

130 The Health Consequences of Smoking —50 Years of Progress Table 4.11 Continued Cardiovascular Disease “Male cigarette smokers have a higher death rate from coronary artery disease than non-smoking males, but it is not clear that the association has causal significance.” (p. 327) Peptic Ulcer “Epidemiological studies indicate an association between cigarette smoking and peptic ulcer which is greater for gastric than for duodenal ulcer.” (p. 340) Tobacco Amblyopia “Tobacco amblyopia (dimness of vision unexplained by an organic lesion) has been related to pipe and cigar smoking by clinical impressions. The association has not been substantiated by epidemiological or experimental studies.” (p. 342) Cirrhosis of the Liver “Increased mortality of smokers from cirrhosis of the liver has been shown in the prospective studies. The data are not sufficient to support a direct or causal association.” (p. 342) Maternal Smoking and Infant Birth Weight “Women who smoke cigarettes during pregnancy tend to have babies of lower birth weight. Information is lacking on the mechanism by which this decrease in birth weight is produced. It is not known whether this decrease in birth weight has any influence on the biological fitness of the newborn.” (p. 343) Smoking and Accidents “Smoking is associated with accidental deaths from fires in the home. No conclusive information is available on the effects of smoking on traffic accidents.” (p. 345) Morphological Constitution of Smokers “The available evidence suggests the existence of some morphological differences between smokers and non-smokers, but is too meager to permit a conclusion.” (p. 387) “The overwhelming evidence points to the conclusion that smoking—its beginning, habituation, and occasional discontinuation— is to a large extent psychologically and socially determined. This does not rule out physiological factors, especially in respect to habituation, nor the existence of predisposing constitutional or hereditary factors.” (p. 377) Source: U.S. Department of Health, Education, and Welfare 1964. 101 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

131 Surgeon General’s Report Table 4.12 Conclusions from previous Surgeon General’s reports related to smoking and all-cause mortality Year Conclusion 1964 “Cigarette smoking is associated with a 70 percent increase in the age specific death rates of males, and to a lesser extent with increased death rates of females. The total number of excess deaths causally related to cigarette smoking in the U.S. population cannot be accurately estimated. In view of the continuing and mounting evidence from many sources, it is the judgment of the Committee that cigarette smoking contributes substantially to mortality from certain specific diseases and to the overall death rate.” (p. 31) 1967 “1. Cigarette smokers have substantially higher rates of death and disability than their nonsmoking counterparts in the population. This means that cigarette smokers tend to die at earlier ages and experience more days of disability than comparable nonsmokers. 2. A substantial portion of earlier deaths and excess disability would not have occurred if those affected had never smoked.” (p. 3) 1968 “Previous findings reported in 1967 indicate that cigarette smoking is associated with an increase in overall (supplement mortality and morbidity and leads to a substantial excess of deaths in those people who smoke.” (p. 3) to 1967) “1. Overall mortality rates for cigarette smokers are about 70 percent higher than those for nonsmokers. 1978 2. Overall mortality risk increases with the amount smoked. For the two-pack-a-day cigarette smoker, the risk of premature death is approximately twice that of the nonsmoker. 3. Overall mortality ratios of smokers compared to nonsmokers are highest at earlier ages and decline with increasing age. For cigarette smokers, the risk of premature death is twice that of nonsmokers at age 40. 4. Overall mortality ratios are higher for those who begin smoking at a young age compared to those who begin later. For those who begin smoking before the age of 15, the risk of premature death is about 86 percent higher than that for nonsmokers.” (pp. 44–5) 1979 “1. The overall mortality ratio for all male current cigarette smokers, irrespective of quantity, is about 1.7 (70 percent excess) compared to nonsmokers. 2. Mortality ratios increase with amount smoked. The two-pack-a-day male smoker has a mortality ratio of 2.0 compared to nonsmokers. 3. Overall mortality ratios are directly proportional to the duration of cigarette smoking. The longer one smokes, the greater the risk of dying. 4. Overall mortality ratios are higher for those who initiated their cigarette smoking at younger ages compared to those who began smoking later. 5. Overall mortality ratios are higher among cigarette smokers who inhale than among those who do not.” (p. 1-10) 1980 “1. The mortality ratio for women who smoke cigarettes is about 1.2 or 1.3. 2. Mortality ratios for women increase with the amount smoked. In the largest prospective study the mortality ratio was 1.63 for the two-pack-a-day smoker as compared to nonsmokers. 3. Mortality ratios are generally proportional to the duration of cigarette smoking; the longer a woman smokes, the greater the excess risk of dying. 4. Mortality ratios tend to be higher for those women who begin smoking at a young age as compared to those who begin smoking later.” (p. 6) 1989 “Smoking is responsible for more than one of every six deaths in the United States. Smoking remains the single most important preventable cause of death in our society.” (p. 11) Chapter 4 102

132 The Health Consequences of Smoking —50 Years of Progress Table 4.12 Continued Year Conclusion 2001 “1. Cigarette smoking plays a major role in the mortality of U.S. women. 2. The excess risk for death from all causes among current smokers compared with persons who have never smoked increases with both the number of years of smoking and the number of cigarettes smoked per day.” (p. 12) 2004 “There have been more than 12 million premature deaths attributable to smoking since the first published Surgeon General’s report on smoking and health in 1964. Smoking remains the leading preventable cause of premature death in the United States.” (p. 30) a 2006 “Secondhand smoke causes premature death and disease in children and in adults who do not smoke.” (p. 11) a Exposure to secondhand smoke. 103 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

133 Surgeon General’s Report References . Cancer Facts & Figures 2009 Surveillance, Epidemiology, and End Results (SEER) American Cancer Society. Program (www.seer.cancer.gov) SEER*Stat Database: Atlanta (GA): American Cancer Society, 2009. . Cancer Facts & Figures 2013 American Cancer Society. Mortality – All COD, Aggregated With State, Total Atlanta (GA): American Cancer Society, 2013. U.S. (1969-2010) , National Cancer Institute, DCCPS, Surveil- Baron JA, Byers T, Greenberg ER, Cummings KM, Swan son M. Cigarette smoking in women with cancers of the lance Research Program, Surveillance Systems Branch, breast and reproductive organs. Journal of the National released April 2013. Underlying mortality data provided Cancer Institute 1986;77(3):677–80. by NCHS (www.cdc.gov/nchs). Cutler D, Deaton A, Lleras-Muney A. The determinants U. S. Burden of Disease Collaborators. The state of U.S. health, 1990–2010: burden of diseases, injuries, and 2006; Journal of Economic Perspectives of mortality. - 20(3):97–120. risk factors. JAMA: the Journal of the American Medi cal Association 2013;310(6):591–608. Doll R, Hill AB. The mortality of doctors in relation to British their smoking habits; a preliminary report. U.S. Department of Health and Human Services. The Medical Journal 1954;1(4877):1451–5. Health Consequences of Smoking for Women. A Report Hammond EC, Horn D. The relationship between human . Washington: U.S. Department of the Surgeon General of Health and Human Services, Public Health Service, smoking habits and death rates: a follow-up study of Office of the Assistant Secretary for Health, Office on 187,766 men. Journal of the American Medical Asso- ciation 1954;155(15):1316–28. Smoking and Health, 1980. Hoyert DL, Anderson RN. Age-adjusted death rates: trend U.S. Department of Health and Human Services. The data based on the year 2000 standard population. Health Consequences of Smoking: Cancer. A Report of 2001;49(9):1–6. National Vital Statistics Reports the Surgeon General . Rockville (MD): U.S. Department Kannel WB, Dawber TR, Kagan A, Revotskie N, Stokes of Health and Human Services, Public Health Service, Office on Smoking and Health, 1982. DHHS Publica- J 3rd. Factors of risk in the development of coronary heart disease—six year follow-up experience. The tion No. (PHS) 82-50179. 1961; Framingham Study. Annals of Internal Medicine U.S. Department of Health and Human Services. The Health Consequences of Smoking: Cardiovascular Dis 55:33–50. - ease. A Report of the Surgeon General - Kim V, Criner GJ. Chronic bronchitis and chronic obstruc . Rockville (MD): tive pulmonary disease. American Journal of Respira- U.S. Department of Health and Human Services, Public 2013;187(3):228–37. Health Service, Office on Smoking and Health, 1983. tory and Critical Care Medicine DHHS Publication No. (PHS) 84-50204. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adair- U.S. Department of Health and Human Services. Rohani H, Amann M, Anderson HR, Andrews KG, Aryee The Health Consequences of Smoking: Chronic Obstruc M, et al. A comparative risk assessment of burden of - disease and injury attributable to 67 risk factors and . tive Lung Disease. A Report of the Surgeon General - risk factor clusters in 21 regions, 1990–2010: a system Rockville (MD): U.S. Department of Health and Human atic analysis for the Global Burden of Disease Study Services, Public Health Service, Office on Smoking and 2010. Lancet 2012;380(9859):2224–60. Health, 1984. DHHS Publication No. (PHS) 84-50205. The Emperor of All Maladies: A Biography of Mukherjee S. U.S. Department of Health and Human Services. The Cancer . New York: Scribner, 2010. Health Consequences of Involuntary Smoking: A National Center for Health Statistics. Chronic obstructive Report of the Surgeon General . Rockville (MD): U.S. http://www.cdc.gov/copd/>; pulmonary disease, 2012; < Department of Health and Human Services, Public accessed: March 1, 2013. Health Service, Centers for Disease Control, Center Health, United National Center for Health Statistics. - for Health Promotion and Education, Office on Smok States, 2012: With Special Feature on Emergency ing and Health, 1986. DHHS Publication No. (CDC) Care . Hyattsville (MD): U.S. Department of Health and 87-8398. Human Services, Centers for Disease Control and Pre- U.S. Department of Health and Human Services. Reduc- vention, National Center for Health Statistics, 2013. ing the Health Consequences of Smoking: 25 Years of Petty TL. The history of COPD. International Journal of . Rockville Progress. A Report of the Surgeon General 2006;1(1): Chronic Obstructive Pulmonary Disease (MD): U.S. Department of Health and Human Services, 3–14. Chapter 4 104

134 The Health Consequences of Smoking —50 Years of Progress Pre- U.S. Department of Health and Human Services. Public Health Service, Centers for Disease Control, venting Tobacco Use Among Youth and Young Adults: National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, . Atlanta (GA): U.S. A Report of the Surgeon General Department of Health and Human Services, Centers 1989. DHHS Publication No. (CDC) 89-8411. The U.S. Department of Health and Human Services. for Disease Control and Prevention, National Center Health Benefits of Smoking Cessation: A Report of the for Chronic Disease Prevention and Health Promotion, Surgeon General . Rockville (MD): U.S. Department Office on Smoking and Health, 2012. Smok- of Health and Human Services, Centers for Disease U.S. Department of Health, Education, and Welfare. ing and Health: Report of the Advisory Committee to Control, National Center for Chronic Disease Preven - . the Surgeon General of the Public Health Service tion and Health Promotion, Office on Smoking and Health, 1990. DHHS Publication No. (CDC) 90-8416. Washington: U.S. Department of Health, Education, and Welfare, Public Health Service, Center for Disease U.S. Department of Health and Human Services. Prevent- ing Tobacco Use Among Young People: A Report of the Control, 1964. PHS Publication No. 1103. U.S. Department of Health, Education, and Welfare. . Atlanta (GA): U.S. Department of Surgeon General The Health Consequences of Smoking. A Public Health Health and Human Services, Public Health Service, Service Review: 1967 . Washington: U.S. Department of Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Pro Health, Education, and Welfare, Public Health Service, - Health Services and Mental Health Administration, motion, Office on Smoking and Health, 1994. Oral 1967. PHS Publication No. 1696. U.S. Department of Health and Human Services. Health in America: A Report of the Surgeon General . U.S. Department of Health, Education, and Welfare. The Rockville (MD): U.S. Department of Health and Human Health Consequences of Smoking. 1968 Supplement - . Washington: to the 1967 Public Health Service Review Services, National Institutes of Health, National Insti U.S. Department of Health, Education, and Welfare, tute of Dental and Craniofacial Research, 2000. Women U.S. Department of Health and Human Services. Public Health Service, 1968. DHEW Publication No. and Smoking. A Report of the Surgeon General . Rock- 1696 (Supplement). The ville (MD): U.S. Department of Health and Human U.S. Department of Health, Education, and Welfare. Health Consequences of Smoking. 1969 Supplement Services, Public Health Service, Office of the Surgeon to the 1967 Public Health Service Review . Washington: General, 2001. The U.S. Department of Health and Human Services. U.S. Department of Health, Education, and Welfare, Health Consequences of Smoking: A Report of the Sur - Public Health Service, 1969. DHEW Publication No. . Atlanta (GA): U.S. Department of Health geon General 1696-2. U.S. Department of Health, Education, and Welfare. and Human Services, Centers for Disease Control and The - Health Consequences of Smoking. A Report of the Sur Prevention, National Center for Chronic Disease Pre - geon General: 1972 vention and Health Promotion, Office on Smoking and . Washington: U.S. Department of Health, Education, and Welfare, Public Health Service, Health, 2004. U.S. Department of Health and Human Services. The Health Services and Mental Health Administration, Health Consequences of Involuntary Exposure to 1972. DHEW Publication No. (HSM) 72-7516. . U.S. Department of Health, Education, and Welfare. The Tobacco Smoke: A Report of the Surgeon General Atlanta (GA): U.S. Department of Health and Human - Health Consequences of Smoking. A Report of the Sur Services, Centers for Disease Control and Prevention, . Washington: U.S. Department of geon General, 1973 Health, Education, and Welfare, Public Health Service, Coordinating Center for Health Promotion, National Health Services and Mental Health Administration, - Center for Chronic Disease Prevention and Health Pro motion, Office on Smoking and Health, 2006. 1973. DHEW Publication No. (HSM) 73-8704. U.S. Department of Health and Human Services. How The U.S. Department of Health, Education, and Welfare. Tobacco Smoke Causes Disease—The Biology and - Health Consequences of Smoking. A Report of the Sur Behavioral Basis for Smoking-Attributable Disease: - geon General, 1977–1978 . Washington: U.S. Depart A Report of the Surgeon General . Atlanta (GA): U.S. ment of Health, Education, and Welfare, Office of the Department of Health and Human Services, Centers Assistant Secretary for Health, Office on Smoking and for Disease Control and Prevention, National Center Health, 1979. DHEW Publication No. (CDC) 79-50065. for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2010. 105 Advances in Knowledge of the Health Consequences of Smoking: From 1964–2014

135 Surgeon General’s Report U.S. Department of Health, Education, and Welfare. Wellcome Trust Case Control Consortium. Genome- - Smoking and Health. A Report of the Surgeon wide association study of 14,000 cases of seven com . Washington: U.S. Department of Health, General mon diseases and 3,000 shared controls. Nature 2007; Education, and Welfare, Office of the Assistant Secre - 447(7145):661–78. tary for Health, Office on Smoking and Health, 1979. World Health Organization. WHO mortality database DHEW Publication No. (PHS) 79-50066. 2012, 2012; ; accessed: August 5, 2013. American Journal of Human of GWAS discovery. 2012;90(1):7–24. Genetics Chapter 4 106

136 Chapter 5 Nicotine 109 Introduction Toxicokinetics and Acute Toxicity of Nicotine 109 109 Toxicokinetics 111 Acute Toxicity of Nicotine Summary 112 Pathophysiology of Nicotine Addiction 112 112 Summary of Evidence from the 2010 Surgeon General’s Report Trajectory of Addiction 113 Health Consequences of Nicotine Exposure 113 Cancer 113 Genotoxicity 114 Effects of Nicotine on Carcinogenic Pathways 114 Experimental Animal Studies for Carcinogenicity 115 115 Human Studies Summary 116 Cardiovascular Diseases 116 Immune Function and Related Disorders 117 Reproductive Health Outcomes 117 Fetal Growth Restriction 118 Preterm Delivery 118 Stillbirth, Perinatal Mortality, and Sudden Infant Death Syndrome 118 Congenital Malformations 119 Summary 119 Lung Development 119 Summary 120 Cognitive Function 120 Cognitive Function and the Nicotinic Acetylcholine Receptor System 121 Effects of Nicotine on Cognitive Function in Healthy Adult Smokers and Nonsmokers 121 Critical Periods of Exposure in the Nervous System 121 Other Vulnerable Populations 123 Tobacco Industry Influence 124 Summary 125 Evidence Summary 125 Conclusions 126 Implications 126 127 References 107

137

138 The Health Consequences of Smoking —50 Years of Progress Introduction disease. A number of new noncombustible products (e.g., Nicotine has been addressed in multiple previous reports of the Surgeon General. Most notably, the 1988 electronic cigarettes) have been marketed by the tobacco The Health Consequences of Surgeon General’s report, industry and other manufacturers that provide nicotine concluded that cigarettes Smoking: Nicotine Addiction, through the oral and inhaled routes. Use of such products and tobacco products are addicting and that “Nicotine is is projected by some to take an increasing market share - the drug in tobacco that causes addiction” (U.S. Depart over the next decade (Citigroup Global Markets 2011). Additionally, nicotine replacement therapy (NRT) remains ment of Health and Human Services [USDHHS] 1988, a mainstay of cessation aids and many former smokers - How Tobacco Smoke Causes Dis p. 9). The 2010 report, addressed the mechanisms by which nicotine leads ease, may remain on such therapy for periods of time longer than recommended and approved by the U.S. Food and to addiction, providing full coverage of pharmacology, - Drug Administration (West and Russell 1985; Hajek et al. genetic factors, manifestations of addiction, and epidemi 1988; Hughes et al. 1991; Hughes 1998). ologic aspects (USDHHS 2010). The topic of trajectories Given the possibility of increasing exposure of the of addiction and relapse was also addressed and further population to nicotine obtained from products other than covered in regard to adolescents and young adults in the conventional cigarettes, this chapter considers the acute 2012 report, Preventing Tobacco Use Among Youth and (USDHHS 2012). and longer-term adverse consequences of nicotine. The Young Adults chapter also provides background for the consideration This chapter addresses the acute toxicity of nicotine of future policy directions in Chapter 16, “A Vision for and the effects of longer-term exposure on reproductive outcomes, lung growth and development, neurocogni - Ending the Epidemic: A Society Free of Tobacco-Related Death and Disease.” tive function and cognitive decline, psychiatric morbid - ity, immune function, cancer risk, and cardiovascular Toxicokinetics and Acute Toxicity of Nicotine Nicotine is the major chemical component respon - Toxicokinetics sible for addiction in tobacco products (USDHHS 1988; Stolerman and Jarvis 1995; Royal College of Physicians of Nicotine, 3-(1-methyl-2-pyrrolidinyl) pyridine, is a - London 2000; Balfour 2004). The risk for nicotine addic volatile alkaloid with a molecular weight of 162.23. The tion depends on the dose of nicotine delivered and the - absorption and elimination via renal excretion of nico way it is delivered; the potential for addiction increases tine are highly dependent on pH. At a high (alkaline) pH, 1 with the dose delivery rate, the rate of absorption, and = 8.5) is in the non-ionized state, which nicotine (pKa the attained concentration of nicotine (Henningfield passes more easily through lipoprotein membranes than and Keenan 1993; de Wit and Zacny 1995; Stitzer and the ionized (charged) state (Stratton et al. 2001). Nico- - de Wit 1998). For an in-depth discussion of the pharma tine in its un-ionized state can be readily absorbed across cokinetics of nicotine as related to addiction, see the phar - the epithelium of the lung, the oral mucosa, and the nose, macokinetics section of Chapter 4 in the 2010 Surgeon and through the skin. Nicotine in tobacco smoke inhaled General’s report (USDHHS 2010). Similarly, the toxicity - because of the large sur into the lung is rapidly absorbed caused by nicotine is dependent on dose, dose duration face area of the alveoli and small airways and the dissolu - - and frequency, route of exposure, formulation of the nico tion of nicotine in the fluid coating the lung’s epithelial tine product, and interpersonal variability as addressed in layer, which has a physiological pH that facilitates absorp- the 2010 report. This section discusses the toxicokinetics tion. Similarly, nicotine from oral tobacco products that and the acute toxicity of nicotine. 1 The logarithmic measure of the acid disassociation constant, which represents the pH of a solution in which half of the acid molecules are ionized. 109 Nicotine

139 Surgeon General’s Report owitz 2001). The time course of nicotine accumulation in have an alkaline pH is readily absorbed through the oral mucosa, but more gradually than via the lungs. Nicotine - the brain and other body organs, and the resultant phar can be well absorbed in the small intestine, because of its macologic effects, are highly dependent on route and rate more alkaline pH and large surface area. However, nicotine of dosing. The lag time between a puff on a cigarette until nicotine reaches the brain is 10–20 seconds (Henningfield is poorly absorbed from the stomach, because its acidic environment results in greater ionized nicotine. In addi - and Keenan 1993; de Wit and Zachy 1995; Stitzer and de Wit 1998; Rose et al. 1999). tion, unlike ingestion, nicotine’s bioavailability is greater More than 80% of nicotine absorbed into the body through the lung or through the oral mucosa, because undergoes metabolism in the liver, principally by CYP2A6, nicotine reaches the systemic circulation before passing through the liver where it is metabolized (first-pass metab UDP-glucuronosyltransferase, and flavin-containing - olism). Arterial concentrations of nicotine from smoking monooxygenase (Cashman et al. 1992; Park et al. 1993; Benowitz and Jacob 1994; Benowitz et al. 2009). Several are higher than venous concentrations (Figure 5.1). Across metabolites of nicotine reach the central nervous sys - studies, the ratios of arterial to venous concentration tem (CNS) after acute administration of nicotine (Crooks range from 2.3–10 (Henningfield et al. 1993; Gourlay and Benowitz 1997; Rose et al. 1999). Less than 5% of nicotine and Dwoskin 1997). Nornicotine is both a metabolite of is protein-bound in the plasma (Benowitz et al. 1982). It nicotine and a minor tobacco alkaloid. Researchers have distributes extensively to body tissues, including the liver, - observed similar behavioral effects from nicotine and nor nicotine. However, because nornicotine is present only as kidney, spleen, lung, and brain and also accumulates in gastric juice and saliva, breast milk, skeletal muscle, and a minor metabolite, it is unclear whether it has significant fetal serum and amniotic fluid (Dahlstrom et al. 1990; pharmacologic or toxicologic effects in nicotine users. - Breese et al. 1997; Perry et al. 1999; Dempsey and Ben Less data are available on cotinine, a major metabolite of Figure 5.1 Venous blood concentrations of nicotine over time for various nicotine delivery systems 20 15 NicoDerm CQ (nicotine patch) Cigarette (nicotine delivery, 1–2 mg) 10 Polacrilex (nicotine gum; nicotine delivery, 4 mg) Oral snuff 5 Nicotrol (nicotine patch) Plasma nicotine concentration (ng/mL) Nasal spray (nicotine delivery, 1 mg) 0 100 110 120 08 09 0 0203 -10 01 04 050607 Minutes Adapted from Fant et al. 1999 with permission from Elsevier, ©1999. Source: mg Note: = milligrams; = nanograms per milliliter. ng/mL 110 Chapter 5

140 The Health Consequences of Smoking —50 Years of Progress Table 5.1 Animal studies on acute toxicity of nicotine Study objective/endpoint Route of exposure Species tested Study Mice, rabbits i.p. Determine LD50 Larson et al. 1945 i.p. Determine LD50 Hicks and Sinclair, 1947 Rats i.p. Determine LD50 Rats Yamamoto et al. 1966 Lazutka et al. 1969 Determine LD16, LD50, LD100 Oral, inhalation Mice, rats Stalhandske and Slanina 1970 i.p. Mice Determine difference in response to LD50 between young and old rats Mice i.p. Determine LD50 by mouse strain, age, gender; ED50 of Tepper et al. 1979 onset of tremor Okamoto et al. 1992 Rats i.p. Determine time to convulsions Okamoto et al. 1994 Rats i.p. Determine difference in response to LD50 between young and old rats Yuen et al. 1995 Rats Oral (water) Examine acute hepatotoxicity ED50 LD16 = intraperitoneal; i.p. = median dose where 50% of sample subjects achieve a predefined endpoint; Note: = dosage of a LD50 = dosage of a given drug required to kill 50% of a test population; given drug required to kill 16% of a test population; = dosage of a given drug required to kill 100% of a test population. LD100 nicotine (Benowitz and Jacob 1994; Keenan et al. 1994). Acute toxicologic data on nicotine is limited. Such For discussion of the pharmacodynamics of nicotine in information comes from three sources: (1) animal studies, the brain, see the section on “Pathophysiology of Nicotine (2) studies investigating nicotine as a therapeutic agent Addiction” in Chapter 4 of the 2010 Surgeon General’s (including NRT), and (3) poisonings involving nicotine. A report (USDHHS 2010). few acute toxicological studies performed on animals are available (Table 5.1). These studies contribute basic LD50 (dose causing 50% lethality) values primarily in rats and Hicks and Sinclair 1947; Yama- mice (Larson et al. 1945; Acute Toxicity of Nicotine Tepper et al. 1979), Lazutka et al. 1969; moto et al. 1966; as well as examining the effects of age and gender, and - Nicotine exerts its effects via stimulation of the nico endpoints other than lethality, such as hepatotoxicity and tinic acetylcholine receptors (nAChRs), which are located time to convulsions. However, the studies available do not in the CNS, at interganglionic junctions of the autonomic adequately characterize acute toxicity. Studies investigat- nervous system, and on target organs throughout the body ing nicotine as a therapeutic agent in humans are limited as part of the parasympathetic autonomic nervous system - in predicting the acute toxicity of nicotine. These stud (USDHHS 2010). As a result of the global expression of ies are better at documenting adverse effects rather than - these receptors, their stimulation causes broad physi overt toxicity, as the doses administered are chosen, in ologic effects. Although the nicotine intoxication syn- part, because they are considered subtoxic. Mild adverse drome is not fully characterized, symptoms of mild acute effects, as defined by the World Health Organization’s toxicity might include nausea and vomiting, progressing (WHO’s) Collaborating Center for International Drug with increased exposure to cholinergic syndrome, which - Monitoring (WHO 1972), of nicotine given as pharma includes diarrhea, increased salivation, increased respira- cologic treatment for nicotine addiction have been com - tory secretions, and bradycardia. Severe poisonings can Barrueco et al. 2005 monly reported ( ). Studies examining progress further to seizures and respiratory depression. nicotine’s potential role to treat ulcerative colitis using Countering the development of acute toxicity is the rela - nicotine patches or enemas provide similar findings with - tively rapid development of tolerance with repeated expo regard to adverse effects (Nikfar et al. 2010; Lunney and sure (Benowitz et al. 1987; Okamoto et al. 1992). Leong 2012). 111 Nicotine

141 Surgeon General’s Report mg for adults, with the fatal dose for youth expected to be Numerous poisonings have been documented in the literature since the use of nicotine as a pesticide became lower, but not determined specifically. A study by Lazutka widespread in the early part of the twentieth century. and colleagues (1969), in a Russian language publication, These studies describe patients exposed to doses associ - is commonly cited in support of these figures. However, Lazutka and colleagues make no such estimation. Fur - ated with toxicity via one or more routes of exposure, and ther, a systematic literature search was performed using a resulting predicted clinical course of acute toxicity as noted previously in this section. However, the literature OVID MEDLINE for nicotine (focusing on ‘toxicity’ n = - also notes exceptions, including a rapid progression to 744 and ‘poisonings’ n = 134), as well as a search of data bases such as the Hazardous Substances Data Bank and near fatal symptoms after a relatively low exposure to a piece of 2 milligrams (mg) nicotine gum that was chewed Haz-Map using Toxnet; however, no study was located as a briefly and discarded – never swallowed ( Mensch and source for an estimate of the dose that is fatal to humans Holden 1984), as well as a patient receiving a relatively and the figure of 50–60 mg is poorly documented. large dose, 240 mg nicotine, in an accidental subcutane- Brady et ous administration that proved to be nonfatal ( al. 1979). In both instances, the affected persons were Summary active cigarette smokers. The case report involving the 2 - mg gum did not specifically document nicotine intoxica In its un-ionized state, nicotine readily enters the tion; rather, a clinical diagnosis was made. Yet, despite the body, regardless of the mode of administration. It has abundance of case reports, it appears that there has not known acute toxicity, reflecting its pharmacologic activ - - been a systematic assessment of the literature to charac ity. There is a potential for poisoning from ingestion of terize the dose-response relationship. Finally, the human nicotine-containing products. oral fatal dose is commonly reported to be between 50–60 Pathophysiology of Nicotine Addiction carry a low risk of establishing addiction and are generally Summary of Evidence from the substantially easier to discontinue than tobacco products 2010 Surgeon General’s Report - (Henningfield et al. 2011; WHO 2012). Conversely, ciga rettes have been researched, designed, and manufactured Dependence on nicotine is characterized by both the to increase the likelihood that initiation will lead to depen - persistence of a drug-taking behavior and the emergence dence and difficulty achieving cessation due to contents of withdrawal symptoms upon the abrupt cessation of and emissions in addition to nicotine (e.g., acetaldehyde, nicotine administration (Wikler 1973; Levine 1974; Stew- ammonia compounds, and menthol); design features that art et al. 1984; Ludwig 1986; O’Brien et al. 1990; Hughes may increase free-base nicotine and produce larger puffs and Hatsukami 1992; Koob et al. 1993; Markou et al. 1993, (filter-tip ventilation); and other factors that reduce the 1998; American Psychiatric Association 1994; Kenny and concerns for smokers and increase the attractiveness of - Markou 2001; USDHHS 2010). Therefore, both the neu the products (USDHHS 2010, 2012). rosubstrates (brain structures, pathways, and systems) nAChRs are ligand-gated ion channels composed of - mediating the reinforcing effects of acute administra five membrane-spanning subunits that combine to form a tion of nicotine and those mediating the nicotine with - functional receptor (Lindstrom et al. 1996; Role and Berg drawal syndrome are relevant to nicotine addiction. The 1996; Albuquerque et al. 1997; Lèna and Changeux 1998, physiological systems that develop adaptations to repeated 1999; Dani 2000; Gotti et al. 2006). As a result of actions nicotine administration, and lead to the emergence of at the nAChR sites, nicotine stimulates the release of withdrawal signs on cessation of nicotine administration, most neurotransmitters throughout the brain (Araujo et are likely to intersect with systems that mediate the acute al. 1988; Toide and Arima 1989; McGehee and Role 1995; effects of nicotine (Markou et al. 1998; Kenny and Markou Gray et al. 1996; Role and Berg 1996; Wilkie et al. 1996; - 2001). That is, nicotine addiction develops as a neurobio Albuquerque et al. 1997; Alkondon et al. 1997; Kenny et logic adaptation to chronic nicotine exposure. However, al. 2000; Grady et al. 2001). Therefore, various transmitter all forms of nicotine delivery do not pose an equal risk systems are likely to be involved in the rewarding effects in establishing or maintaining nicotine addiction. NRT of nicotine and in the adaptations that occur in response medicines, which are designed to minimize addiction risk, 112 Chapter 5

142 —50 Years of Progress The Health Consequences of Smoking Hildebrand et al. 1998, 1999; Carboni et al. 2000). Nor to chronic exposure to nicotine, which give rise to depen - - dence and to withdrawal responses. adrenergic and serotonergic systems may also play a role - in withdrawal. Decreased glutamate transmission appears The positive reinforcing aspects of nicotine addic to mediate the affective aspects of withdrawal, but GABA tion primarily results from the release of dopamine in the ventral tegmental area region of the brain (Grenhoff transmission does not appear to change with withdrawal. et al. 1986; Nisell et al. 1994a,b, 1997; Pidoplichko et al. 1997; Watkins et al. 2000; Picciotto and Corrigall 2002; - Balfour 2004). Nicotine stimulates nAChRs on glutama Trajectory of Addiction tergic terminals that release glutamate, an excitatory neurotransmitter, which results in an increased release of The addiction caused by the nicotine in tobacco - dopamine in the nucleus accumbens and the frontal cor - smoke is critical in the transition of smokers from experi tex (Gray et al. 1996; Gioanni et al. 1999; Fu et al. 2000; mentation to sustained smoking and, subsequently, in Grillner and Svensson 2000; Mansvelder and McGehee - the maintenance of smoking for the majority of smok 2000; Reid et al. 2000). Nicotine also excites nAChRs on ers who want to quit (USDHHS 2010, 2012). Substantial gamma-aminobutyric acid (GABA)-releasing terminals longitudinal research has shown that smoking typically (Schilström et al. 1998; Mansvelder and McGehee 2000). begins with experimental use of cigarettes and that the Thus, levels of GABA, an inhibitory neurotransmitter, are transition to regular smoking can occur relatively quickly, also increased by nicotine. However, the interplay between with the smoking of as few as 100 cigarettes (USDHHS the quick desensitization of nAChRs on the GABA neuron 2012). Longitudinal studies show that there are indi- and the higher doses of nicotine required to desensitize vidual trajectories of smoking as tracked by the index nAChRs on the glutamate neuron results in an increase in of numbers of cigarettes smoked daily. These trajecto - dopamine levels (Schilström et al. 1998; Mansvelder and - ries are variable, with some smokers quickly progress - McGehee 2000). A critical role may also be played by nic ing to regular smoking and others doing so more slowly otine-induced increases in norepinephrine transmission, (USDHHS 2010, 2012). Research is in progress on the pos- although the role of this transmitter system in nicotine sible role of genetic factors in determining the trajectory of dependence has not been investigated as extensively as nicotine use. that of the dopamine, glutamate, and GABA systems. The The 2012 Surgeon General’s report makes clear roles of endocannabinoids, serotonin, and endogenous - that addiction can begin in people who begin experi opiates in nicotine addiction are less certain. For further menting with tobacco use during their teenage years discussion of neurosubstrates, see ‘Neurosubstrates of (USDHHS 2012). Although the phenotype of addiction - Nicotine Reinforcement’ in the “Pathophysiology of Nico - is not so well defined as with adults, symptoms of with tine Addiction” section of Chapter 4 in the 2010 Surgeon drawal occur among youth who become regular smokers. General’s report. As documented in that report, the longitudinal studies The neurophysiological mechanisms associated with show several different patterns of smoking uptake, with withdrawal symptoms may vary with the type of symp- some young people rapidly escalating their use to a typical toms experienced (e.g., somatic vs. affective). The nAChRs pattern of regular use and others doing so more slowly. appear to be involved in both the somatic and affective Some adolescents may be able to smoke on an experi- components of nicotine withdrawal. Decreased mesolim - mental or intermittent basis without becoming addicted bic dopaminergic transmission seems to mediate various (USDHHS 2012). aspects of the withdrawal syndrome (Fung et al. 1996; Health Consequences of Nicotine Exposure promote cancer based on experimental studies that have Cancer limitations in replicating human exposure and on mecha- (Lee et al. nistic studies, but human evidence is lacking Nicotine is a highly bioactive compound with effects 2005, 2012; Dasgupta and Chellappan 2006; Zheng et al. ranging from being a natural pesticide in tobacco leaves 2007; Catassi et al. 2008; Chen et al. 2008b, 2010; Egleton to causing addiction in tobacco users. For cancer, there et al. 2008) . Nicotinic receptors are found not only in the is some biological basis for proposing that nicotine may 113 Nicotine

143 Surgeon General’s Report antioxidants or nicotinic receptor inhibitors (Ginzkey et brain but throughout the body; for example, in muscle, al. 2012). One study found that cotinine, and not nicotine, (Improgo et al. 2011; lung, endothelia, kidney, and skin Cardinale et al. 2012; Hurst et al. 2013) was genotoxic by the bacterial genotoxicity luminescence . These receptors trigger a number of cellular pathways involved in carcino test, but another was null for the Ames assay and SCE - (Doolittle et al. 1995; Yim and Hee 1995) . Some induction genesis. The presence of nicotinic cholinergic receptors reports indicate that nicotine can lead to the formation throughout the normal lung and in lung tumors has been (Schuller 2009; Improgo et al. 2011) of DNA adducts using the ultrasensitive technique accel- . well documented This section reviews the current literature that relates to erator mass spectroscopy (Cheng et al. 2003). Although the hypothesis that nicotine may contribute to the carci cigarette smoke is highly genotoxic, a comparison of Ames - nogenic process. The evidence comes from experimental mutagenicity for cigarette smoke from cigarettes with dif - cell culture and animal studies, and from human studies fering nicotine yields did not indicate different mutagenic - including epidemiologic. potential, suggesting that there was no additional contri The potential for nicotine to contribute to the risk of bution by nicotine (Chen et al. 2008a). incident cancer or cancer recurrence is important due to Effects of Nicotine on Carcinogenic Pathways the number of smokers who have quit by using NRT, some of whom use NRT for long durations to remain smoking There are numerous studies that focus on lung cells abstinent, and other smokers who switch to alternate and cells from other organs relating to nicotine exposure. sources of nicotine (e.g., e-cigarettes or smokeless tobacco A wide range of effects has been reported in cellular sys - products). Although using NRT or other noncombusted tems, including at doses similar to those in the blood of sources of nicotine is different than smoking in evident smokers (Cardinale et al. 2012). The presence of nAChRs ways, the possibility of increased risk in long-term users throughout the lung has been well documented via pro - compared to those who use such products only briefly tein studies and demonstration of the presence of tran - - for cessation merits consideration. Thus, when con scripts for both normal tissues and lung tumors (Improgo templating the available evidence, coming largely from et al. 2011). These receptors are important for triggering laboratory experiments, the following questions need to (Schuller 2009). many signaling pathways in lung cells be addressed: (1) What is the cancer risk for those who In lung cells, nicotine has been shown to: (1) inhibit quit smoking but use long-term NRT or other sources of apoptosis including apoptosis induced by chemotherapy nicotine compared with those who continue to smoke? (2) , (Maneckjee and Minna 1990, 1994; Cardinale et al. 2012) What is the cancer risk of a lifetime pattern of repeatedly which involves the PI3-K-Akt pathway and attendant posi - quitting with NRT and relapsing, but smoking fewer life- tive effects on Bcl-2 and negative effects on BAD and BAX time cigarettes overall? (3) What is the cancer risk of long- ; (2) (West et al. 2003; Jin et al. 2004; Xin and Deng 2005) term NRT use without relapse to smoking or sustained affect proliferation by stimulating the release of epidermal switching to a noncombusted nicotine source compared growth factor and, therefore, activation of the Ras-Raf- with long-term abstinence without NRT or other source of ERK cascade (Dasgupta and Chellappan 2006; Carlisle et nicotine or relapse to smoking? This section will address al. 2007; Paleari et al. 2008); and (3) stimulating fibro- these questions. nectin production activating ERK, PI3-K, mTOR, and the (Dasgupta et al. 2006). Also, there δ β expression of PPAR- / Genotoxicity is evidence that nicotine may promote metastases because of stimulation of cell motility and migration, loss of adhe- There are mixed data for a genotoxic effect of nico - sion, and inducing the transition of a well-differentiated tine. Most studies were negative that used the Ames assay epithelial cell to a highly invasive carcinoma via epithe - (including urine of rats exposed to nicotine), chromosomal lial-mesenchymal transition (Catassi et al. 2008; Egleton aberration and sister chromatid exchange (SCE) assays in et al. 2008; Cardinale et al. 2012). Chinese hamster ovary cells, and the bacterial genotoxicity An important consideration for cancer survival and luminescence test (Mizusaki et al. 1977; Riebe et al. 1982; metastasis is angiogenesis. A variety of mechanisms are . In contrast, Doolittle et al. 1991, 1995; Yim and Hee 1995) - stimulated by nicotine to promote angiogenesis; for exam two studies were positive for chromosomal aberration and ple, promoting endothelial cell migration, proliferation, SCEs (Riebe and Westphal 1983; Trivedi et al. 1990), one survival, and tube formation (Cardinale et al. 2012; Lee was positive for micronuclei formation that was inhibited . Nicotine directly binding to nicotinic and Cooke 2012) with antioxidants (Argentin and Cicchetti 2004) , one was + – receptors in endothelial cells induced endothelial cell /POLA mutation Escherichia coli POLA positive for an tube migration by stimulating VEGF in lung cancer cells (Riebe et al. 1982), and another using nasal mucosal assay (Conklin et al. 2002; Heeschen et al. 2002; Li and Wang cells was positive by the Comet assay, which is inhibited by 114 Chapter 5

144 —50 Years of Progress The Health Consequences of Smoking 2006; Ng et al. 2007) . Lower doses of nicotine in vitro and metastases (Heeschen et al. 2001; Jarzynka et al. 2006; induce endothelial cell proliferation, while higher doses Al-Wadei et al. 2009; Davis et al. 2009). . These effects also Other studies have investigated the potential for induce cytotoxicity (Villablanca 1998) - nicotine to promote the carcinogenic effects of 4-(methyl- occur via stimulation of nicotinic receptors in the endo thelia. The angiogenic effect of nicotine involves MAPK, nitrosamino)-1-(3-pyridyl)-1-butanone (NNK). Maier and (Heeschen et al. 2002) . - colleagues (2011) conducted a series of studies to deter PI3K/Akt, and NF- κ B activation Angiogenesis has been shown in a variety of tumor cells, mine if nicotine would promote the carcinogenesis such as breast, colon, and lung, implanted in a chick cho induced by NNK. The results were null. The investigators - used several models, including a crossed A/J and C57BL/6 (Heeschen et rioallantoic membrane, and other systems mouse, a mutant animal model prone to develop k-Ras al. 2002; Mousa and Mousa 2006). lung tumors, and a syngeneic lung-cancer graft model Limited research has addressed whether the nico- with NNK-transformed lung cancer cells. The dosing of tine in tobacco smoke somehow alters the toxicity of nicotine, albeit by drinking water, was specifically intended tobacco smoke. Chen and colleagues (2008a) conducted - to be similar to the levels human smokers receive when various in vitro studies comparing cigarettes with differ ing amounts of nicotine, and where nicotine was added using NRT. In a separate study, Murphy and colleagues (2011) studied the A/J mouse and did not find a difference back to the condensate. They found that nicotine attenu - ated the cytotoxicity of cigarette smoke through inhibi- in tumorigenesis whether the nicotine was given before or after NNK, compared to NNK alone. tion of apoptotic pathways, increased proliferative activity, and increased cell survival. There was no evidence of an In summary, the findings of animal studies do not support the hypothesis that nicotine is a complete car - effect on the gap junction intracellular communication, cinogen. It is a tumor promoter in some experimental which is considered to be a marker of tumor promo- tion effects. models, although not for tobacco-specific nitrosamines. Studies examining other classes of tobacco smoke car - cinogens (e.g., polycyclic aromatic hydrocarbons) would Experimental Animal Studies for Carcinogenicity need to be performed to better define the potential cancer Several studies in experimental animals also did risk inferred from animal studies. not indicate that nicotine alone is tumorigenic (Martin et al. 1979; Waldum et al. 1996; Hecht 2003; Murphy et Human Studies al. 2011). These studies have included the inhalational Very little human data are on human cancer risk route of exposure, fetal exposure, and exposure through maternal milk. The only exception to the null findings is relating to nicotine. The Lung Health Study is the only the report of nicotine inducing sarcomas in the muscle study that provides information about long-term users of . This study was not designed to (Murray et al. 2009) NRT or uterus of exposed A/J mice; other tumors, including directly examine nicotine’s potential cancer risk. It was those in the lung, were not observed in that same study (Galitovskiy et al. 2012) a 5-year randomized trial to assess the effects of smok - . The A/J mouse model is used ing cessation and reduction on chronic lung disease and for assessing the carcinogenic effects of cigarette smoke lung function. Among 5,887 persons initially enrolled, the in inducing lung tumors. However, the lack of nicotine researchers continued to follow them for an additional 7 induction of lung tumors may be related to the dose and Study participants were offered NRT years (n = 3,220). route of exposure. without consideration of randomization or study design. As a tumor promoter, nicotine has been reported Although they were encouraged to use NRT for only 6 to increase the frequency of tumors induced by agents months, many continued to use it long term. A total of 75 such as nicotine-derived nitrosamine ketone, and lung cancers were diagnosed among smokers and quitters , 7,12-dimethylbenz(a)anthracene (Chen and Squier 1990) of the extended surveillance group, but the use of NRT (Gurkalo and N- methyl-N ́-nitro- N -nitrosoguanidine was not associated with lung cancer (or other cancers). Volfson 1982), and N -[4-(5-nitro-2-furyl)-2-thiazolyl] for - A major limitation was the short follow-up period of only (LaVoie et al. 1985) . Other studies showed that mamide 7 additional years. Notwithstanding the limitations, this nicotine had no effect in promoting tumors related to other study at least does not indicate a strong role for nicotine in (Habs and Schmahl 1984) and had an anti- N -nitrosamines promoting carcinogenesis in humans, and clearly the risk, tumor effect in some cases (Zeller and Berger 1989) . In a if any, is less than continued smoking. different tumor promotion model, nicotine induced lung Another approach to examining whether nicotine tumors in hamsters in the presence of hyperoxia (Schuller could contribute to carcinogenesis would be to consider et al. 1995). In addition, studies using cancer xenograft its delivery in the context of long-term smokeless tobacco models have shown that nicotine promotes tumor growth 115 Nicotine

145 Surgeon General’s Report chromosome 1 (i.e., CHRNB2 ), chromosome 8 ( CHRNB3 ), use. Smokeless tobacco products used in northern Euro- CHRNA4 and chromosome 20 ( pean countries appear to result in a substantially reduced ), it is not known how much exposure to many tobacco smoke carcinogens, because of an effect there is, if any, by these genes on carcinogen- smokeless tobacco does not undergo combustion. Epi- (Thorgeirs- esis independently of an effect on tobacco use demiologic studies of smokeless tobacco indicate that it son et al. 2008; Bierut 2009, 2010; Johnson et al. 2010; Li et al. 2011; Russo et al. 2011; Sarginson et al. 2011; increases the risk of oral cavity, esophageal and pancreatic cancers, (IARC 2012) at least for some forms of smokeless Sorice et al. 2011; Timofeeva et al. 2011; Wassenaar et al. tobacco. The associated risks for these sites are less than 2011; Broms et al. 2012; Budulac et al. 2012; Kapoor et al. the risk of these cancers from smoking; however, high 2012). Separately, there are data on CYP2A6 genetics and nicotine metabolism that show associations with smoking rates of oral cancers in India and Sudan are attributable to behavior, nicotine levels, and lung cancer risk (Wassenaar (Accortt et al. 2005; the use of smokeless tobacco products et al. 2011; Gold and Lerman 2012; Liu et al. 2013; Zhu Boffetta et al. 2005, 2008; Luo et al. 2007) . The risks for et al. 2013). many cancers commonly associated with smoking are not elevated by long-term smokeless tobacco use. This pattern Summary of risk suggests that in humans nicotine may not have a strong tumor promoting effect. Further, although levels There is insufficient data to conclude that nicotine of nicotine are similar for smokers and smokeless tobacco causes or contributes to cancer in humans, but there is users, the risk of cancer of the oral cavity, esophagus, and evidence showing possible oral, esophageal, or pancreatic - pancreas is less for the smokeless tobacco users, indicat cancer risks. Additionally, there is substantial experi- ing that exposures other than nicotine contribute to the mental evidence indicating that nicotine is bioactive for a cancer process. This conclusion, however, needs to be number of carcinogenic mechanisms in experimental sys - tempered by the possibility that there may be a different tems. Although in vitro data are suggestive of relevant bio - risk due to route of exposure, because smokeless tobacco logical activity, this is not supported overall by the most use leads to nicotine exposure via the oral mucosa and recent experimental animal studies. In humans, there has ingestion, while smoking results in inhalation exposure. been limited research and only one relatively short–term Risks inferred from smokeless tobacco studies may not follow-up study on nicotine and cancer. extend directly to inhalation exposures. There is some evidence that NRT can endogenously lead to the formation of the carcinogenic tobacco-specific Cardiovascular Diseases nitrosamines, NNK and N-nitrosonornicotine (NNN), at least in rats (Carmella et al. 1997) , which concep- The potential role of nicotine in atherogenesis and tually would increase cancer risk if the resultant dose in triggering acute coronary events has been discussed was similar to those which result from smoking or the extensively in the medical literature (USDHHS 2010) and use of smokeless tobacco. A smoking cessation study by reviewed in Chapter 8, “Cardiovascular Diseases,” of this Stepanov and colleagues (2009b) demonstrated that NNK volume. It is likely that the sympathomimetic effects of metabolites were not detectable in persons using NRT nicotine increase heart rate and myocardial contractility, . However, they did find intermittently (Hecht et al. 1999) increase coronary vascular resistance, and reduce insulin high levels of NNN similar to baseline smoking levels in - sensitivity, contributing to some extent to increasing car 13 of 34 participants using NRT gum or lozenges, and in diovascular risk in smokers. However, other mechanisms only 1 of 9 persons using the patch (Stepanov et al. 2009a). by which nicotine might contribute to atherogenesis Although these data indicate a potential cancer risk to have also been proposed (Lee and Cooke 2011). nAChRs NRT users, especially oral users, it is important to realize are found not only in neuronal and muscle cells but also that NNN is only one of the tobacco-specific nitrosamines in endothelial cells and immune cells. Nicotine has been in cigarette smoke. Thus, it will be important to quantify reported to induce the proliferation of vascular smooth the level of risk from long-term use of NRT or other non - muscle cells and migration of cells into blood vessels (Lee combusted sources of nicotine, particularly if long-term - and Cooke 2012). In apoE*deficient mouse models of ath nicotine use from sources other than smoking becomes erosclerosis, oral nicotine was shown to promote plaque more prevalent. Although there is a variety of evidence progression and neovascularization. The primary nicotine that nicotinic receptor polymorphisms play a role in lung receptor in endothelial cells is the alpha 7 homomeric cancer risk and in determining the amount of tobacco use, - nicotine receptor. In mice deficient in this receptor sub CHRNA3 , CHRNA4 , the genes on chromosome 15 (i.e., type, the effect of nicotine in augmenting angiogenesis , ), , and CHRNB3 , CHRNB2 CHRNB4 CHRNA6 CHRNA5 , 116 Chapter 5

146 —50 Years of Progress The Health Consequences of Smoking is blunted. Tolerance develops to many of the effects of can suppress cellular immunity both in vivo and in vitro. Nicotine suppresses the production of antibodies in B - nicotine with prolonged exposure, both in people and ani mals. Chronic oral administration of nicotine was shown cells, reduces the proliferation of T cells, and induces an anergy-like state where signaling via the T cell receptor to abolish the augmenting effect of nicotine on angiogenic is attenuated (Geng et al. 1995, 1996). These effects have et al. 2010). Thus, responses to limb ischemia (Konishi it is unclear whether the short-term effects of nicotine been linked to the impaired host defense response to bac - - teria and viruses in nicotine-treated animals. in enhancing angiogenesis persist with long-term expo In summary, as reviewed here and discussed in sure, as seen with users of tobacco or other nicotine- delivering products. more detail in Chapter 10, “Other Specific Outcomes,” A genomewide association study found an associa- there is compelling evidence that nicotine affects cellu - tion between a gene cluster on chromosome 15 and an lar immunity, either directly by interacting with nicotinic cholinoceptors or indirectly via its effects on the nervous increased risk of peripheral arterial obstructive disease (Thorgeirsson et al. 2008). Since this gene cluster is system. Whether these effects contribute to the overall adverse effects of cigarette smoke on immunity is less strongly associated with the level of nicotine dependence, well-understood. it is not clear whether the association indicates a direct role of nicotine in atherosclerosis. Reproductive Health Outcomes Immune Function and Related Pregnancy is accompanied by a complex series Disorders of maternal physiological adjustments to support fetal - growth and homeostasis. Basic characteristics of embryo Nicotine has both stimulatory and suppressive effects - logic and fetal development include cell growth, differenti on the immune system, and levels of nicotine, inferred ation, interaction, and migration. Teratogenic factors can from urine markers, have been linked with both induction disturb one or more of these processes, resulting in abnor - of and protection from immunologically mediated disease malities in fetal structure or function, including growth (Cloëz-Tayarani and Changeux 2007). Nicotine exerts its retardation, malformations due to abnormal growth or effects via pentameric nicotinic cholinoceptors that vary morphogenesis, and altered CNS performance (Hacker et in their alpha and beta subunit composition ( USDHHS al. 2010). In addition, there is a growing appreciation that 2010). Nicotine can act directly on cells, but in vivo it is teratogenic substances can have effects throughout the also a direct activator of the sympathetic nervous system, duration of pregnancy, and that those effects can be more which itself can have strong immune-regulatory effects. subtle than gross anatomic anomalies (Yaffe and Aranda Aged-smoke extracts that still contain all of the nicotine of - 2011). Thus, for women of reproductive age, a compre fresh smoke but lack reactive intermediates are much less hensive exploration of the known and potential harms of active in immune assays than freshly prepared, oxidant- the range of available tobacco products, all of which con - rich extracts (Laan et al. 2004; Bauer et al. 2008). Nicotine tain nicotine, is needed. The health effects of smoking and patches or mecamylamine (a full antagonist) or nicotine of components in tobacco smoke, including nicotine, on partial antagonists (e.g., varenicline), which are used as reproduction are reviewed in Chapter 9, “Reproductive - adjuncts in smoking cessation, are not immune-modula Outcomes.” A focused review of what is known about the tory in humans (Cahill et al. 2008), and snus (the nico- effects of nicotine on maternal and fetal health outcomes tine-rich low nitrosamine smokeless tobacco product that is presented here. is used widely in Sweden) does not replicate the effects of Cigarette use before and/or during pregnancy smoking. This interpretation is consistent with research remains a major cause of reduced fertility as well as mater - - with macrophages where the effects of smoking on immu nal, fetal, and infant morbidity and mortality (see Chapter nity were linked to oxidation (McMaster et al. 2008). 9) and over 400,000 live-born infants in the United States This highly contradictory literature is further rein - are exposed in utero to tobacco from maternal smoking forced by studies on human immune effector cells linked annually (Hamilton et al. 2012; Tong et al. in press). Con - to atherosclerosis where nicotine was found to stimulate, ditions causally associated with maternal prenatal smok- not suppress, dendritic cells as part of adaptive immune ing include preterm delivery and fetal growth restriction, responses (Aicher et al. 2003). However, a large body of placenta previa, placental abruption, sudden infant death evidence suggests that nicotine acting via the alpha 7 syndrome (SIDS), some congenital anomalies, ectopic subunit that contains neuronal nicotinic cholinoceptors 117 Nicotine

147 Surgeon General’s Report Preterm Delivery pregnancy, and reduced preeclampsia risk. Maternal pre- natal smoking has also been associated with stillbirth and Maternal smoking is associated with a 27% increase spontaneous abortion (USDHHS 2004). in the risk of preterm delivery (Shah and Bracken 2000) - Much of what can currently be inferred about nico and several studies have also found an increased risk of tine and reproductive health comes from studies compar - preterm delivery in smokeless tobacco users (Gupta and ing the effects of prenatal smokeless tobacco use with the Sreevidya 2004; Baba et al. 2012; England et al. 2013). In effects of prenatal smoking because smokeless tobacco Sweden, snus use and smoking during pregnancy were products do not expose users to products of combustion, both associated with increased risks of preterm birth, and - but all contain nicotine. In addition, some smokeless prod the magnitudes of the associations were similar (Baba et ucts such as Swedish snus contain lower levels of certain al. 2012). Together, these studies provide evidence that toxicants when compared with conventional smokeless - nicotine increases the risk of preterm delivery. The poten tobacco products (Stepanov et al. 2008). This differen - - tial roles of nicotine and products of combustion in pre tial exposure between snus and other smokeless products term delivery are discussed in detail in Chapter 9. allows researchers to study the health effects of smokeless tobacco while reducing the likelihood that adverse out- Stillbirth, Perinatal Mortality, and Sudden Infant comes will mistakenly be attributed to nicotine. Studies of Death Syndrome health outcomes in randomized trials of nicotine therapy Studies of stillbirth have also been conducted in pregnant women offer additional insights. among smokeless tobacco users. Studies in India and Swe - den showed an increased risk of stillbirth in women using Fetal Growth Restriction smokeless tobacco (Krishna 1978; Gupta and Subramoney It has been believed for decades that in utero expo - - Wikström et al. 2010). In the study conducted in Swe 2006; sure to cigarette smoke causes fetal growth restriction den, when antenatal bleeding and small-for-gestational- through nicotine-mediated vasoconstriction of utero- age deliveries were excluded, the smoking-related risk of placental vessels (Lambers and Clark 1996). However, this stillbirth was markedly reduced although the elevated risk hypothesis has been questioned because it is not likely for snus users remained the same. These findings suggest that nicotine’s vasoconstrictive effects are sufficient to that the mechanisms underlying the associations between overcome placental circulatory reserve (Benowitz and smoking and stillbirth and between smokeless tobacco use Dempsey 2004). Further evidence against a vasoconstric- and stillbirth both involve nicotine, but other factors may - tive mechanism comes from studies of pregnancy out also contribute to increased risk in smokers (Wikström et comes in smokeless tobacco users. These studies have al. 2010). consistently demonstrated only modest contributions of - The effects of nicotine on the brainstem, cardio smokeless tobacco to reduced infant birth weight (Eng - pulmonary integration, fetal and neonatal responses to land et al. 2003, 2012; Gupta and Sreevidya 2004; Juarez hypoxic stress, and arousal in early infancy are reviewed and Merlo 2013). Results from a population-based study in Chapter 9. For example, it has been hypothesized that in Sweden conducted from 1999–2010 suggest that tobacco-related changes in autonomic function and/ - smokeless tobacco (snus) use increases the risk for deliv or arousal could increase the risk of SIDS, although a ering a small-for-gestational-age infant (for term births, mechanistic pathway has not been established (American adjusted odds ratio [AOR] = 1.21; 95% confidence inter - Academy of Pediatrics Task Force on Sudden Infant Death val [CI], 1.02–1.43). The effect was smaller in magnitude Syndrome 2011). Studies of human infants have shown an for smokeless tobacco than for cigarette smoking (AOR = association between prenatal exposure to cigarette smoke 2.27; 95% CI, 2.62–2.91) (Baba et al. 2013). In a trial of and impaired recovery from hypoxia in preterm infants 250 women randomized to a nicotine patch (15 mg) or (Schneider et al. 2008) and an association with impaired to placebo for 11 weeks, the researchers found that there - arousal patterns that correlates with cotinine levels (Rich was no difference between the two arms in quit rates or in ardson et al. 2009). Maternal prenatal cigarette use has saliva cotinine, but birth weight was significantly higher also been associated with increased obstructive apnea and in the NRT group (186g [95% CI, 35–336g]), possibly due decreased arousal in response to apnea events in infants to reduced cigarette smoking and exposure to products of - (Sawnani et al. 2004). Additional data suggest that mater - combustion. Taken together, these studies support a mod nal prenatal smokeless tobacco use also increases infants’ est role for nicotine in fetal growth restriction. risk of apnea, of a similar magnitude to that seen with maternal smoking (Gunnerbeck et al. 2011). 118 Chapter 5

148 —50 Years of Progress The Health Consequences of Smoking Extensive animal research has generated plausible Lung Development and generalizable models to explain how nicotine could increase the risk of SIDS and perinatal mortality (Slotkin The 2004 Surgeon General’s report concluded and Seidler 1988); these models are reviewed in Chapter that “the evidence is sufficient to infer a causal relation- 9. In one such model, the fetal/infant protective response ship between maternal smoking during pregnancy and a to hypoxia is impaired. During parturition, the fetus reduction of lung function in infants” (p. 27). This conclu - normally experiences significant hypoxia, but is able to sion was based on epidemiologic studies that consistently respond with a massive release of catecholamines from demonstrated an inverse dose-response relationship the adrenal medulla ( Lagercrantz and Bistolelli 1977; between the number of cigarettes smoked per day during Lagercrantz and Slotkin 1986) in order to maintain blood pregnancy by the mother and the level of lung function . In the fetus and neonate, the flow to the brain and heart and pulmonary compliance in the newborn. The 2006 adrenal gland responds directly to hypoxia, independent of Surgeon General’s report expanded the conclusions of his direct mechanism persists until central reflexes, and t the 2004 report to address the duration of effects after chromaffin cells differentiate after the development of infancy: “The evidence is sufficient to infer a causal rela- - (Slotkin 1998). However, pre splanchnic nerve function tionship between maternal smoking during pregnancy natal nicotine exposure in rat models causes immature and persistent adverse effects on lung function across chromaffin cells in the adrenal gland to differentiate pre - childhood” (p. 399). The report further concluded that the maturely, resulting in loss of the normal direct stimula- - (referring to sec “evidence shows that parental smoking tion of the adrenal gland by hypoxia, complete absence of ondhand smoke exposure and maternal smoking during catecholamine release, and impaired cardiac response in reduces the maximum achieved level (of lung pregnancy) the presence of hypoxia (Slotkin 1998). The effect is a tem- although not to a degree (on average) that would growth), porary loss of a critical protective response to hypoxia and, impair individuals” (p. 400). “Nonetheless, a reduced theoretically, is accompanied by a temporary increased peak level increases the risk for future chronic lung dis - mortality risk (Slotkin 1998). ease, and there is heterogeneity of the effect so that some exposed children may have a much greater reduction Congenital Malformations than the mean” (p. 400). This section considers studies on the mechanisms underlying the relationship between In this report, the evidence was determined to be maternal smoking and the infant or child’s lung develop- - sufficient to support a causal relationship between mater ment and function and the potential role of nicotine in of nal smoking and orofacial clefts, and to be suggestive these mechanisms. a causal relationship for clubfoot, cryptorchidism, gas- Human lung development begins in the embryonic troschisis, and some types of congenital heart defects (see stage and extends through early adulthood. During fetal Chapter 9). The 2010 Surgeon General’s report examined lung growth, structural and vascular development take the biological basis for increased risk of congenital defects - place and major airway branching and mesenchymal pro in infants of mothers who smoke and specifically consid - liferation are complete by the end of the second trimester. ered the potential role of nicotine (USDHHS 2010); this Alveolarization (marked by septation and multiplication report updates that review. A number of potential mecha- of alveoli) begins in the third trimester of pregnancy and nisms were cited by which nicotine having crossed the multiplication of alveolar number continues to 2–3 years placenta, could contribute to defects. of age, when lungs reach the full adult quantity of approxi - - mately 300 million alveoli. Alveolar size and surface, how Summary ever, increase until after adolescence as the lungs grow The evidence supports the hypothesis that nicotine (Joshi and Kotecha 2007). Lung development is tightly plays a key role in mediating adverse effects of smoking on regulated, and intrauterine and postnatal environmen- reproductive health, including preterm delivery and still - - tal factors can interfere with this complicated set of pro birth. Smoking has been linked to diverse adverse health cesses. The alveolar phase of development is particularly outcomes for the developing fetus and experimental sensitive to late-pregnancy and postnatal insults (Harding research and pharmacologic understanding indicate that and Maritz 2012). nicotine specifically has a role in causing them. 119 Nicotine

149 Surgeon General’s Report As reviewed in previous Surgeon General’s reports, an increase in nAChRs in the lungs (Sekhon et al. 1999; the clinical and epidemiologic data strongly support that Fu et al. 2009), increased collagen deposition in airway walls, and increases in the numbers of alveolar type II and maternal smoking in pregnancy has lasting effects on lung development. Studies of infants exposed in utero to neuroendocrine cells (Sekhon et al. 1999, 2002). Coincid - ing with these changes are alterations in smooth muscle tobacco smoke show evidence of impaired lung function with reduced respiratory compliance, forced expiratory and vascular tension, perhaps explaining the effects of flow, and tidal breathing ratio, consistent with impaired maternal smoking on infant lung function (Stocks and airways development (Hanrahan et al. 1992; Tager et al. Dezateux 2003). Other hypothesized mechanisms through 1995; Lødrup Carlsen et al. 1997; Stocks and Dezateux which nicotine could affect lung development include premature onset of cell 2003). Maternal prenatal smoking has also been associ- decreased rep- differentiation and lication and impaired alveolar development—resulting - ated with impaired lung function with reduced small air way flow rates in school-age offspring (Cunningham et al. from altered expression or deposition of elastin (Pierce and Nguyen 2002; Stocks and Dezateux 2003). - 1994, 1995), even after adjustment for the offspring’s cur Together, these findings indicate that nicotine is a rent and past exposure to secondhand smoke (Gilliland et primary mediator of many of the adverse effects of mater - al. 2000), and with deficits in measures of airflow among nal smoking on fetal lung development. However, the adolescents, especially among those with a history of early- mechanisms involved remain incompletely understood. onset asthma (Gilliland et al. 2003). There is also evidence to suggest that exposure to prenatal tobacco smoke could result in an acceleration of lung aging and an increased Summary susceptibility to obstructive lung disease, lasting beyond Studies reviewed in the 2004 and 2006 Surgeon childhood (Maritz and Harding 2011). General’s reports and subsequently published data col - Numerous studies using animal models have been lectively show that prenatal tobacco exposure affects the conducted to develop a better understanding of the structure and function of the lung; these effects may mechanisms through which maternal smoking affects have consequences that last into childhood beyond, as fetal and infant lungs. These studies are summarized in lung development and growth are completed. Studies in several review articles (Stocks and Dezateux 2003; Maritz rhesus monkeys, which have lung development similar - 2008; Maritz and Harding 2011). Studies in primates spe - to that of humans, and in other animal models consis cifically examining nicotine exposure have demonstrated tently show that nicotine may be the primary mediator of decreased lung size and volume, increased type I and type many of the adverse effects of maternal smoking on fetal III collagens, decreased elastin in the lung parenchyma, lung development. increased alveolar volume, and increased airway wall area (Sekhon et al. 1999, 2001, 2002). Animal studies have also demonstrated decreased expiratory flow rates and Cognitive Function increased pulmonary resistance with nicotine exposure, similar to findings in human studies (Hanrahan et al. Researchers have suggested that smoking may have 1992; Cunningham et al. 1995; Tager et al. 1995; Dezateux cognition-enhancing properties (West 1993; Heishman et al. 1999). Primate studies further suggest that nicotine- et al. 2010), such as improvements in sustained atten - induced changes in airway wall thickness or stiffness could tion, reaction time, and memory (Evans and Drobes be an underlying cause of altered lung function (Pierce - 2008; Poorthuis et al. 2009; Heishman et al. 2010). Ini and Nguyen 2002; Sekhon et al. 2002). Finally, nicotine tial reports of improved cognitive function were based on - exposure in fetal lambs has been associated with acceler empirical evidence from smokers (Bell et al. 1999); thus, - ated maturation of lung acini and reduced proximal air these observations could reflect the mitigation of cogni- way conductance (Sandberg et al. 2004), hyperreactive tive impairment from nicotine withdrawal, enhancement proximal airways, and changes in proximal airway wall of smokers’ cognitive function independent of nicotine’s composition with associated defects in airflow (Sandberg effects on withdrawal symptoms, or both. Interest in the et al. 2011). effects of nicotine on cognition has since expanded to At the molecular level, nicotine crosses the placenta include healthy nonsmokers and individuals with under - and binds nAChRs in numerous locations in the lung, - lying neuropsychiatric conditions accompanied by cogni including bronchial epithelial cells, alveolar epithelial tive deficits. Concurrently, there is a growing awareness cells, neuroendocrine cells, submucosal glands, airway and of the potential harms of nicotine exposure during certain vascular smooth muscle cells, fibroblasts, and pulmonary vulnerable stages of brain development, such as during macrophages (Pierce and Nguyen 2002). Nicotine admin - fetal and adolescent growth (Dwyer et al. 2008; Duncan et istration to pregnant rhesus monkeys is associated with 120 Chapter 5

150 —50 Years of Progress The Health Consequences of Smoking al. 2009; Poorthuis et al. 2009; Bublitz and Stroud 2012; Whether there are direct effects of nicotine on cog- nitive function (positive or negative) in nonabstinent Goriounova and Mansvelder 2012). This section reviews smokers and in healthy nonsmoking adults is less clear. In the evidence on the effects of nicotine on cognitive func - a recent meta-analysis of double-blind, placebo-controlled tion in general (in smokers and nonsmokers), and in studies examining the acute effects of nicotine (adminis - potentially vulnerable populations. - tered mainly as nicotine replacement product) on cogni tive function in nonsmokers and smokers abstinent for 2 Cognitive Function and the Nicotinic Acetylcholine Receptor System hours or less, nicotine was found to result in cognitive enhancement in six of nine performance domains: fine Underlying the purported connection between nico- motor, alerting attention-accuracy and response time tine and cognitive enhancement is the role of nAChRs in (RT), orienting attention and RT, short-term episodic attention, learning, memory, and cortical plasticity (Wal- memory accuracy, and working memory RT (Heishman - lace and Bertrand 2013). nAChRs are receptors that nor et al. 2010). To separate the effects of nicotine on symp - mally bind endogenous neurotransmitter acetylcholine, toms of withdrawal versus its direct effects, the results but are also particularly responsive to nicotine. nAChRs were stratified by smoking status. The effects on alerting are abundant in brain regions associated with learning - attention accuracy and short-term episodic memory accu and memory, including the prefrontal cortex (Poorthuis racy were significant in smokers but not in nonsmokers; - et al. 2009), and in primate and rodent models, deple effects on alerting attention RT were significant in non - tion of acetylcholine in the prefrontal cortex results in smokers but not in smokers; effects on working memory impaired attentional performance (Poorthuis et al. 2009; RT were significant in both smokers and nonsmokers, and 2 nAChRs are especially β Wallace and Bertrand 2013). in the remaining outcomes there were insufficient num - abundant in the brain and have a high affinity for nicotine bers of studies on smokers to conduct stratified analysis. (Evans and Drobes 2008; Poorthuis et al. 2009; Herman Thus, nicotine may have some positive effects on cognitive and Sofuoglu 2010). Recent evidence from animal studies performance that are unique to nonsmokers. No studies suggests that β 2 nAChRs play a critical role in regulating meeting the inclusion criteria for the review addressed - attention (Howe et al. 2010; Poorthuis et al. 2013a). Addi learning or executive function. tional research has demonstrated that nicotine interferes with cholinergic control of β 2 nAChRs in the prefrontal Critical Periods of Exposure in the cortex in mice, which could result in acute impairment of Nervous S ystem attention and alterations of the prefrontal cortex network, and lead to long-term effects on attention (Poorthuis et al. Across the lifespan, there are several developmen - 2013a). Mice lacking the tal windows during which exposure to nicotine may have 2 nAChR subunit demonstrate β deficits in executive function (Granon et al. 2003). adverse consequences. In the fetus, nicotine targets neu - rotransmitter receptors in the brain, potentially resulting in abnormalities in cell proliferation and altering synaptic Effects of Nicotine on Cognitive Function in activity (Slotkin 1998). The effects of prenatal exposure Healthy Adult Smokers and Nonsmokers to nicotine on the fetal nervous system are summarized In adults, the negative effects of nicotine with- earlier in this chapter and elsewhere in this report (see drawal on cognitive function have been documented Chapter 9). in both humans and animals, and the administration of Human brain development continues far longer nicotine during withdrawal mitigates cognitive impair - than was previously realized. In particular, areas involved ment (Evans and Drobes 2008). In dependent smokers, in higher cognitive function such as the prefrontal cortex abstinence from smoking is associated with reductions continue to develop throughout adolescence (the period in working memory and sustained attention (Evans and during which individuals are most likely to begin smoking) Drobes 2008), and adverse effects on attention can be and into adulthood (Poorthuis et al. 2009; Goriounova and - seen as early as 30 minutes after smoking the last ciga - Mansvelder 2012). During this extended period of matu rette (Hendricks et al. 2006). Nicotine withdrawal is also ration, substantial neural remodeling occurs, including commonly accompanied by symptoms of negative affect synaptic pruning and changes in dopaminergic input, as (anxiety and depression) (Edwards and Kendler 2011) and well as changes in gray and white matter volume. The relief of this symptom may be an important element of density of projections from the amygdala to the prefron- addiction in smokers (Baker et al. 2004). Because nega - tal cortex increases, suggesting that there is substantial tive affect and attentional control are related, the effects development of the connectivity between the emotional of smoking on these two domains could be interrelated and cognitive areas of the brain (Durston et al. 2001; Ernst (Evans and Drobes 2008). and Fudge 2009). The cholinergic system, which matures 121 Nicotine

151 Surgeon General’s Report ship. The results of the latter study suggest that there may resulted in increased membrane protein concentration in in adolescence, plays a central role in maturation of cogni - tive function and reward (Poorthuis et al. 2009). - the hippocampus, consistent with cell damage and/or cell be mechanisms contributing to cognitive decline in addi tion to and independent of respiratory and cardiovascular loss, in female rats, but not in males (Trauth et al. 1999). - Smoking during adolescence has been associ Male rats with nicotine exposure demonstrated a loss of - disease; however, whether nicotine plays a role in acceler ated with lasting cognitive and behavioral impairments, ating cognitive decline is unknown. including effects on working memory and attention, a dopaminergic response to nicotine more than a month although causal relationships are difficult to establish in after exposure ended, while females exhibited deficits in the presence of potential confounding factors (Goriunova - Other Vulnerable Populations hippocampal norepinephrine content and turnover dur and Mansvelder 2012). In addition, functional magnetic ing the month after nicotine exposure (Trauth et al. 2001). Although the contribution of nicotine to the effects Because estrogen regulates hippocampal cell proliferation resonance imaging in humans showed that young adult of smoking on cognitive decline is unclear, there has been - smokers had reduced prefrontal cortex activation dur in an adult rat, there may be interactions between the a great deal of interest in applications of nicotine as a ing attentional tasks when compared with nonsmoking - effects of nicotine and the hormonal milieu in the adoles treatment for several conditions characterized by cogni - controls. Diminished prefrontal cortex activity correlated cent (Trauth et al. 1999). tive deficits, including Alzheimer’s disease and Parkin - with duration of smoking, supporting the hypothesis that Corresponding behavioral studies of adolescent rats son’s disease. These disorders have underlying deficits in have also shown effects of nicotine exposure. Exposed smoking could have long-lasting effects on cognition the cholinergic system, and it has been hypothesized that females exhibited reduced grooming during exposure and (Musso et al. 2007). nicotine and/or nicotine analogs may be effective in atten - reduced locomotion and rearing after cessation of expo- Animal studies provide evidence that nicotine expo- uating symptoms or slowing disease progression. This sure; these results were not seen in exposed adult rats, sure during adolescence has effects on the brain that dif- hypothesis is further supported by research (reviewed ear - which show increased grooming in both genders and no fer from exposure during other periods of development. lier in this chapter) suggesting that acute administration decrease in locomotion (Trauth et al. 2000). Adolescent Studies in rodents show that nicotine induces changes of nicotine has cognitive-enhancing properties. In addi- - rats, tested 5 weeks after nicotine exposure ended, demon in gene expression in the brain to a greater degree with tion, some early observational studies showed evidence for adolescent exposure than during other periods of develop - strated an increase in premature responses and a reduc- a reduced risk of Alzheimer’s in smokers, suggesting that tion in correct responses when given a serial reaction time ment (Schochet et al. 2005; Polesskaya et al. 2007). DNA components in tobacco smoke, such as nicotine, may have - - microarrays in female rats demonstrated that gene expres test; this effect was not seen with adult exposure (Cou protective properties. A growing body of evidence now notte et al. 2009). sion in response to nicotine was most pronounced around links smoking to an increased risk for Alzheimer’s disease the age of puberty and the effects of nicotine on gene Thus, adolescents appear to be particularly vulner - (Almeida et al. 2002; Anstey et al. 2007; Hernán et al. 2008; expression were most dramatic in the hippocampus, with able to the adverse effects of nicotine on the CNS. Based Purnell et al. 2009) rather than a reduced risk; however, on existing knowledge of adolescent brain development, upregulation of growth factors and cyclic AMP signaling research on nicotine as a treatment for this condition (and Arc results of animal studies, and limited data from studies of pathways (Polesskaya et al. 2007). Expression of the for Parkinson’s disease) continues. - adolescent and young adult smokers, it is likely that nico gene (implicated in synaptic plasticity, learning, memory, - Other disorders associated with cognitive and atten and addiction) was upregulated in the prefrontal cortex in tine exposure during adolescence adversely affects cogni - tional impairment, such as schizophrenia and attention adolescent rats exposed to nicotine, and to a much greater tive function and development. Therefore, the potential deficit hyperactivity disorder (ADHD), are characterized by long-term cognitive effects of exposure to nicotine in this extent than in adult rats (Schochet et al. 2005). a very high prevalence of smoking among those affected. age group are of great concern. Nicotine exposure during adolescence also appears It has been proposed that individuals with these disorders to cause long-term structural and functional changes in - The effects of nicotine exposure on cognitive func smoke in order to alleviate the symptoms of their disease, the brain (Dwyer et al. 2009). Exposure of adolescent rats tion after adolescence and young adulthood are unknown. and a number of clinical trials using nicotine as a thera- to nicotine resulted in upregulation of nAChRs in the There are data to suggest that smoking accelerates some peutic agent have been conducted. midbrain, cerebral cortex, and hippocampus that was still aspects of cognitive decline in adults, and that these effects appear to be mediated by an increased risk of respiratory present 4 weeks after the end of the exposure, in contrast Alzheimer’s Disease Swan and Lessov-Schlaggar and cardiovascular disease to adult rats in which upregulation had disappeared by 4 ( 2007; Almeida et al. 2011 Alzheimer’s disease is a common form of dementia weeks. Receptor upregulation was more pronounced in ). However, in a cohort study male adolescent rats than females (Trauth et al. 1999). of more than 7,000 men and women, the authors found in which individuals experience ongoing deterioration Indices of cell damage and size in rats with adolescent nic - that current male smokers and recent former smokers of cognitive abilities. Although smoking is recognized as a risk factor for Alzheimer’s disease (Peters et al. 2008; otine exposure indicate reduced cell number and size in had a greater 10-year decline in global cognition and the cerebral cortex, midbrain, and hippocampus (Trauth - executive function than never smokers (with the great Cataldo et al. 2010), acute nicotine administration has been reported to improve some Alzheimer’s symptoms, et al. 2000). Structural changes in prefrontal cortex neu est adverse effect on executive function); these differences - rons have also been described, including increased den- such as recall, visual attention, and mood (Lopez-Arrieta were not explained by other health behaviors or measures, including heart disease and stroke, and measures of lung and Sanz 2001). The plausibility of this effect is supported dritic length and spine density (Brown and Kolb 2001). 2 as the exposure Some effects of nicotine exposure appear to be gen - by studies of Alzheimer’s disease patients showing defi function. An analysis using pack-years - measure provided evidence of a dose-response relation- cits in cholinergic systems and a loss of nicotinic binding der-selective. For example, adolescent nicotine exposure sites (Whitehouse et al. 1982 ). However, evidence from randomized trials to support improvement of Alzheimer’s symptoms from nicotine treatment is sparse. In a 2001 2 Pack-years: the number of years of smoking multiplied by the number of packs of cigarettes smoked per day. Cochrane review updated in 2010, the authors found no Chapter 5 122

152 The Health Consequences of Smoking —50 Years of Progress double-blind, placebo-controlled, randomized trials of ship. The results of the latter study suggest that there may be mechanisms contributing to cognitive decline in addi - - treatment for Alzheimer’s disease with nicotine and con cluded that there is no evidence to recommend nicotine tion to and independent of respiratory and cardiovascular disease; however, whether nicotine plays a role in acceler as a treatment for Alzheimer’s disease (Lopez-Arrieta and - Sanz 2001). ating cognitive decline is unknown. Parkinson’s Disease Other Vulnerable Populations Parkinson’s disease is a degenerative hypokinetic Although the contribution of nicotine to the effects movement disorder. Most patients with Parkinson’s dis - of smoking on cognitive decline is unclear, there has been a great deal of interest in applications of nicotine as a ease will also eventually develop cognitive impairment— treatment for several conditions characterized by cogni with deficits in attention, executive and visual-spatial - functions, and memory—and subsequent dementia. In - tive deficits, including Alzheimer’s disease and Parkin son’s disease. These disorders have underlying deficits in Parkinson’s disease, both the dopaminergic and choliner - gic systems undergo degeneration, which leads to deficits the cholinergic system, and it has been hypothesized that nicotine and/or nicotine analogs may be effective in atten in dopamine and acetylcholine at synapses; thus, nicotinic - mechanisms may play a role in cognitive deficits. In con- uating symptoms or slowing disease progression. This - hypothesis is further supported by research (reviewed ear trast to Alzheimer’s, data consistently support that smok - ers are at reduced risk for developing Parkinson’s disease lier in this chapter) suggesting that acute administration of nicotine has cognitive-enhancing properties. In addi- (Ritz et al. 2007; Wirdefeldt et al. 2011), and twin studies have reported a 20–30% tion, some early observational studies showed evidence for reduction of Parkinson’s disease risk for ever smoking or regular smoking in monozygotic a reduced risk of Alzheimer’s in smokers, suggesting that components in tobacco smoke, such as nicotine, may have and dizygotic, same-gender male twin pairs who were protective properties. A growing body of evidence now discordant for Parkinson’s disease ( Tanner et al. 2002; links smoking to an increased risk for Alzheimer’s disease Wirdefeldt et al. 2005). This suggests that genetic factors (Almeida et al. 2002; Anstey et al. 2007; Hernán et al. 2008; contributing to both Parkinson’s disease and smoking are not responsible for the apparent smoking and Parkinson’s Purnell et al. 2009) rather than a reduced risk; however, disease association. research on nicotine as a treatment for this condition (and Two studies have examined the association between for Parkinson’s disease) continues. smokeless tobacco use and risk of Parkinson’s disease: a - Other disorders associated with cognitive and atten tional impairment, such as schizophrenia and attention case-control study found a significant inverse association (odds ratio [OR] 0.18; 95% CI, 0.04–0.82, in ever users vs. deficit hyperactivity disorder (ADHD), are characterized by never users of smokeless tobacco) (Benedetti et al. 2000) a very high prevalence of smoking among those affected. and a prospective cohort study that assessed Parkinson’s It has been proposed that individuals with these disorders disease mortality as the outcome found a relative risk of smoke in order to alleviate the symptoms of their disease, 0.22 (95% CI, 0.07–0.67) for current users of smokeless and a number of clinical trials using nicotine as a thera- tobacco versus never users (O’Reilly et al. 2005). These peutic agent have been conducted. studies add support for a protective role for nicotine. However, there are few controlled trials of the effects of Alzheimer’s Disease nicotine on cognitive function in patients with Parkin - Alzheimer’s disease is a common form of dementia son’s disease, and results have been inconsistent (Kelton in which individuals experience ongoing deterioration et al. 2000; Vieregge et al. 2001; Lemay et al. 2004; Holmes of cognitive abilities. Although smoking is recognized as et al. 2011). a risk factor for Alzheimer’s disease (Peters et al. 2008; Cataldo et al. 2010), acute nicotine administration has ADHD and Schizophrenia been reported to improve some Alzheimer’s symptoms, Several neuropsychiatric disorders characterized such as recall, visual attention, and mood (Lopez-Arrieta by attention-related cognitive defects are characterized and Sanz 2001). The plausibility of this effect is supported - by studies of Alzheimer’s disease patients showing defi by high prevalence of smoking, including ADHD and schizophrenia. It has been suggested that smoking may be cits in cholinergic systems and a loss of nicotinic binding ). particularly reinforcing for individuals with these condi - sites (Whitehouse et al. 1982 However, evidence from tions because of the cognitive-enhancing effects of nico- randomized trials to support improvement of Alzheimer’s tine. Because cholinergic systems play an important role symptoms from nicotine treatment is sparse. In a 2001 2 Pack-years: the number of years of smoking multiplied by the number of packs of cigarettes smoked per day. Cochrane review updated in 2010, the authors found no in functional impairments in certain neurodegenerative 123 Nicotine

153 Surgeon General’s Report Reynolds both developed research programs to explore diseases, it also has been suggested that individuals with the potential uses of nicotine and analogs in the treat - - attention-related cognitive defects may benefit from treat ment with nicotine through nicotine’s role as a choliner ment of neurological disorders (R.J. Reynolds 1993). In - the early 1990s, R.J. Reynolds established both its “Nico - gic agonist (Singh et al. 2004; Kumari and Postma 2005; - Evans and Drobes 2008). Some research suggests that nic tine Pharmacology and Neurodegenerative Disease Pro- otine may improve attention performance in individuals gram” and later Targacept, a pharmaceutical company, for with ADHD and schizophrenia (Evans and Drobes 2008). the purpose of discovering therapeutic uses of nicotinic ADHD is a common disorder of childhood with compounds. Tobacco industry documents indicate that symptoms of inattention and hyperactivity/impulsiv- diversification into the pharmaceutical industry was seen not only as potentially profitable but also as a strategy to ity. Behavioral inhibition and delay aversion deficits are believed to be factors contributing to impulsive behav improve the tobacco industry’s corporate image (Vagg and - ior. Other features, such as poor planning, and deficits Chapman 2005). Data from observational studies describing the in working memory and cognitive flexibility, are more recently recognized traits. Limited research suggests protective effects of smoking on the risk of Parkinson’s disease and Alzheimer’s disease and the high prevalence that nicotine might improve the symptoms and measures of smoking among individuals with ADHD and schizo of behavioral inhibition, delay aversion, and recogni- - tion memory in individuals with ADHD (Gehricke et al. phrenia are often cited in industry-sponsored and non- 2006, 2009). industry-sponsored literature as evidence to support the Schizophrenia is a chronic disorder marked by therapeutic applications of nicotine. However, there is evidence that the tobacco industry influenced many of delusions, hallucinations, thought disorder, and negative symptoms such as flattening of affect. The evidence sug- these epidemiologic studies of smoking and psychiatric disorders. For example, an analysis of publications on the gests that dysregulation of cholinergic systems is involved relationship between smoking and Alzheimer’s disease - in altered sensory physiology and individuals with schizo that controlled for authors’ industry affiliation revealed phrenia have decreased dopaminergic activity in the that pooled ORs for studies without industry funding prefrontal cortex (Punnoose and Belgamwar 2006). The prevalence of smoking in individuals with schizophrenia were neutral or indicated an increased risk with smoking, depending on study design, while industry-affiliated stud- is high, perhaps as the result of an effort of patients to relieve symptoms associated with the disorder (Kumari ies indicated a reduced risk (Cataldo et al. 2010). Studies of tobacco industry documents have also revealed that the and Postma 2005). Specifically, it has been suggested industry sought to influence scientific attitudes regard that nicotine-induced release of dopamine could improve - ing the role of smoking in schizophrenia (Prochaska et attention and processing symptoms and sensory-gating al. 2008). Tobacco industry documents indicate that the deficits in schizophrenia, and that nicotine treatment - could attenuate antipsychotic-induced cognitive impair - industry funded research for the specific purpose of per petuating the belief that smoking improves symptoms ment and extrapyramidal symptoms, through nicotine’s in schizophrenic patients, advocated for exceptions for effects on dopamine release (Alder et al. 1993; Newhouse smoking in hospitalized psychiatric patients, and funded et al. 2004; Birkett et al. 2007; Evans and Drobes 2008). studies of medicinal uses of nicotine analogs to treat men- However, in a 2012 Cochrane Review update, the authors tal illnesses (Prochaska et al. 2008). - reviewed all randomized controlled trials in which nico Evidence of the tobacco industry’s interest in the tine or tobacco and placebo were administered to patients with schizophrenia or schizophrenia-like illness and cognitive-enhancing properties of nicotine comes from - a 1997 review of publications investigating the effects of found no studies that met the inclusion criteria. A num tobacco and nicotine on cognitive performance. Turner ber of studies were excluded because they were a crossover and Spilich (1997) found that authors acknowledging design, which was determined to be inappropriate because tobacco industry funding were much less likely than schizophrenia is an unstable condition and nicotine may nonindustry-funded authors to report a negative effect of have carryover effects (Punnoose and Belgamwar 2006). nicotine on cognitive performance. Nonindustry-funded authors reported both positive and negative findings, Tobacco Industry Influence while industry-funded authors reported positive findings The tobacco industry has a long-standing interest in almost exclusively (Turner and Spilich 1997). Studies of nicotine and neurocognitive functioning and psychiatric this type using more recent published articles are needed disease. The tobacco industry has invested in pharma- to better understand current industry influences on the ceutical applications of nicotine and nicotine analogs for scientific literature. decades (Vagg and Chapman 2005). Philip Morris and R.J. 124 Chapter 5

154 —50 Years of Progress The Health Consequences of Smoking smokers during withdrawal. However, less is known about It is difficult to estimate the extent to which indus - try-generated research activities have influenced scientific the acute effects in nonabstinent smokers and in nonsmok- thinking regarding the effects of nicotine on cognitive ers, and about the effects of long-term nicotine exposure on cognitive performance. Human and animal evidence performance and on nicotine’s therapeutic applications. Authors’ industry affiliations and potential conflicts of show detrimental effects on cognition from smoking dur - interest reported in publications may go unnoticed by ing aging. Evidence also shows that exposure to cigarette readers, may be difficult to identify, or may not be disclosed - smoke and to nicotine has adverse effects on fetal and ado lescent brain development, which could result in lasting at all. Reviews and other articles citing industry-affiliated deficits in cognitive function. Furthermore, withdrawal studies generally did not include author affiliations or from tobacco in dependent-users results in cognitive potential conflicts of interest at all, leaving the readers unaware of possible industry influences. A growing con - impairment. Among individuals with attention-related - cern about conflicts of interest resulting from funding cognitive defects, nicotine has been proposed as a poten tial treatment because of its effect as a cholinergic agonist. through the tobacco industry is reflected in the National However, randomized controlled trials to demonstrate Institute on Drug Abuse (NIDA) advice to its grantees that safety and efficacy of nicotine treatment in individuals - “Receiving funding from the tobacco industry may com with these disorders are lacking, and the long-term effects promise the perceived objectivity of their research results, which in turn could impact the overall credibility of their of low-dose, chronic nicotine exposure on individuals research findings, including its interpretation, acceptance with neuropsychiatric disorders are unknown. Because and implementation” (NIDA n.d.). nAChRs are distributed extensively across the central and peripheral nervous systems, studying the effects of nico- tine across the behavioral spectrum, rather than on iso - lated domains, may reveal adverse effects and may help Summary establish whether the potential benefits of nicotine are clinically meaningful (Heishman 1998). Evidence shows that acute nicotine administration has some modest cognition-enhancing effects in adult Evidence Summary known about the effects of long-term nicotine exposure This chapter complements reviews in prior reports on cognitive performance and how nicotine withdrawal and in other sections of this report on the potential tox - icity of nicotine, a pharmacologically active agent that impairs cognition. Previous reports have reached causal readily enters the body and is distributed throughout. conclusions related to nicotine and addiction (USDHHS 1988, 2010, 2012). Evidence in this chapter considers the Nicotine activates multiple biological pathways that are particular vulnerability of adolescents and other groups relevant to fetal growth and development, immune func - to nicotine. Beyond the use of NRT cessation aids, the - tion, the cardiovascular system, the CNS, and carcino genesis. Experimental research documents that nicotine therapeutic roles for nicotine have not been established, in spite of clinical research, some carried out by the plays a key role in several adverse consequences of mater - tobacco industry. - nal smoking for the fetus, including altered lung develop Acute toxicity of nicotine, reflecting its pharmaco- ment, and has effects on the developing brain. Evidence logic activity, is well established. There is a potential for supports that acute nicotine administration has modest cognition-enhancing properties in adult smokers during poisoning from ingestion of nicotine-containing products. withdrawal and in adult nonsmokers. However, little is 125 Nicotine

155 Surgeon General’s Report Conclusions 4. The evidence is sufficient to infer that nicotine The evidence is sufficient to infer that at high-enough 1. adversely affects maternal and fetal health during doses nicotine has acute toxicity. pregnancy, contributing to multiple adverse out - 2. comes such as preterm delivery and stillbirth. The evidence is sufficient to infer that nicotine acti- vates multiple biological pathways through which smoking increases risk for disease. 5. The evidence is suggestive that nicotine exposure - during adolescence, a critical window for brain devel 3. - opment, may have lasting adverse consequences for The evidence is sufficient to infer that nicotine expo brain development. sure during fetal development, a critical window for brain development, has lasting adverse consequences The evidence is inadequate to infer the presence or 6. for brain development. absence of a causal relationship between exposure to nicotine and risk for cancer. Implications smoke should consider and discuss with their health care Large numbers of people are exposed to nicotine through products other than conventional cigarettes, providers the potential risk to the fetus from continuing including NRT, smokeless tobacco, and new nicotine- - to smoke and from using NRT. There is a strong recom containing noncombustible products. The fetus will be mendation from the U.S. Preventive Services Task Force for health care providers to ask pregnant women about exposed to nicotine without other smoke components if the mother uses these products. The number of people tobacco use and provide the appropriate counseling. exposed to nicotine long-term may grow under a num - - The possibility of increasing chronic nicotine expo sure in the population from various nicotine-containing ber of potential future scenarios; for example, expanding - use of multiple products or the replacement of conven- products for the long-term merits further research. Can tional combustible cigarettes with other nicotine delivery cer, cardiovascular, and neurocognitive outcomes are of concern. The evidence is already sufficient to provide systems (see Chapter 15, “The Changing Landscape of appropriately cautious messages to pregnant women and Tobacco Control: Current Status and Future Directions”), women of reproductive age as well as adolescents about or increased appeal and uptake of nicotine product use - because of their apparent relative safety in comparison the use of nicotine-containing products such as smoke to cigarettes. In considering such scenarios, information less tobacco and electronic cigarettes, and newer forms - of nicotine-containing tobacco products, as alternatives will be needed on the risks of long-term exposure to nico to smoking. tine, including the consequences for reproductive health All tobacco products contain toxicants, so all and adolescent cognitive development, compared with tobacco product use poses some health risks. Because of cigarette smoking, and no tobacco products use at all. the potential for fetal and adolescent nicotine exposure The evidence reviewed in this chapter, in other chapters in this report, and in previous reports shows that long-term to have long-term detrimental effects on brain develop- nicotine use may have adverse consequences for those ment, measures should be taken to ensure that nicotine exposed and it clearly harms the developing fetus. The is not perceived by the public as a cognitive-enhancing latest U.S. Public Health Service guidelines acknowledge substance. It also does not have an established role in the this risk and have not made a specific recommendation on management of people with a severe mental illness. the use of NRT during pregnancy. Pregnant women who 126 Chapter 5

156 —50 Years of Progress The Health Consequences of Smoking References Accortt NA, Waterbor JW, Beall C, Howard G. Can Anstey KJ, von Sanden C, Salim A, O’Kearney R. Smoking - as a risk factor for dementia and cognitive decline: a cer incidence among a cohort of smokeless tobacco Cancer Causes and Control American Journal users (United States). meta-analysis of prospective studies. 2005;16(9):1107–15. 2007;166(4):367–78. of Epidemiology Araujo DM, Lapchak PA, Collier B, Quirion R. Character Adler LE, Hoffer LD, Wiser A, Freedman R. Normalization - of auditory physiology by cigarette smoking in schizo- ization of N-[3H]methylcarbamylcholine binding sites and effect of N-methylcarbamylcholine on acetylcho- American Journal of Psychiatry phrenic patients. line release in rat brain. 1993;150(12):1856–61. Journal of Neurochemistry 1988;51(1):292–9. Aicher A, Heeschen C, Mohaupt M, Cooke JP, Zeiher AM, Argentin G, Cicchetti R. Genotoxic and antiapoptotic Dimmeler S. Nicotine strongly activates dendritic effect of nicotine on human gingival fibroblasts. Toxi- cell-mediated adaptive immunity: potential role for 2004;79(1):75–81. cological Sciences Circulation progression of atherosclerotic lesions. Baba S, Wikström AK, Stephansson O, Cnattingius S. 2003;107(4):604–11. Albuquerque EX, Alkondon M, Pereira EF, Castro NG, Influence of smoking and snuff cessation on risk of preterm birth. European Journal of Epidemiology Schrattenholz A, Barbosa CT, Bonfante-Cabarcas R, 2012;27(4):297–304. Aracava Y, Eisenberg HM, Maelicke A. Properties of Baba S, Wikström AK, Stephansson O, Cnattingius S. neuronal nicotinic acetylcholine receptors: pharma- - Changes in snuff and smoking habits in Swedish preg cological characterization and modulation of synaptic nant women and risk for small for gestational age Journal of Pharmacology and Experimental function. 1997;280(3):1117–36. Therapeutics BJOG: An International Journal of Obstetrics births. and Gynaecology 2013;120(4):456–62. Alkondon M, Pereira EF, Barbosa CT, Albuquerque EX. - Neuronal nicotinic acetylcholine receptor activa Baker TB, Piper ME, McCarthy DE, Majeskie MR, Fiore tion modulates gamma-aminobutyric acid release MC. Addiction motivation reformulated: an affective processing model of negative reinforcement. from CA1 neurons of rat hippocampal slices. Journal Psycho- logical Review 2004;111(1):33–51. of Pharmacology and Experimental Therapeutics 1997;283(3):1396–411. Balfour DJ. The neurobiology of tobacco dependence: a Almeida OP, Garrido GJ, Alfonso H, Hulse G, Lauten - - preclinical perspective on the role of the dopamine pro schlager NT, Hankey GJ, Flicker L. 24-month effect Nico- jections to the nucleus accumbens [corrected]. of smoking cessation on cognitive function and brain 2004;6(6):899–912. tine & Tobacco Research Barrueco M, Otero MJ, Palomo L, Jimenez-Ruiz C, Tor structure in later life. Neuroimage 2011;55(4):1480–9. - - Almeida OP, Hulse GK, Lawrence D, Flicker L. Smoking recilla M, Romero P, Riesco JA. Adverse effects of phar as a risk factor for Alzheimer’s disease: contrasting macological therapy for nicotine addiction in smokers evidence from a systematic review of case-control and following a smoking cessation program. Nicotine & cohort studies. Addiction 2002;97(1):15–28. 2005;7(3):335–42. Tobacco Research Al-Wadei HA, Plummer HK 3rd, Schuller HM. Nicotine Bauer CM, Dewitte-Orr SJ, Hornby KR, Zavitz CC, Lichty stimulates pancreatic cancer xenografts by systemic BD, Stämpfli MR, Mossman KL. Cigarette smoke sup - increase in stress neurotransmitters and suppression of presses type I interferon-mediated antiviral immunity the inhibitory neurotransmitter gamma-aminobutyric Journal of Inter- in lung fibroblast and epithelial cells. acid. Carcinogenesis 2009;30(3):506–11. 2008;28(3):167–79. feron and Cytokine Research American Academy of Pediatrics Task Force on Sud- Bell SL, Taylor RC, Singleton EG, Henningfield JE, Heish - den Infant Death Syndrome. SIDS and other sleep- man SJ. Smoking after nicotine deprivation enhances related infant deaths: expansion of recommendations - cognitive performance and decreases tobacco crav for a safe infant sleeping environment. Pediatrics ing in drug abusers. Nicotine & Tobacco Research 2011;128(5):e1341–e67. 1999;1(1):45–52. American Psychiatric Association. Diagnostic and Statis - Benedetti MD, Bower JH, Maraganore DM, McDonnell tical Manual of Mental Disorders . 4th ed. Washington: SK, Peterson BJ, Ahlskog JE, Schaid DJ, Rocca WA. American Psychiatric Association, 1994. Smoking, alcohol, and coffee consumption preceding Neurology Parkinson’s disease: a case-control study. 2000;55(9):1350–8. 127 Nicotine

157 Surgeon General’s Report Benowitz N, Dempsey D. Pharmacotherapy for smok Brown RW, Kolb B. Nicotine sensitization increases den - - ing cessation during pregnancy. Nicotine & Tobacco dritic length and spine density in the nucleus accum- 2004;6(Suppl 2):S189–S202. Brain Research Research bens and cingulate cortex. 2001; 899(1-2):94–100. Benowitz NL, Hukkanen J, Jacob P 3rd. Nicotine chemis- - Bublitz MH, Stroud LR. Maternal smoking during preg try, metabolism, kinetics and biomarkers. Handbook of 2009;(192):29–60. nancy and offspring brain structure and function: Experimental Pharmacology review and agenda for future research. - Benowitz NL, Jacob P 3rd. Metabolism of nicotine to coti Nicotine & nine studied by a dual stable isotope method. Clinical 2012;14(4):388–97. Tobacco Research Pharmacology and Therapeutics 1994;56(5):483–93. Budulac SE, Vonk JM, Postma DS, Siedlinski M, Timens Benowitz NL, Jacob P 3rd, Jones RT, Rosenberg J. Interin W, Boezen MH. Nicotinic acetylcholine receptor vari- - ants are related to smoking habits, but not directly to dividual variability in the metabolism and cardiovascu - PloS One 2012;7(3):e33386. COPD. lar effects of nicotine in man. Journal of Pharmacology Cahill K, Stead LF, Lancaster T. Nicotine receptor partial 1982;221(2):368–72. and Experimental Therapeutics Benowitz NL, Lake T, Keller KH, Lee BL. Prolonged Cochrane Database of agonists for smoking cessation. absorption with development of tolerance to toxic Systematic Reviews 2008, Issue 3. Art No.: CD006103. Clinical effects after cutaneous exposure to nicotine. DOI: 10.1002/14651858.CD006103.pub3. Pharmacology and Therapeutics Carboni E, Bortone L, Giua C, Di Chiara G. Dissociation 1987;42(1):119–20. of physical abstinence signs from changes in extracel Bierut LJ. Nicotine dependence and genetic variation in - the nicotinic receptors. Drug and Alcohol Dependence lular dopamine in the nucleus accumbens and in the prefrontal cortex of nicotine dependent rats. 2009;104(Suppl 1):S64–S9. Drug and 2000;58(1-2):93–102. Alcohol Dependence Bierut LJ. Convergence of genetic findings for nicotine Cardinale A, Nastrucci C, Cesario A, Russo P. Nicotine: dependence and smoking related diseases with chro - specific role in angiogenesis, proliferation and apopto- Trends in Pharmacological Sci - mosome 15q24-25. 2012;42(1):68–89. Critical Reviews in Toxicology sis. 2010;31(1):46–51. ences Carlisle DL, Liu X, Hopkins TM, Swick MC, Dhir R, Sieg- Birkett P, Sigmundsson T, Sharma T, Toulopoulou T, fried JM. Nicotine activates cell-signaling pathways Griffiths TD, Reveley A, Murray R. Reaction time and through muscle-type and neuronal nicotinic ace- sustained attention in schizophrenia and its genetic tylcholine receptors in non-small cell lung cancer 2007;95(1-3): Schizophrenia Research predisposition. Pulmonary Pharmacology and Therapeutics cells. 76–85. 2007;20(6):629–41. Boffetta P, Aagnes B, Weiderpass E, Andersen A. Smoke - Carmella SG, Borukhova A, Desai D, Hecht SS. Evidence less tobacco use and risk of cancer of the pancreas - for endogenous formation of tobacco-specific nitrosa International Journal of Cancer and other organs. mines in rats treated with tobacco alkaloids and sodium 2005;114(6):992–5. nitrite. 1997;18(3):587–92. Carcinogenesis Boffetta P, Hecht S, Gray N, Gupta P, Straif K. Smoke - Cashman JR, Park SB, Yang ZC, Wrighton SA, Jacob P less tobacco and cancer. Lancet Oncology 2008;9(7): 3rd, Benowitz NL. Metabolism of nicotine by human 667–75. liver microsomes: stereoselective formation of trans- - Brady ME, Ritschel WA, Saelinger DA, Cacini W, Patter Chemical Research in Toxicology nicotine N’-oxide. son AJ. Animal model and pharmacokinetic interpre - 1992;5(5):639–46. International tation of nicotine poisoning in man. Cataldo JK, Prochaska JJ, Glantz SA. Cigarette smoking Journal of Clinical Pharmacology and Biopharmacy is a risk factor for Alzheimer’s Disease: an analysis 1979;17(1):12–7. Journal of controlling for tobacco industry affiliation. Breese CR, Marks MJ, Logel J, Adams CE, Sullivan B, Col - 2010;19(2):465–80. Alzheimer’s Disease lins AC, Leonard S. Effect of smoking history on [3H] Catassi A, Servent D, Paleari L, Cesario A, Russo P. Mul - nicotine binding in human postmortem brain. Jour- tiple roles of nicotine on cell proliferation and inhibi - nal of Pharmacology and Experimental Therapeutics tion of apoptosis: implications on lung carcinogenesis. 1997;282(1):7–13. 2008;659(3):221–31. Mutation Research Broms U, Wedenoja J, Largeau MR, Korhonen T, Pitkani- Chen J, Higby R, Tian D, Tan D, Johnson MD, Xiao Y, Kel- emi J, Keskitalo-Vuokko K, Happola A, Heikkila KH, lar KJ, Feng S, Shields PG. Toxicological analysis of - Heikkila K, Ripatti S, et al. Analysis of detailed phe Toxicology low-nicotine and nicotine-free cigarettes. notype profiles reveals CHRNA5-CHRNA3-CHRNB4 2008a;249(2-3):194–203. - gene cluster association with several nicotine depen Nicotine & Tobacco Research dence traits. 2012;14(6): 720–33. Chapter 5 128

158 —50 Years of Progress The Health Consequences of Smoking - Chen RJ, Ho YS, Guo HR, Wang YJ. Rapid activation of Dasgupta P, Chellappan SP. Nicotine-mediated cell prolif Stat3 and ERK1/2 by nicotine modulates cell prolifera - eration and angiogenesis: new twists to an old story. tion in human bladder cancer cells. Toxicological Sci- 2006;5(20):2324–8. Cell Cycle ences 2008b;104(2):283–93. Dasgupta P, Rastogi S, Pillai S, Ordonez-Ercan D, Morris Chen RJ, Ho YS, Guo HR, Wang YJ. Long-term nicotine M, Haura E, Chellappan S. Nicotine induces cell prolif - exposure-induced chemoresistance is mediated by eration by beta-arrestin-mediated activation of Src and Journal of Clinical Investigation Rb-Raf-1 pathways. activation of Stat3 and downregulation of ERK1/2 via 2006;116(8):2208–17. - nAChR and beta-adrenoceptors in human bladder can Toxicological Sciences - Davis R, Rizwani W, Banerjee S, Kovacs M, Haura E, Cop cer cells. 2010;115(1):118–30. Chen YP, Squier CA. Effect of nicotine on pola D, Chellappan S. Nicotine promotes tumor growth 7,12-dimethylbenz[a]anthracene carcinogenesis in and metastasis in mouse models of lung cancer. PloS 2009;4(10):e7524. One Journal of the National Cancer hamster cheek pouch. de Wit H, Zacny J. Abuse potential of nicotine replacement 1990;82(10):861–4. Institute 1995;4(6):456–68. Cheng Y, Li HL, Wang HF, Sun HF, Liu YF, Peng SX, Liu CNS Drugs therapies. Dempsey DA, Benowitz NL. Risks and benefits of nicotine - KX, Guo ZY. Inhibition of nicotine-DNA adduct for mation in mice by six dietary constituents. Food and Drug Safety to aid smoking cessation in pregnancy. 2003;41(7):1045–50. 2001;24(4):277–322. Chemical Toxicology Dezateux C, Stocks J, Dundas I, Fletcher ME. Impaired Citigroup Global Markets Inc. Tobacco: What if the airway function and wheezing in infancy: the influence London: Citi Investment Last Smoker Quits in 2050? Research & Analysis, 2011. of maternal smoking and a genetic predisposition to Cloez-Tayarani I, Changeux JP. Nicotine and serotonin asthma. American Journal of Respiratory and Critical Care Medicine 1999;159(2):403–10. in immune regulation and inflammatory processes: a perspective. Journal of Leukocyte Biology 2007;81(3): Doolittle DJ, Rahn CA, Lee CK. The effect of exposure to - 599–606. nicotine, carbon monoxide, cigarette smoke or ciga Conklin BS, Zhao W, Zhong DS, Chen C. Nicotine and rette smoke condensate on the mutagenicity of rat 1991;260(1):9–18. cotinine up-regulate vascular endothelial growth fac - urine. Mutation Research Doolittle DJ, Winegar R, Lee CK, Caldwell WS, Hayes AW, American Journal tor expression in endothelial cells. de Bethizy JD. The genotoxic potential of nicotine and 2002;160(2):413–8. of Pathology 1995;344 Mutation Research its major metabolites. Counotte DS, Spijker S, Van de Burgwal LH, Hogenboom (3-4):95–102. F, Schoffelmeer AN, De Vries TJ, Smit AB, Pattij T. Duncan JR, Garland M, Myers MM, Fifer WP, Yang M, Long-lasting cognitive deficits resulting from adoles- Kinney HC, Stark RI. Prenatal nicotine-exposure cent nicotine exposure in rats. Neuropsychopharma- - alters fetal autonomic activity and medullary neu cology 2009;34(2):299–306. rotransmitter receptors: implications for sudden infant Crooks PA, Dwoskin LP. Contribution of CNS nicotine Journal of Applied Physiology death syndrome. 2009; metabolites to the neuropharmacological effects of 107(5):1579–90. Biochemical Pharma- nicotine and tobacco smoking. Durston S, Hulshoff Pol HE, Casey BJ, Giedd JN, Buitelaar 1997;54(7):743–53. cology - JK, van Engeland H. Anatomical MRI of the develop Cunningham J, Dockery DW, Gold DR, Speizer FE. Racial Journal of the ing human brain: what have we learned? differences in the association between maternal smok - American Academy of Child and Adolescent Psychiatry ing during pregnancy and lung function in children. 2001;40(9):1012-20. American Journal of Respiratory and Critical Care Dwyer JB, Broide RS, Leslie FM. Nicotine and brain devel - 1995;152(2):565–9. Medicine Birth Defects Research. Part C, Embryo Today opment. Cunningham J, Dockery DW, Speizer FE. Maternal smok- 2008;84(1):30–44. - ing during pregnancy as a predictor of lung func Dwyer JB, McQuown SC, Leslie FM. The dynamic effects American Journal of Epidemiology tion in children. Pharmacology and of nicotine on the developing brain. 1994;139(12):1139–52. Therapeutics 2009;122(2):125–39. - Dahlstrom A, Lundell B, Curvall M, Thapper L. Nico Edwards AC, Kendler KS. Nicotine withdrawal-induced tine and cotinine concentrations in the nursing negative affect is a function of nicotine dependence mother and her infant. Acta Paediatrica Scandinavica and not liability to depression or anxiety. Nicotine & 1990;79(2):142–7. Tobacco Research 2011;13(8):677–85. Dani JA. Properties underlying the influence of nicotinic Egleton RD, Brown KC, Dasgupta P. Nicotinic acetylcho- receptors on neuronal excitability and epilepsy. Epilep- line receptors in cancer: multiple roles in proliferation 2000;41(8):1063–5. sia 129 Nicotine

159 Surgeon General’s Report Trends in Pharmacological and inhibition of apoptosis. Gehricke JG, Whalen CK, Jamner LD, Wigal TL, Steinhoff Sciences 2008;29(3):151–8. K. The reinforcing effects of nicotine and stimulant medication in the everyday lives of adult smokers with England LJ, Kim SY, Shapiro-Mendoza CK, Wilson HG, - ADHD: a preliminary examination. Kendrick JS, Satten GA, Lewis CA, Tucker MJ, Cal Nicotine & Tobacco laghan WM. Effects of maternal smokeless tobacco Research 2006;8(1):37–47. Geng Y, Savage SM, Johnson LJ, Seagrave J, Sopori ML. use on selected pregnancy outcomes in Alaska Native Acta Obstetricia et Gyne Effects of nicotine on the immune response. I. Chronic women: a case-control study. - 2013;92(6):648–55. cologica Scandinavica - exposure to nicotine impairs antigen receptor-medi England LJ, Kim SY, Shapiro-Mendoza CK, Wilson HG, ated signal transduction in lymphocytes. Toxicology Kendrick JS, Satten GA, Lewis CA, Whittern P, Tucker and Applied Pharmacology 1995;135(2):268–78. MJ, Callaghan WM. Maternal smokeless tobacco use Geng Y, Savage SM, Razani-Boroujerdi S, Sopori ML. Effects of nicotine on the immune response. II. Chronic in Alaska Native women and singleton infant birth Journal of size. Acta Obstetricia et Gynecologica Scandinavica nicotine treatment induces T cell anergy. Immunology 1996;156(7):2384–90. 2012;91(1):93–103. England LJ, Levine RJ, Mills JL, Klebanoff MA, Yu KF, Gilliland FD, Berhane K, Li YF, Rappaport EB, Peters JM. Effects of early onset asthma and in utero exposure to Cnattingius S. Adverse pregnancy outcomes in snuff maternal smoking on childhood lung function. Ameri- users. American Journal of Obstetrics and Gynecology 2003;189(4):939–43. can Journal of Respiratory and Critical Care Medicine 2003;167(6):917–24. Ernst M, Fudge JL. A developmental neurobiological model of motivated behavior: anatomy, connectivity Gilliland FD, Berhane K, McConnell R, Gauderman WJ, Neuroscience and and ontogeny of the triadic nodes. Vora H, Rappaport EB, Avol E, Peters JM. Maternal 2009;33(3):367–82. Biobehavioral Reviews smoking during pregnancy, environmental tobacco Evans DE, Drobes DJ. Nicotine self-medication of cogni - smoke exposure and childhood lung function. Thorax 2009; 2000;55(4):271–6. Addiction Biology tive-attentional processing. 14(1):32–42. Ginzkey C, Stueber T, Friehs G, Koehler C, Hackenberg Fant RV, Owen LL, Henningfield JE. Nicotine replace S, Richter E, Hagen R, Kleinsasser NH. Analysis of - nicotine-induced DNA damage in cells of the human Primary Care: Clinics in Office Practice ment therapy. 2012;208(1):23–9. Toxicology Letters respiratory tract. 1999;26(3):633–52. Gioanni Y, Rougeot C, Clarke PB, Lepouse C, Thierry AM, Fu XW, Lindstrom J, Spindel ER. Nicotine activates and Vidal C. Nicotinic receptors in the rat prefrontal cortex: up-regulates nicotinic acetylcholine receptors in bron- chial epithelial cells. American Journal of Respiratory - increase in glutamate release and facilitation of medio Cell and Molecular Biology 2009;41(1):93–9. dorsal thalamo-cortical transmission. European Jour- 1999;11(1):18–30. nal of Neuroscience Fu Y, Matta SG, Gao W, Brower VG, Sharp BM. Systemic Gold AB, Lerman C. Pharmacogenetics of smoking ces - nicotine stimulates dopamine release in nucleus sation: role of nicotine target and metabolism genes. accumbens: re-evaluation of the role of N-methyl-D- Human Genetics 2012;131(6):857–76. aspartate receptors in the ventral tegmental area. Jour- Goriounova NA, Mansvelder HD. Short- and long-term nal of Pharmacology and Experimental Therapeutics - consequences of nicotine exposure during adoles 2000;294(2):458–65. cence for prefrontal cortex neuronal network func - Fung YK, Schmid MJ, Anderson TM, Lau YS. Effects of Cold Spring Harbor Perspectives in Medicine tion. nicotine withdrawal on central dopaminergic systems. 2012;2(12):a012120. Pharmacology, Biochemistry and Behavior 1996; Gotti C, Zoli M, Clementi F. Brain nicotinic acetylcholine 53(3):635–40. receptors: native subtypes and their relevance. Trends Galitovskiy V, Chernyavsky AI, Edwards RA, Grando SA. in Pharmacological Sciences 2006;27(9):482–91. Muscle sarcomas and alopecia in A/J mice chroni - Gourlay SG, Benowitz NL. Arteriovenous differences in cally treated with nicotine. 2012;91(21- Life Sciences plasma concentration of nicotine and catecholamines 22):1109–12. - and related cardiovascular effects after smoking, nico Gehricke JG, Hong N, Whalen CK, Steinhoff K, Wigal TL. tine nasal spray, and intravenous nicotine. Clinical Effects of transdermal nicotine on symptoms, moods, Pharmacology and Therapeutics 1997;62(4):453–63. and cardiovascular activity in the everyday lives of Grady SR, Meinerz NM, Cao J, Reynolds AM, Picciotto - smokers and nonsmokers with attention-deficit/hyper MR, Changeux JP, McIntosh JM, Marks MJ, Collins AC. activity disorder. Psychology of Addictive Behaviors Nicotinic agonists stimulate acetylcholine release from 2009;23(4):644–55. 130 Chapter 5

160 —50 Years of Progress The Health Consequences of Smoking Hecht SS. Tobacco carcinogens, their biomarkers and mouse interpeduncular nucleus: a function mediated by a different nAChR than dopamine release from stria- tobacco-induced cancer. 2003; Nature Reviews: Cancer 3(10):733–44. 2001;76(1):258–68. Journal of Neurochemistry tum. Granon S, Faure P, Changeux JP. Executive and social Hecht SS, Carmella SG, Chen M, Dor Koch JF, Miller AT, behaviors under nicotinic receptor regulation. Pro- Murphy SE, Jensen JA, Zimmerman CL, Hatsukami ceedings of the National Academy of Sciences of the DK. Quantitation of urinary metabolites of a tobacco- Can- specific lung carcinogen after smoking cessation. United States of America 2003;100(16):9596–601. - cer Research Gray R, Rajan AS, Radcliffe KA, Yakehiro M, Dani JA. Hip 1999;59(3):590–6. - pocampal synaptic transmission enhanced by low con Heeschen C, Jang JJ, Weis M, Pathak A, Kaji S, Hu RS, centrations of nicotine. Nature 1996;383(6602):713–6. Tsao PS, Johnson FL, Cooke JP. Nicotine stimulates - Grenhoff J, Aston-Jones G, Svensson TH. Nicotinic effects angiogenesis and promotes tumor growth and athero Nature Medicine sclerosis. on the firing pattern of midbrain dopamine neurons. 2001;7(7):833–9. Heeschen C, Weis M, Aicher A, Dimmeler S, Cooke JP. A Acta Physiologica Scandinavica 1986;128(3):351–8. novel angiogenic pathway mediated by non-neuronal Grillner P, Svensson TH. Nicotine-induced excitation of midbrain dopamine neurons in vitro involves iono Journal of Clinical nicotinic acetylcholine receptors. - Synapse Investigation 2002;110(4):527–36. tropic glutamate receptor activation. 2000; 38(1):1–9. Heishman SJ. What aspects of human performance are 1998;93(3): Addiction truly enhanced by nicotine? Gunnerbeck A, Wikstrom AK, Bonamy AK, Wickstrom 317–20. R, Cnattingius S. Relationship of maternal snuff use Pediatrics and cigarette smoking with neonatal apnea. Heishman SJ, Kleykamp BA, Singleton EG. Meta-analysis of the acute effects of nicotine and smoking on human 2011;128(3):503–9. Psychopharmacology performance. Gupta PC, Subramoney S. Smokeless tobacco use, birth 2010;210(4): weight, and gestational age: population based, prospec - 453–69. Hendricks PS, Ditre JW, Drobes DJ, Brandon TH. The tive cohort study of 1,217 women in Mumbai, India. - Psy early time course of smoking withdrawal effects. British Medical Journal 2004;328(7455):1538. chopharmacology 2006;187(3):385–96. Gupta PC, Subramoney S. Smokeless tobacco use and risk Henningfield JE, Hatsukami DK, Zeller M, Peters E. Con- Epidemi- of stillbirth: a cohort study in Mumbai, India. ference on abuse liability and appeal of tobacco prod - ology 2006;17(1):47–51. ucts: conclusions and recommendations. Drug and Gurkalo VK, Volfson NI. Nicotine influence upon the Alcohol Dependence 2011;116(1):1–7. development of experimental stomach tumors. Archiv Henningfield JE, Keenan RM. Nicotine delivery kinetics für Geschwulstforschung 1982;52(4):259–65. and abuse liability. Journal of Consulting and Clinical Habs M, Schmahl D. Influence of nicotine on N-nitro- 1993;61(5):743–50. Psychology somethylurea-induced mammary tumors in rats. Henningfield JE, Stapleton JM, Benowitz NL, Grayson RF, 1984;62(Suppl 2):105–8. Klinische Wochenschrift London ED. Higher levels of nicotine in arterial than in Hacker and Moore’s Hacker NF, Gambone JC, Hobel SJ. - venous blood after cigarette smoking. Drug and Alco Essentials of Obstetrics and Gynecology . 5th ed. Phila- 1993;33(1):23–9. hol Dependence delphia: Saunders Elsevier, 2010. Herman AI, Sofuoglu M. Cognitive effects of nicotine: Hajek P, Jackson P, Belcher M. Long-term use of nicotine 2010;15(3): genetic moderators. Addiction Biology chewing gum. Occurrence, determinants, and effect on 250–65. JAMA: the Journal of the American Medi- weight gain. Hernán MA, Alonso A, Logroscino G. Cigarette smoking cal Association 1988;260(11):1593–6. and dementia: potential selection bias in the elderly. Hamilton BE, Martin JA, Ventura SJ. Births: prelimi - Epidemiology 2008;19(3):448–50. National Vital Statistics Reports nary data for 2011. Hicks CS, Sinclair DA. Toxicities of the optical isomers of 2012;61(5):1–18. nicotine and nornicotine. Australian Journal of Exper- Hanrahan JP, Tager IB, Segal MR, Tosteson TD, Castile imental Biology and Medical Science 1947;25:83–6. RG, Van Vunakis H, Weiss ST, Speizer FE. The effect Hildebrand BE, Nomikos GG, Hertel P, Schilstrom B, of maternal smoking during pregnancy on early infant Svensson TH. Reduced dopamine output in the nucleus American Review of Respiratory Disease lung function. accumbens but not in the medial prefrontal cortex in 1992;145(5):1129–35. rats displaying a mecamylamine-precipitated nicotine Harding R, Maritz G. Maternal and fetal origins of lung withdrawal syndrome. Brain Research 1998;779(1- Seminars in Fetal & Neonatal disease in adulthood. 2):214–25. 2012;17(2):67–72. Medicine 131 Nicotine

161 Surgeon General’s Report Hildebrand BE, Panagis G, Svensson TH, Nomikos GG. Juárez SP, Merlo J. The effect of Swedish snuff (snus) PloS One Behavioral and biochemical manifestations of meca- on offspring birthweight: a sibling analysis. 2013;8(6):e65611. mylamine-precipitated nicotine withdrawal in the rat: Kapoor M, Wang JC, Bertelsen S, Bucholz K, Budde JP, role of nicotinic receptors in the ventral tegmental Neuropsychopharmacology 1999;21(4):560–74. Hinrichs A, Agrawal A, Brooks A, Chorlian D, Dick D, area. - et al. Variants located upstream of CHRNB4 on chro Holmes AD, Copland DA, Silburn PA, Chenery HJ. Acute nicotine enhances strategy-based semantic processing mosome 15q25.1 are associated with age at onset PloS One of daily smoking and habitual smoking. in Parkinson’s disease. International Journal of Neuro - 2012;7(3):e33513. psychopharmacology 2011;14(7):877–85. Howe WM, Ji J, Parikh V, Williams S, Mocaer E, Trocme- Keenan RM, Hatsukami DK, Pentel PR, Thompson TN, Grillo MA. Pharmacodynamic effects of cotinine in Thibierge C, Sarter M. Enhancement of attentional performance by selective stimulation of alpha4beta2(*) abstinent cigarette smokers. Clinical Pharmacology Neuro- nAChRs: underlying cholinergic mechanisms. 1994;55(5):581–90. and Therapeutics psychopharmacology Kelton MC, Kahn HJ, Conrath CL, Newhouse PA. The 2010;35(6):1391–401. effects of nicotine on Parkinson’s disease. Brain and Hughes JR. Dependence on and abuse of nicotine replace- Cognition Nicotine ment medications. In: Benowitz N, editor. 2000;43(1-3):274–82. Kenny PJ, File SE, Neal MJ. Evidence for a complex influ - Safety and Toxicity . Oxford: Oxford University Press, - 1998:147–57. ence of nicotinic acetylcholine receptors on hippo Journal of Neurochemistry campal serotonin release. Hughes JR, Hatsukami DK. The nicotine withdrawal syn- International Jour- drome: a brief review and update. 2000;75(6):2409–14. nal of Smoking Cessation - 1992;1(2):21–6. Kenny PJ, Markou A. Neurobiology of the nicotine with Pharmacology, Biochemistry and drawal syndrome. Hughes JR, Gust SW, Keenan R, Fenwick JW, Skoog K, 2001;70(4):531–49. Higgins ST. Long-term use of nicotine vs placebo gum. Behavior Konishi H, Wu J, Cooke JP. Chronic exposure to nicotine Archives of Internal Medicine 1991;151(10):1993–8. impairs cholinergic angiogenesis. Hurst R, Rollema H, Bertrand D. Nicotinic acetylcholine Vascular Medicine Pharma- 2010;15(1):47–54. receptors: from basic science to therapeutics. cology and Therapeutics 2013;137(1):22–54. Koob GF, Markou A, Weiss F, Schulteis G. Opponent pro- - Improgo MR, Tapper AR, Gardner PD. Nicotinic acetyl cess and drug dependence: neurobiological mecha - Seminars in Neuroscience 1993;5(5):351–8. nisms. choline receptor-mediated mechanisms in lung cancer. British Journal Krishna K. Tobacco chewing in pregnancy. 2011;82(8):1015–21. Biochemical Pharmacology of Obstetrics and Gynaecology 1978;85(10):726–8. International Agency for Research on Cancer. A Review Kumari V, Postma P. Nicotine use in schizophrenia: the of Human Carcinogens. Part E: Personal Habits and self medication hypotheses. Neuroscience and Biobe- Indoor Combustions. IARC monograph volume 100E. havioral Reviews 2005;29(6):1021–34. Lyon (France): International Agency for Research on Laan M, Bozinovski S, Anderson GP. Cigarette smoke Cancer, 2012. inhibits lipopolysaccharide-induced production of Jarzynka MJ, Guo P, Bar-Joseph I, Hu B, Cheng SY. Estra - inflammatory cytokines by suppressing the activation diol and nicotine exposure enhances A549 bronchio - Jour- of activator protein-1 in bronchial epithelial cells. loalveolar carcinoma xenograft growth in mice through 2004;173(6):4164–70. nal of Immunology the stimulation of angiogenesis. International Journal Lagercrantz H, Bistoletti P. Catecholamine release in 2006;28(2):337–44. of Oncology Pediatric Research the newborn infant at birth. Jin Z, Gao F, Flagg T, Deng X. Nicotine induces multi- 1977;11(8):889–93. site phosphorylation of Bad in association with sup- Lagercrantz H, Slotkin TA. The “stress” of being born. Sci- Journal of Biological Chemistry pression of apoptosis. 1986;254(4):100–7. entific American 2004;279(22):23837–44. Lambers DS, Clark KE. The maternal and fetal physiologic Johnson EO, Chen LS, Breslau N, Hatsukami D, Robbins 1996; effects of nicotine. Seminars in Perinatology T, Saccone NL, Grucza RA, Bierut LJ. Peer smoking and 20(2):115–26. the nicotinic receptor genes: an examination of genetic Landau LI. Tobacco smoke exposure and tracking of lung and environmental risks for nicotine dependence. function into adult life. Paediatric Respiratory Reviews Addiction 2010;105(11):2014–22. 2008;9(1):39–43; quiz –4. Early Joshi S, Kotecha S. Lung growth and development. Larson PS, Haag HB, Finnegan JK. On the relative toxicity Human Development 2007;83(12):789–94. of nicotine and nornicotine. Experimental Biology and Medicine 1945;58(3):231–2. 132 Chapter 5

162 The Health Consequences of Smoking —50 Years of Progress LaVoie EJ, Shigematsu A, Rivenson A, Mu B, Hoffmann D. meta-analysis of case-control studies. Environmental Evaluation of the effects of cotinine and nicotine-N’- and Molecular Mutagenesis 2013;54(2):133–40. oxides on the development of tumors in rats initiated Lødrup Carlsen KC, Jaakkola JJ, Nafstad P, Carlsen KH. In utero exposure to cigarette smoking influences with N-[4-(5-nitro-2-furyl)-2-thiazolyl]formamide. European Respiratory Journal Journal of the National Cancer Institute 1985;75(6): lung function at birth. 1997;10(8):1774–9. 1075–81. - Lazutka FA, Vasiliauskene AP, Gefen Sh G. On the toxi Lopez-Arrieta JM, Rodriguez JL, Sanz F. Nicotine for - cological assessment of the insecticide nicotine sulfate Alzheimer’s disease. Cochrane Database of System 2000, Issue 2. Art No.: CD001749. DOI: [Russian]. Gigiena i Sanitariia 1969;34(5):30–3. atic Reviews 10.1002/14651858.CD001749. Lee HJ, Guo HY, Lee SK, Jeon BH, Jun CD, Lee SK, Park Addictive MH, Kim EC. Effects of nicotine on proliferation, cell Ludwig AM. Pavlov’s “bells” and alcohol craving. Behaviors cycle, and differentiation in immortalized and malig - 1986;11(2):87–91. Journal of Oral Pathology and nant oral keratinocytes. Lunney PC, Leong RW. Review article: Ulcerative colitis, 2005;34(7):436–43. Medicine Alimentary Pharmacol- smoking and nicotine therapy. 2012;36(11-12):997–1008. ogy and Therapeutics Lee J, Cooke JP. The role of nicotine in the pathogenesis 2011;215(2):281–3. Luo J, Ye W, Zendehdel K, Adami J, Adami HO, Boffetta of atherosclerosis. Atherosclerosis Lee J, Cooke JP. Nicotine and pathological angiogenesis. P, Nyren O. Oral use of Swedish moist snuff (snus) 2012;91(21-22):1058–64. Life Sciences and risk for cancer of the mouth, lung, and pancreas Lee W, Ray R, Bergen AW, Swan GE, Thomas P, Tyndale in male construction workers: a retrospective cohort RF, Benowitz NL, Lerman C, Conti DV. DRD1 associa Lancet study. - 2007;369(9578):2015–20. tions with smoking abstinence across slow and normal Maier CR, Hollander MC, Hobbs EA, Dogan I, Linnoila RI, Dennis PA. Nicotine does not enhance tumorigenesis - nicotine metabolizers. Pharmacogenetics and Genom in mutant K-ras-driven mouse models of lung cancer. 2012;22(7):551–4. ics 2011;4(11):1743–51. Cancer Prevention Research Lemay S, Chouinard S, Blanchet P, Masson H, Soland V, Beuter A, Bedard MA. Lack of efficacy of a nicotine Maneckjee R, Minna JD. Opioid and nicotine receptors affect growth regulation of human lung cancer cell transdermal treatment on motor and cognitive deficits Progress in Neuro-Psychophar- in Parkinson’s disease. lines. Proceedings of the National Academy of Sciences macology and Biological Psychiatry of the United States of America 1990;87(9):3294–8. 2004;28(1):31–9. Léna C, Changeux JP. Allosteric nicotinic receptors, Maneckjee R, Minna JD. Opioids induce while nicotine 1998; suppresses apoptosis in human lung cancer cells. human pathologies. Cell Journal of Physiology, Paris 1994;5(10):1033–40. Growth and Differentiation 92(2):63–74. Mansvelder HD, McGehee DS. Long-term potentiation - Léna C, Changeux JP. The role of beta 2-subunit-con of excitatory inputs to brain reward areas by nicotine. taining nicotinic acetylcholine receptors in the brain Neuron 2000;27(2):349–57. explored with a mutant mouse. Annals of the New York Maritz GS. Nicotine and lung development. Birth Defects 1999;868:611–6. Academy of Sciences Research. Part C, Embryo Today 2008;84(1):45–53. Levine DG. “Needle freaks”: compulsive self-injection Maritz GS, Harding R. Life-long programming implica- by drug users. 1974; American Journal of Psychiatry tions of exposure to tobacco smoking and nicotine 131(3):297–300. before and soon after birth: evidence for altered lung Li P, McCollum M, Bracamontes J, Steinbach JH, Akk development. International Journal of Environmental G. Functional characterization of the alpha5(Asn398) Research and Public Health 2011;8(3):875–98. variant associated with risk for nicotine dependence in Markou A, Kosten TR, Koob GF. Neurobiological similari - Molecular the alpha3beta4alpha5 nicotinic receptor. ties in depression and drug dependence: a self-medi - 2011;80(5):818–27. Pharmacology Neuropsychopharmacology cation hypothesis. 1998; Li XW, Wang H. Non-neuronal nicotinic alpha 7 receptor, 18(3):135–74. Life Sci- a new endothelial target for revascularization. Markou A, Weiss F, Gold LH, Caine SB, Schulteis G, Koob ences 2006;78(16):1863–70. Psychopharmacol- GF. Animal models of drug craving. Lindstrom J, Anand R, Gerzanich V, Peng X, Wang F, 1993;112(2-3):163–82. ogy Wells G. Structure and function of neuronal nicotinic Martin JC, Martin DD, Radow B, Day HE. Life span and acetylcholine receptors. Progress in Brain Research pathology in offspring following nicotine and meth - 1996;109:125–37. amphetamine exposure. Experimental Aging Research Liu T, Xie CB, Ma WJ, Chen WQ. Association between 1979;5(6):509–22. CYP2A6 genetic polymorphisms and lung cancer: a 133 Nicotine

163 Surgeon General’s Report Nisell M, Marcus M, Nomikos GG, Svensson TH. Differen - McGehee DS, Role LW. Physiological diversity of nicotinic tial effects of acute and chronic nicotine on dopamine acetylcholine receptors expressed by vertebrate neu- 1995;57:521–46. Annual Review of Physiology rons. output in the core and shell of the rat nucleus accum - - McMaster SK, Paul-Clark MJ, Walters M, Fleet M, Anan bens. Journal of Neural Transmission 1997;104(1): darajah J, Sriskandan S, Mitchell JA. Cigarette smoke 1–10. inhibits macrophage sensing of Gram-negative bacte - Nisell M, Nomikos GG, Svensson TH. Infusion of nicotine ria and lipopolysaccharide: relative roles of nicotine - in the ventral tegmental area or the nucleus accum - bens of the rat differentially affects accumbal dopa British Journal of Pharmacology and oxidant stress. 2008;153(3):536–43. Pharmacology and Toxicology mine release. 1994a; Mensch AR, Holden M. Nicotine overdose from a single 75(6):348–52. Chest 1984;86(5):801–2. piece of nicotine gum. Nisell M, Nomikos GG, Svensson TH. Systemic nicotine- induced dopamine release in the rat nucleus accum - Mizusaki S, Okamoto H, Akiyama A, Fukuhara Y. Relation - bens is regulated by nicotinic receptors in the ventral between chemical constituents of tobacco and muta tegmental area. genic activity of cigarette smoke condensate. Synapse Mutation 1994b;16(1):36–44. 1977;48(3-4):319–25. - O’Brien CP, Childress AR, McLellan T, Ehrman R. Integrat Research Mousa S, Mousa SA. Cellular and molecular mechanisms ing systemic cue exposure with standard treatment in Addictive Behav- recovering drug dependent patients. of nicotine’s pro-angiogenesis activity and its potential 1990;15(4):355–65. impact on cancer. Journal of Cellular Biochemistry iors Okamoto M, Kita T, Okuda H, Nakashima T. Tolerance 2006;97(6):1370–8. - Murphy SE, von Weymarn LB, Schutten MM, Kassie F, to the convulsions induced by daily nicotine treat ment in rats. Japanese Journal of Pharmacology 1992; Modiano JF. Chronic nicotine consumption does not influence 4-(methylnitrosamino)-1-(3-pyridyl)-1-buta- 59(4):449–55. none-induced lung tumorigenesis. Cancer Prevention Okamoto M, Kita T, Okuda H, Tanaka T, Nakashima T. Effects of aging on acute toxicity of nicotine in rats. Research 2011;4(11):1752–60. Pharmacology and Toxicology Murray RP, Connett JE, Zapawa LM. Does nicotine replace 1994;75(1):1–6. - O’Reilly EJ, McCullough ML, Chao A, Henley SJ, Calle EE, ment therapy cause cancer? Evidence from the Lung Thun MJ, Ascherio A. Smokeless tobacco use and the Nicotine & Tobacco Research 2009; Health Study. Movement Disor- risk of Parkinson’s disease mortality. 11(9):1076–82. ders 2005;20(10):1383–4. Musso F, Bettermann F, Vucurevic G, Stoeter P, Konrad A, - Paleari L, Catassi A, Ciarlo M, Cavalieri Z, Bruzzo C, Ser Winterer G. Smoking impacts on prefrontal attentional vent D, Cesario A, Chessa L, Cilli M, Piccardi F, et al. Psychophar- network function in young adult brains. Role of alpha7-nicotinic acetylcholine receptor in macology 2007;191(1):159–69. human non-small cell lung cancer proliferation. Cell National Institute on Drug Abuse. Points to Consider 2008;41(6):936–59. Proliferation - Regarding Tobacco Industry Funding of NIDA Appli - Park SB, Jacob P 3rd, Benowitz NL, Cashman JR. Stere cants, n.d.; ; accessed: August 7, 2013. Perry DC, Davila-Garcia MI, Stockmeier CA, Kellar KJ. Newhouse P, Singh A, Potter A. Nicotine and nicotinic Increased nicotinic receptors in brains from smokers: receptor involvement in neuropsychiatric disorders. membrane binding and autoradiography studies. Jour- Current Topics in Medicinal Chemistry 2004;4(3): nal of Pharmacology and Experimental Therapeutics 267–82. 1999;289(3):1545–52. Ng MK, Wu J, Chang E, Wang BY, Katzenberg-Clark R, Peters R, Poulter R, Warner J, Beckett N, Burch L, Bul- Ishii-Watabe A, Cooke JP. A central role for nicotinic pitt C. Smoking, dementia and cognitive decline in the cholinergic regulation of growth factor-induced endo- BMC Geriatrics elderly, a systematic review. 2008;8:36. Arteriosclerosis, Thrombosis, thelial cell migration. Philip Morris International. Philip Morris International 2007;27(1):106–12. and Vascular Biology (PMI) enters into a patent purchase agreement of new Nikfar S, Ehteshami-Ashar S, Rahimi R, Abdollahi M. Sys- technology with the potential to reduce the harm of tematic review and meta-analysis of the efficacy and smoking [press release], May 26, 2011; ; accessed: July 13, 2013.>. 134 Chapter 5

164 —50 Years of Progress The Health Consequences of Smoking Picciotto MR, Corrigall WA. Neuronal systems underly- Riebe M, Westphal K. Studies on the induction of sister- - ing behaviors related to nicotine addiction: neural cir chromatid exchanges in Chinese hamster ovary cells 1983; cuits and molecular genetics. Journal of Neuroscience by various tobacco alkaloids. Mutation Research 2002;22(9):3338–41. 124(3-4):281–6. Riebe M, Westphal K, Fortnagel P. Mutagenicity testing, in Pidoplichko VI, DeBiasi M, Williams JT, Dani JA. Nicotine activates and desensitizes midbrain dopamine neurons. bacterial test systems, of some constituents of tobacco. 1982;101(1):39–43. Nature Mutation Research 1997;390(6658):401–4. Pierce RA, Nguyen NM. Prenatal nicotine exposure and Ritz B, Ascherio A, Checkoway H, Marder KS, Nelson LM, American Journal of Respira- Rocca WA, Ross GW, Strickland D, Van Den Eeden SK, abnormal lung function. Gorell J. Pooled analysis of tobacco use and risk of tory Cell and Molecular Biology 2002;26(1):10–3. Polesskaya OO, Fryxell KJ, Merchant AD, Locklear LL, Ker Parkinson disease. Archives of Neurology 2007;64(7): 990–7. KF, McDonald CG, Eppolito AK, Smith LN, Wheeler Role LW, Berg DK. Nicotinic receptors in the devel - TL, Smith RF. Nicotine causes age-dependent changes in gene expression in the adolescent female rat brain. Neuron opment and modulation of CNS synapses. Neurotoxicology and Teratology 1996;16(6):1077–85. 2007;29(1):126–40. Rose JE, Behm FM, Westman EC, Coleman RE. Arterial Poorthuis RB, Bloem B, Schak B, Wester J, de Kock CP, Mansvelder HD. Layer-specific modulation of the pre - nicotine kinetics during cigarette smoking and intra- venous nicotine administration: implications for addic frontal cortex by nicotinic acetylcholine receptors. - tion. Drug and Alcohol Dependence 2013a;23(1):148–61. Cerebral Cortex 1999;56(2):99–107. Poorthuis RB, Bloem B, Verhoog MB, Mansvelder HD. Nicotine Addiction Royal College of Physicians of London. Layer-specific interference with cholinergic signaling in Britain: A Report of the Tobacco Advisory Group of the Royal College of Physicians in the prefrontal cortex by smoking concentrations . London: Royal College of nicotine. Journal of Neuroscience 2013b;33(11): of Physicians of London, 2000. 4843–53. Russo P, Cesario A, Rutella S, Veronesi G, Spaggiari L, Poorthuis RB, Goriounova NA, Couey JJ, Mansvelder HD. Galetta D, Margaritora S, Granone P, Greenberg DS. Nicotinic actions on neuronal networks for cognition: Impact of genetic variability in nicotinic acetylcholine - general principles and long-term consequences. receptors on nicotine addiction and smoking cessa Bio- 2011; tion treatment. Current Medicinal Chemistry chemical Pharmacology 2009;78(7):668–76. 18(1):91–112. Prochaska JJ, Hall SM, Bero LA. Tobacco use among Sandberg K, Poole SD, Hamdan A, Arbogast P, Sundell individuals with schizophrenia: what role has the - HW. Altered lung development after prenatal nico tobacco industry played? 2008; Schizophrenia Bulletin Pediatric Research tine exposure in young lambs. 34(3):555–67. 2004;56(3):432–9. Punnoose S, Belgamwar MR. Nicotine for schizophre - Sandberg KL, Pinkerton KE, Poole SD, Minton PA, Sun- 2006, Cochrane Database of Systematic Reviews nia. dell HW. Fetal nicotine exposure increases airway Issue 1. Art No.: CD004838. DOI: 10.1002/14651858. responsiveness and alters airway wall composition in CD004838.pub2. young lambs. Respiratory Physiology & Neurobiology Purnell C, Gao S, Callahan CM, Hendrie HC. Cardiovascu - 2011;176(1-2):57–67. lar risk factors and incident Alzheimer disease: a sys- Sarginson JE, Killen JD, Lazzeroni LC, Fortmann SP, tematic review of the literature. Alzheimer Disease and Ryan HS, Schatzberg AF, Murphy GM Jr. Markers in Associated Disorders 2009;23(1):1–10. the 15q24 nicotinic receptor subunit gene cluster R.J. Reynolds Tobacco Company. How important is smoker (CHRNA5-A3-B4) predict severity of nicotine addiction understanding and what are our learning needs? 1993. and response to smoking cessation therapy. American RJ Reynolds Collection. Bates No. 510932130/2136. Journal of Medical Genetics. Part B: Neuropsychiatric . 2011;156B(3):275–84. Genetics Reid MS, Fox L, Ho LB, Berger SP. Nicotine stimulation Sawnani H, Jackson T, Murphy T, Beckerman R, Simaka - - of extracellular glutamate levels in the nucleus accum - jornboon N. The effect of maternal smoking on respi Synapse bens: neuropharmacological characterization. ratory and arousal patterns in preterm infants during 2000;35(2):129–36. sleep. American Journal of Respiratory and Critical - Richardson HL, Walker AM, Horne RS. Maternal smok Care Medicine 2004;169(6):733–8. ing impairs arousal patterns in sleeping infants. Sleep 2009;32(4):515–21. 135 Nicotine

165 Surgeon General’s Report Sorice R, Bione S, Sansanelli S, Ulivi S, Athanasakis E, Schilström B, Svensson HM, Svensson TH, Nomikos GG. Lanzara C, Nutile T, Sala C, Camaschella C, D’Adamo Nicotine and food induced dopamine release in the nucleus accumbens of the rat: putative role of alpha7 P, et al. Association of a variant in the CHRNA5-A3-B4 nicotinic receptors in the ventral tegmental area. Neu- gene cluster region to heavy smoking in the Italian population. 1998;85(4):1005–9. roscience European Journal of Human Genetics Schneider J, Mitchell I, Singhal N, Kirk V, Hasan SU. Pre- 2011;19(5):593–6. natal cigarette smoke exposure attenuates recovery Stalhandske T, Slanina P. Lethal brain concentrations of Ameri- from hypoxemic challenge in preterm infants. nicotine in mice of different ages. Acta Pharmacolog- 1970;28(3):233–40. can Journal of Respiratory and Critical Care Medicine ica et Toxicologica Stepanov I, Carmella SG, Briggs A, Hertsgaard L, Lind 2008;178(5):520–6. - Schochet TL, Kelley AE, Landry CF. Differential expression - gren B, Hatsukami D, Hecht SS. Presence of the carcin of arc mRNA and other plasticity-related genes induced ogen N’-nitrosonornicotine in the urine of some users of oral nicotine replacement therapy products. by nicotine in adolescent rat forebrain. Neuroscience Cancer 2009a;69(21):8236–40. 2005;135(1):285–97. Research Schuller HM. Is cancer triggered by altered signalling Stepanov I, Carmella SG, Han S, Pinto A, Strasser AA, Nature Reviews: of nicotinic acetylcholine receptors? - Lerman C, Hecht SS. Evidence for endogenous for 2009;9(3):195–205. mation of N’-nitrosonornicotine in some long-term Cancer Schuller HM, McGavin MD, Orloff M, Riechert A, Porter Nicotine & Tobacco Research nicotine patch users. 2009b;11(1):99–105. B. Simultaneous exposure to nicotine and hyperoxia Stepanov I, Jensen J, Hatsukami D, Hecht SS. New and causes tumors in hamsters. Laboratory Investigation traditional smokeless tobacco: comparison of toxicant 1995;73(3):448–56. - Nicotine & Tobacco Research and carcinogen levels. Sekhon HS, Jia Y, Raab R, Kuryatov A, Pankow JF, Whit 2008;10(12):1773–82. sett JA, Lindstrom J, Spindel ER. Prenatal nicotine increases pulmonary alpha7 nicotinic receptor expres Stewart J, de Wit H, Eikelboom R. Role of unconditioned - and conditioned drug effects in the self-administra - sion and alters fetal lung development in monkeys. Psychological Review 1999;103(5):637–47. Journal of Clinical Investigation tion of opiates and stimulants. Sekhon HS, Keller JA, Benowitz NL, Spindel ER. Prenatal 1984;91(2):251–68. - Stitzer ML, de Wit H. Abuse liability of nicotine. In: Ben nicotine exposure alters pulmonary function in new - born rhesus monkeys. . New Nicotine Safety and Toxicity owitz NL, editor. American Journal of Respira- tory and Critical Care Medicine 2001;164(6):989–94. York: Oxford University Press, 1998:119–31. Stocks J, Dezateux C. The effect of parental smoking on Sekhon HS, Keller JA, Proskocil BJ, Martin EL, Spindel Respi- lung function and development during infancy. ER. Maternal nicotine exposure upregulates collagen 2003;8(3):266–85. rology gene expression in fetal monkey lung. Association Stolerman IP, Jarvis MJ. The scientific case that nicotine Ameri- with alpha7 nicotinic acetylcholine receptors. Psychopharmacology 1995;117(1):2–10; is addictive. can Journal of Respiratory Cell and Molecular Biology discussion 4–20. 2002;26(1):31–41. Stratton K, Shetty P, Wallace R, Bondurant S, editors. In: Shah NR, Bracken MB. A systematic review and meta- Clearing the Smoke: Assessing the Science Base for analysis of prospective studies on the association Washington: National Acad- Tobacco Harm Reduction. between maternal cigarette smoking and preterm emies Press, 2001. delivery. American Journal of Obstetrics and Gynecol - Swan GE, Lessov-Schlaggar CN. The effects of tobacco ogy 2000;182(2):465–72. Neuro- smoke and nicotine on cognition and the brain. Singh A, Potter A, Newhouse P. Nicotinic acetylcho- psychology Review 2007;17(3):259–73. line receptor system and neuropsychiatric disorders. Tager IB, Ngo L, Hanrahan JP. Maternal smoking during 2004;7(12):1096–103. IDrugs pregnancy. Effects on lung function during the first 18 Slotkin TA. Fetal nicotine or cocaine exposure: which one American Journal of Respiratory and months of life. Journal of Pharmacology and Experimental is worse? Critical Care Medicine 1995;152(3):977–83. 1998;285(3):931–45. Therapeutics Tanner CM, Goldman SM, Aston DA, Ottman R, Ellenberg Slotkin TA, Seidler FJ. Adrenomedullary catecholamine J, Mayeux R, Langston JW. Smoking and Parkinson’s - release in the fetus and newborn: secretory mecha Neurology disease in twins. 2002;58(4):581–8. Journal of nisms and their role in stress and survival. Tepper JM, Wilson JR, Schlesinger K. Relations between 1988;10(1):1–16. Developmental Physiology nicotine-induced convulsive behavior and blood and 136 Chapter 5

166 The Health Consequences of Smoking —50 Years of Progress The U.S. Department of Health and Human Services. brain levels of nicotine as a function of sex and age in two inbred strains of mice. Health Consequences of Smoking: A Report of the Sur - Pharmacology, Biochemis- . Atlanta (GA): U.S. Department of Health 1979;10(3):349–53. geon General try and Behavior Thorgeirsson TE, Geller F, Sulem P, Rafnar T, Wiste A, and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Pre - Magnusson KP, Manolescu A, Thorleifsson G, Ste- fansson H, Ingason A, et al. A variant associated with vention and Health Promotion, Office on Smoking and nicotine dependence, lung cancer and peripheral arte - Health, 2004. Nature rial disease. 2008;452(7187):638–42. U.S. Department of Health and Human Services. The Timofeeva MN, McKay JD, Smith GD, Johansson M, Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General . Byrnes GB, Chabrier A, Relton C, Ueland PM, Vollset Atlanta (GA): U.S. Department of Health and Human SE, Midttun O, et al. Genetic polymorphisms in 15q25 and 19q13 loci, cotinine levels, and risk of lung cancer Services, Centers for Disease Control and Prevention, Cancer Epidemiology, Biomarkers and Pre - Coordinating Center for Health Promotion, National in EPIC. - 2011;20(10):2250–61. vention Center for Chronic Disease Prevention and Health Pro motion, Office on Smoking and Health, 2006. - Toide K, Arima T. Effects of cholinergic drugs on extracel lular levels of acetylcholine and choline in rat cortex, U.S. Department of Health and Human Services. How hippocampus and striatum studied by brain dialy - Tobacco Smoke Causes Disease—The Biology and 1989;173(2- sis. Behavioral Basis for Smoking-Attributable Disease: European Journal of Pharmacology . Atlanta (GA): U.S. 3):133–41. A Report of the Surgeon General Tong VT, Dietz PM, Morrow B, D’Angelo DV, Farr SL, Department of Health and Human Services, Centers Rockhill KM, England LJ. Trends in smoking before, for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, during, and after pregnancy—Pregnancy Risk Assess - Office on Smoking and Health, 2010. ment Monitoring System (PRAMS), United States, 40 sites, 2000–2010. Morbidity and Mortality Weekly Pre- U.S. Department of Health and Human Services. . 2013;62(SS-6):1–19. Report venting Tobacco Use Among Youth and Young Adults: . Atlanta (GA): U.S. A Report of the Surgeon General Trauth JA, Seidler FJ, Ali SF, Slotkin TA. Adolescent nic - Department of Health and Human Services, Centers otine exposure produces immediate and long-term for Disease Control and Prevention, National Center changes in CNS noradrenergic and dopaminergic func - tion. Brain Research 2001;892(2):269–80. for Chronic Disease Prevention and Health Promotion, - Office on Smoking and Health, 2012. Trauth JA, Seidler FJ, McCook EC, Slotkin TA. Adoles cent nicotine exposure causes persistent upregulation Vagg R, Chapman S. Nicotine analogues: a review of 2005; Addiction tobacco industry research interests. of nicotinic cholinergic receptors in rat brain regions. 100(5):701–12. 1999;851(1-2):9–19. Brain Research Vieregge A, Sieberer M, Jacobs H, Hagenah JM, Vieregge Trauth JA, Seidler FJ, Slotkin TA. An animal model of ado- P. Transdermal nicotine in PD: a randomized, dou - lescent nicotine exposure: effects on gene expression Neurology ble-blind, placebo-controlled study. 2001; and macromolecular constituents in rat brain regions. 57(6):1032–5. 2000;867(1-2):29–39. Brain Research Villablanca AC. Nicotine stimulates DNA synthesis and - Trivedi AH, Dave BJ, Adhvaryu SG. Assessment of geno proliferation in vascular endothelial cells in vitro. Jour- toxicity of nicotine employing in vitro mammalian test 1998;84(6):2089–98. nal of Applied Physiology Cancer Letters 1990;54(1-2):89–94. system. Waldum HL, Nilsen OG, Nilsen T, Rorvik H, Syversen V, Turner C, Spilich GJ. Research into smoking or nicotine and Sanvik AK, Haugen OA, Torp SH, Brenna E. Long- - human cognitive performance: does the source of fund Life Sciences 1996; term effects of inhaled nicotine. Addiction 1997;92(11):1423–6. ing make a difference? 58(16):1339–46. The U.S. Department of Health and Human Services. Wallace TL, Bertrand D. Importance of the nicotinic ace - - Health Consequences of Smoking: Nicotine Addic tylcholine receptor system in the prefrontal cortex. tion. A Report of the Surgeon General . Atlanta (GA): 2013;85(12):1713–20. Biochemical Pharmacology U.S. Department of Health and Human Services, Public - Wassenaar CA, Dong Q, Wei Q, Amos CI, Spitz MR, Tyn Health Service, Centers for Disease Control, National dale RF. Relationship between CYP2A6 and CHRNA5- Center for Chronic Disease Prevention and Health Pro - CHRNA3-CHRNB4 variation and smoking behaviors motion, Office on Smoking and Health, 1988. DHHS and lung cancer risk. Journal of the National Cancer Publication No. (CDC) 88-8406. Institute 2011;103(17):1342–6. 137 Nicotine

167 Surgeon General’s Report Watkins SS, Stinus L, Koob GF, Markou A. Reward and World Health Organization. WHO Study Group on somatic changes during precipitated nicotine with Tobacco Product Regulation: Report on the Scientific - drawal in rats: centrally and peripherally mediated Basis of Tobacco Product Regulation. Fourth Report WHO Technical Report Series of a WHO Study Group. effects. Journal of Pharmacology and Experimental 967. Geneva (Switzerland): World Health Organization, Therapeutics 2000;292(3):1053–64. 2012. West KA, Brognard J, Clark AS, Linnoila IR, Yang X, Swain Xin M, Deng X. Nicotine inactivation of the proapoptotic SM, Harris C, Belinsky S, Dennis PA. Rapid Akt activa - Journal of function of Bax through phosphorylation. tion by nicotine and a tobacco carcinogen modulates the phenotype of normal human airway epithelial cells. Biological Chemistry 2005;280(11):10781–9. 2003;111(1):81–90. Neonatal and Pediatric Pharmacol- Journal of Clinical Investigation Yaffe SJ, Aranda JV. Addic- . 4th ed. Phila- ogy: Therapeutic Principles in Practice West R. Beneficial effects of nicotine: fact or fiction? delphia: Lippincott, Wiilliams, and Wilkins, 2010. 1993;88(5):589–90. tion West RJ, Russell MA. Effects of withdrawal from long- Yamamoto I, Nagai K, Inoki R. The contents of DOPA and catecholamines in several rat tissues and nicotine- 1985; Psychological Medicine term nicotine gum use. Japanese Journal of Pharmacol- 15(4):891–3. induced convulsions. Whitehouse PJ, Price DL, Struble RG, Clark AW, Coyle 1966;16(3):295–305. ogy JT, Delon MR. Alzheimer’s disease and senile demen Yim SH, Hee SS. Genotoxicity of nicotine and cotinine - in the bacterial luminescence test. tia: loss of neurons in the basal forebrain. Science Mutation Research 1982;215(4537):1237–9. 1995;335(3):275–83. Wikler A. Dynamics of drug dependence. Implications Yuen ST, Gogo AR Jr, Luk IS, Cho CH, Ho JC, Loh TT. The of a conditioning theory for research and treatment. - effect of nicotine and its interaction with carbon tetra Pharmacology and Toxicology Archives of General Psychiatry 1973;28(5):611–6. chloride in the rat liver. Wikström AK, Cnattingius S, Stephansson O. Maternal 1995;77(3):225–30. - Zeller WJ, Berger MR. Nicotine and estrogen metabo use of Swedish snuff (snus) and risk of stillbirth. Epide- miology 2010;21(6):772–8. lism—possible implications of smoking for growth and outcome of treatment of hormone-dependent cancer? Wilkie GI, Hutson P, Sullivan JP, Wonnacott S. Pharma- cological characterization of a nicotinic autoreceptor Discussion of experimental results. Journal of Cancer Research and Clinical Oncology 1989;115(6):601–3. Neurochemical in rat hippocampal synaptosomes. Research 1996;21(9):1141–8. Zheng Y, Ritzenthaler JD, Roman J, Han S. Nicotine stimulates human lung cancer cell growth by inducing Wirdefeldt K, Adami HO, Cole P, Trichopoulos D, Mandel - American Journal of Respira fibronectin expression. J. Epidemiology and etiology of Parkinson’s disease: a 2007;37(6):681–90. - tory Cell and Molecular Biology review of the evidence. European Journal of Epidemi Zhu AZ, Renner CC, Hatsukami DK, Swan GE, Lerman C, 2011;26(Suppl 1):S1–58. ology Benowitz NL, Tyndale RF. The ability of plasma cotinine Wirdefeldt K, Gatz M, Pawitan Y, Pedersen NL. Risk and to predict nicotine and carcinogen exposure is altered protective factors for Parkinson’s disease: a study in 2005;57(1):27–33. Annals of Neurology by differences in CYP2A6: the influence of genetics, Swedish twins. Cancer Epidemiology, Biomarkers and race, and sex. - World Health Organization. International Drug Monitor ing: The Role of National Centres. Report of a WHO Prevention 2013;22(4):708–18. Geneva (Switzerland): World Health Organi- Meeting. zation, 1972. WHO Technical Report Series No. 498. 138 Chapter 5

168 Chapter 6 Cancer Introduction 143 148 Mechanisms of Cancer Induction by Tobacco Smoke Summary 151 Changing Cigarettes and Risk for Lung Cancer Over Time 151 152 Changes in Cigarettes Over the Past Several Decades 153 Changes in Design, Curing, and Composition 154 Differences Across Brands in Toxicant Yields Changes in Tobacco-Specific Nitrosamine and Benzo[ a ]pyrene Levels Over Time 155 155 Differences in Toxicant Yields Across Countries Low-Tar Cigarettes Do Not Reduce Risk of Lung Cancer 157 158 Overall Death Rates for Lung Cancer Indicate Increased Risk of Smoking in Recent Decades Changes Over Time in the Types of Lung Cancer Associated With Smoking 161 Evidence for a Rising Risk of Adenocarcinoma of the Lung in the United States 177 185 Evidence Synthesis 186 Conclusions Implications 186 Liver Cancer 187 187 Conclusions of Previous Surgeon General’s Reports Biologic Basis 188 Epidemiologic Evidence 188 Evidence Synthesis 196 Conclusion 196 196 Implications 197 Colorectal Cancer 198 Conclusions from Previous Surgeon General’s Reports Biologic Basis 198 Description of the Literature Review 199 Epidemiologic Evidence 199 Adenomatous Polyps 199 Colon and Rectal Cancer 199 Evidence Synthesis 203 Conclusion 204 Implications 204 Prostate Cancer 204 Conclusions from Previous Surgeon General’s Reports 205 Biologic Basis 205 206 Description of the Literature Review 139

169 Epidemiologic Evidence 206 206 Incidence and Mortality Stage and Histologic Grade 208 Progression, Case Fatality, and All-Cause Mortality 208 Evidence Synthesis 209 209 Conclusions 209 Implications Breast Cancer 209 210 Biologic Basis—Evidence for Potential Etiologic Mechanisms 210 DNA Adducts Other Cellular Mechanisms 211 Hormones 212 214 Summary Epidemiologic Evidence—Overview 214 215 Active Cigarette Smoking and Risk for Breast Cancer 217 Conclusions from Previous Surgeon General’s Reports 217 Cohort Studies 221 Case-control Studies Adjustment for Selected Covariates 224 225 Meta-analysis of Breast Cancer Risk Associated with Measures of Active Smoking 226 Ever Smoking Cigarette Smoking Status 231 233 Duration of Cigarette Smoking 236 Cigarettes Smoked Per Day Pack-years of Cigarette Smoking 237 Timing of Exposure to Tobacco Smoke 240 242 Menopausal Status 245 Hormone Receptor Status Exposure to Tobacco Smoke and Risk of Second Primary Contralateral Breast Cancer 247 Genetic Susceptibility to Smoking 247 248 Family History BRCA1/BRCA2 248 Carcinogen Metabolism 249 253 Genetic Susceptibility—Summary Summary and Review of Active Cigarette Smoking 254 Major Summary Points for Active Smoking 254 Exposure to Secondhand Smoke and Risk for Breast Cancer 255 Conclusions from Previous Surgeon General’s Reports 257 Cohort Studies 258 260 Case-Control Studies Meta-Analysis of Breast Cancer Risk Associated with Measures of Secondhand Smoke 263 Measures of Exposure to Secondhand Smoke 263 Adjustment for Selected Covariates 265 Most Comprehensive Measures of Passive Smoking 265 Other Categories of Passive Exposure 272 In Utero Exposure to Secondhand Smoke 272 Secondhand Smoke Exposure and Genotype Interaction 274 Summary and Review of Exposure to Secondhand Smoke 274 275 Major Summary Points on Passive Smoking 140

170 Exposure to Tobacco Smoke and Breast Cancer Mortality 275 276 Active Smoking 277 Duration and Intensity of Smoking Hormone Receptor Status 277 277 Exposure to Secondhand Smoke 278 Summary of Exposure to Tobacco Smoke and Breast Cancer Mortality Evidence Synthesis 278 Methodological Issues 278 281 Criteria for Causal Inference Conclusions 283 283 Implications Adverse Health Outcomes in Cancer Patients and Survivors 284 Conclusions of Previous Surgeon General’s Reports 284 284 Biologic Basis 285 Epidemiologic and Clinical Evidence Literature Search and Other Methodologic Considerations 285 Cigarette Smoking and All-Cause Mortality in Cancer Patients 286 Cigarette Smoking and Overall Survival in Cancer Patients 287 Cigarette Smoking and Cancer Mortality in Cancer Patients 287 Cigarette Smoking and Risk of Second Primary Cancers in Cancer Patients 287 288 Cigarette Smoking and Recurrence and Response to Treatment in Cancer Patients Cigarette Smoking and Toxicity Associated with Cancer Treatment 288 Evidence Synthesis 289 Conclusions 291 Implications 291 Evidence Summary 292 Chapter Conclusions 293 Lung Cancer 293 Liver Cancer 293 Colorectal Cancer 293 Prostate Cancer 293 Breast Cancer 293 Adverse Health Outcomes in Cancer Patients and Survivors 293 294 References 141

171

172 The Health Consequences of Smoking —50 Years of Progress Sections of this chapter on the health consequences of smoking are accompanied by evidence tables detailing the studies that were used to evaluate the evidence to assess causality. A supplement to this report is provided that ” where they are called out in S contains these tables. The tables included in the supplement are indicated with an “ the text. Introduction “initiators,” causing permanent changes in cells, while The signature finding of the landmark 1964 Sur - others are “promoters” of the carcinogenic process. The Smoking and Health , was the con - geon General’s report, structure of DNA and the genetic code were identified, but clusion that cigarette smoking was a cause of lung cancer - in men (U.S. Department of Health, Education, and Wel research on DNA, mutations, and cancer was just starting (Table 6.1). Of course, many processes now considered to fare [USDHEW] 1964). At that time, cancer was a highly - feared disease with limited therapeutic options (Mukher - be critical in carcinogenesis (e.g., those involving onco jee 2010). Surgery and radiation therapy were essentially genes, tumor suppressor genes, and epigenetics) had not the only treatment options, as chemotherapy was in its yet been discovered. - Figures 4.3 and 4.4 document trends in cancer mor infancy. The efficacy of chemotherapy for childhood acute lymphoblastic leukemia and for testicular cancer had not tality among men and women for the period 1930–2010 yet been established (Proctor 1995). Chemoprevention, (American Cancer Society [ACS] 2013). However, mortal- ity does not capture the full picture of cancer occurrence, as now used for breast cancer, for example, had not been since it matches incidence (i.e., the occurrence of new - implemented. Screening was employed for only one dis ease, cervical cancer, using the Papanicolaou (Pap) smear. cases) for only those malignancies for which survival is The first trial of mammographic screening for breast can- very poor. For lung cancer, given a 5-year survival rate of around 15%, incidence and death rates are close. In 1964, cer, the Health Insurance Plan (HIP) study, had just been lung cancer was the leading cause of cancer deaths in men, launched (Mukherjee 2010). Many of the most critical having passed colorectal cancer about a decade previously. advances in mechanistic understanding that are relevant Death rates for stomach cancer had declined steadily in to prevention and treatment today had yet to arrive (Table uterine (corpus and cervix) 6.1) (DeVita and Rosenberg 2012). men and women, as had the - From the perspective of 2014, the understanding cancer mortality rate for women. The lung cancer mortal ity rate in 1964 for women was just starting its upward 50 years ago of the pathogenesis and etiology of cancer trajectory. Figure 4.3 charts the continuing course of was also quite limited (Figure 6.1) (DeVita and Rosenberg lung cancer death rates, showing an eventual plateau and 2012). Radiation was a long-established cause of multiple decline in men. Figure 4.4 shows a long upward course types of cancer; the increased risk of lung cancer in radon- and then the beginning of a decline in women. exposed uranium miners was established; and follow-up Overall, cancer survival has also improved in the of the atomic bomb survivors had documented their increased risk of acute leukemia. Clinical experience and United States. In 1953, relative 5-year survival for people with cancer was only 35% (DeVita and Rosenberg 2012). epidemiologic studies were documenting links between By 1977, the figure was 49% and the most recent data occupational exposures, including asbestos and nickel from the National Cancer Institute’s (NCI’s) Surveillance, oxides, and cancer. The wave of epidemiologic studies that Epidemiology, and End Results (SEER) Program for cases focused on lifestyle and risk of cancer was just starting, diagnosed between 2003–2009 and followed through 2010 and relatively little attention was given to viruses and bac- was 68% (NCI 2013). teria as causes of cancer. - The process of carcinogenesis was commonly under Since 1973, the incidence of cancer has been tracked stood as prolonged and involving multiple stages, leading in some states and metropolitan areas through the SEER to uncontrolled cell replication (Armitage and Doll 1954; Program. Figures 6.2 and 6.3 show trends for age-adjusted - Shimkin 1977). The 1964 Surgeon General report’s discus - incidence of cigarette-caused cancers across the span cov sion of carcinogenesis referred to “...a slow multi-stage ered by the SEER data among men and women. Among process” (p. 142) and pointed out that some chemicals are men, incidence rates of lung, colorectal, oropharyngeal, 143 Cancer

173 Surgeon General’s Report Table 6.1 Singular discoveries and major events in the cancer field and changing relative survival rates for persons with cancer in the United States, 1863–2006 Year Relative 5-year survival rate Discovery or event 1863 Cellular origin of cancer (Virchow) Seed-and-soil hypothesis (Paget) 1889 Chromosomal mutations in cancer (Boveri) 1914 1937 Founding of the National Cancer Institute 1944 Transmission of cellular information by DNA (Avery) 1950 Availability of cancer drugs through CCNSC 1953 35% Report on structure of DNA 1961 Breaking of the genetic code 1970 Reverse transcriptase Restriction enzymes 1971 Passage of National Cancer Act of 1971 Hybridomas and monoclonal antibodies 50% 1975 Tracking of cancer statistics by SEER Program 1976 Cellular origin of retroviral oncogenes 1979 Epidermal growth factor and receptor Suppression of tumor growth by 1981 P53 1982 RAS oncogenes Discovery of 1984 G proteins and cell signaling 1986 Retinoblastoma gene 1990 First decrease in cancer incidence and mortality 1991 Association between mutation in APC gene and colorectal cancer Genetic cancer syndromes 1994 BRCA1 and breast cancer Association between Sequencing of the human genome 2000 Epigenetics in cancer 2002 Micro-RNAs in cancer First decrease in total number of deaths from cancer 2005 68% 2006 Tumor stromal interaction Source: Adapted from DeVita and Rosenberg 2012 using data from Chang et al. 1982. Reprinted with permission from Massachusetts , © 2012. Medical Society SEER = Surveillance, Epidemiology, and End Results Program of the Note : CCNSC = Cancer Chemotherapy National Service Center; National Cancer Institute. Chapter 6 144

174 —50 Years of Progress The Health Consequences of Smoking Figure 6.1 Timeline of pivotal events in cancer prevention HPV vaccine developed, 1985 Viruses and cancer Vaccine prevents Chemoprevention hepatitis and hepatoma, Tamoxifen prevention trials, 1981 Tobacco and cancer 1989 Link discovered between HPV and cervical cancer, Proof of principle: 1976 chemoprevention works, 1990 Hepatitis linked to hepatoma, 1974 Lung-cancer incidence and mortality begin to fall, First vaccine against hepatitis B, 1990–1991 1974 BCG prevents Tobacco advertising on radio and bladder cancer, television banned in U.S., 1991 1970 Antiestrogen drugs prevent DCIS, Hepatitis B discovered, 1995 1967 Tamoxifen reduces Tamoxifen discovered, breast cancer 1967 incidence, 1998 Warning labels on Hypothesis that tobacco FDA approves HPV is linked to lung cancer, cigarette packages, vaccine to prevent 1912 1965 cervical cancer, 2000 Surgeon General’s report Finasteride on risks of smoking, reduces prostate HPV discovered, 1964 cancer incidence, 1907 2003 Experimental evidence Aspirin prevents links lung cancer to smoking, colon cancer, 1950 2003 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020 Incidence per 100,000 444 473 511 400 (age adjusted) Mortality 200 215 178 151 per 100,000 (age adjusted) Source: DeVita and Rosenberg 2012. Reprinted with permission from Massachusetts Medical Society, © 2012. = ductal carcinoma in situ; DCIS = bacille Calmette-Guérin; BCG Notes: = human FDA HPV = U.S. Food and Drug Administration; papilloma virus. Cancer 145

175 Surgeon General’s Report ongoing assessment of progress in cancer control. The stomach, and laryngeal cancers have declined over time, but rates for kidney and liver cancers continue to rise. most recent report reveals a decline in the incidence of The trend is similar among women, with the exception of lung cancer for both men and women in the first decade of lung cancer for which incidence rates increased in the two the twenty-first century (Jemal et al. 2013). For men, the rate declined by 2.0% annually during this decade, while decades since 1975, and reached a plateau since the mid- 1990s, before declining in 2007 ( the annual decline was 0.2% for women. . Howlader et al. 2013) In addition to the SEER areas, the rest of the nation and This chapter reviews the evidence on smoking - the District of Columbia are covered by the National Pro and cancer for malignancies for which the evidence was gram of Cancer Registries (NPCR) of the Centers for Dis previously found to be inadequate or was insufficient - to reach a causal conclusion. Specifically, four cancer Annual Report ease Control and Prevention (CDC). The , a collaborative breast, colon and rectum, liver, and to the Nation on the Status of Cancer sites are covered  publication by the ACS, the North American Association prostate  and also the changing cigarette and risk for lung cancer over time. The chapter also covers the rela of Central Cancer Registries, CDC, and NCI, provides an - Figure 6.2 Surveillance, Epidemiology, and End Results (SEER) age-adjusted incidence, selected sites, males, 1975–2010 100 Lung Colorectal Bladder Kidney Oropharynx Pancreas Liver 10 Stomach Esophagus Larynx AML Incidence rate per 100,000 (log scale) 1 1995 2000 2005 2010 1975 1980 1985 1990 Year of diagnosis Howlader et al. 2013. Source: The data are for nine SEER areas (San Francisco, Connecticut, Detroit, Hawaii, Iowa, New Mexico, Seattle, Utah, and Atlanta). Note: = acute AML Rates are per 100,000 and are age-adjusted to the 2000 U.S. standard population (19 age groups − Census P25-1130). myeloid leukemia. 146 Chapter 6

176 The Health Consequences of Smoking —50 Years of Progress Surveillance, Epidemiology, and End Results (SEER) age-adjusted incidence, selected sites, females, Figure 6.3 1975–2010 100 Lung Colorectal Pancreas 10 Kidney Bladder Cervical Oropharynx Stomach Liver AML Incidence rate per 100,000 (log scale) Esophagus Larynx 1 1990 2005 1975 1980 2010 1985 1995 2000 Year of diagnosis Howlader et al. 2013. Source: The data are for nine SEER areas (San Francisco, Connecticut, Detroit, Hawaii, Iowa, New Mexico, Seattle, Utah, and Atlanta). Note: AML Rates are per 100,000 and are age-adjusted to the 2000 U.S. standard population (19 age groups − Census P25-1130). = acute myeloid leukemia. tionship between smoking and the outcome of cancer, a 2006). Figure 1.1A shows those malignancies for which - the Surgeon General’s reports classified the relationship topic not previously addressed in the reports of the Sur with smoking as causal. The chapter begins with an over - geon General on smoking and health. Previous reviews view of the mechanisms by which smoking causes cancer, related to cancer were included in the 2004 Surgeon based on the indepth coverage of this topic in the 2010 General’s report on active smoking (U.S. Department of Surgeon General’s report How Tobacco Smoke Causes Health and Human Services [USDHHS] 2004) and in the (USDHHS 2010). 2006 report on exposure to secondhand smoke (USDHHS Disease Cancer 147

177 Surgeon General’s Report Mechanisms of Cancer Induction by Tobacco Smoke Classic studies demonstrating the covalent binding it is reproduced here as Figure 6.4 (Hecht 1999, 2012a). This section will present a brief overview of the relevant of carcinogens, or their reactive electrophilic metabolites, steps in Figure 6.4 and a more detailed discussion of some - to cellular macromolecules (including DNA) were pub lished at about the same time as the 1964 Surgeon Gener - recent findings pertinent to this overall mechanism. People begin to smoke cigarettes at a relatively al’s report on smoking and health (USDHEW 1964; Miller young age, typically have difficulty stopping, and may and Miller 1976). Building on these seminal observations, continue to smoke for decades. Nicotine is addictive, many researchers explored this mechanistic concept in detail and confirmed it for different classes of chemical but is not a direct chemical carcinogen (see Chapter 5, carcinogens; that line of research continues even today “Nicotine”) (Maier et al. 2011; Murphy et al. 2011). How - ever, by creating and sustaining addiction, it leads to the (Searle 1984; Loebe and Harris 2008; Penning 2011). Tobacco smoke, with its multiple carcinogens, recapitu prolonged exposure to tobacco smoke that increases can - - cer risk for smokers. When smokers inhale smoke, each lates the classic mechanisms established in these studies. The general concept of exposure to carcinogens, metabo - cigarette puff delivers a mixture of carcinogens and toxi - obacco smoke contains more than 7,000 chemi- cants. T lism to reactive intermediates, and DNA damage leading cals, and at least 69 of these can cause cancer (USDHHS to mutations in critical genes has been established as one 2010). These include polycyclic aromatic hydrocarbons major mechanism by which tobacco smoke causes cancer. (PAHs); tobacco-specific nitrosamines; aromatic amines; This topic was discussed in some detail in Chapter 5 of the and volatile carcinogens such as formaldehyde, acet- 2010 Surgeon General’s report. A mechanistic framework encompassing these steps and related phenomena was aldehyde, 1,3-butadiene, and benzene (as well as vari- presented in that report and in related publications, and ous metals). Pathway for causation of cancer by carcinogens in tobacco smoke Figure 6.4 Modified from U.S. Department of Health and Human Services 2010. Source: 148 Chapter 6

178 —50 Years of Progress The Health Consequences of Smoking Most constituents of cigarette smoke, including can induce inflammation resulting in enhanced pneu - the carcinogens, are compounds foreign to the human mocyte proliferation, activation of nuclear factor-kappa body and, consequently, are acted upon by metaboliz- k B (NF- B), and tumor promotion (Takahashi et al. 2010). ing enzymes designed to detoxify them. These enzymes, Cigarette smoke also has cocarcinogens which, while not -transferases, S including cytochrome P-450, glutathione carcinogenic themselves, enhance the smoke’s carcino - and UDP-glucuronosyl transferases and sulfotranferases, genic effects. Further, cigarette smoke induces oxidative catalyze the conversion of these foreign compounds to damage and gene promoter methylation, processes that more water-soluble products that can be easily excreted also likely contribute to cancer development. from the body. But during this process, certain reactive In the last few years, there have been some develop - compounds may be formed as intermediates. Examples of ments that were not fully covered in the 2010 Surgeon these reactive intermediates include electrophilic carboca - - General’s report, but are pertinent to a fuller understand tions or epoxides that can bind covalently to nucleophilic ing of the mechanisms of carcinogenesis by cigarette sites in DNA, including the nitrogen and oxygen atoms smoke. They are discussed briefly here. These binding products are known of DNA nucleobases. Addiction to nicotine results from its binding to as DNA adducts and are critical in carcinogenesis if they nicotinic acetylcholine receptors (nAChRs). An associa- . Persons with rare are not fixed by DNA repair enzymes tion between common variants in the CHRNA5-CHRNA3- syndromes in which DNA repair is deficient, such as Xero- CHRNB4 nAChRs subunit gene cluster on chromosome derma pigmentosum , are highly prone to cancer devel - 15q25 and the risk of lung cancer was reported in three opment; people with this syndrome develop skin cancer genome-wide association studies (Amos et al. 2008; Hung because of the multiple types of DNA damage that result et al. 2008; Thorgeirsson et al. 2008). These genes are from exposure to sunlight (Weinberg 2007). strongly associated with nicotine dependence (Saccone There is convincing evidence for the presence of et al. 2007), and multiple studies have confirmed and DNA adducts in the lungs and other tissues of smok- amplified these observations (Saccone et al. 2009, 2010; ers in amounts generally higher than those found in Timofeeva et al. 2011; Wang et al. 2011; Ware et al. 2011; - nonsmokers. While many of these adducts remain uniden Wassenaar et al. 2011). These results are likely due to tified, a number of studies have characterized specific car - changes in smoking behavior causing an increased uptake cinogen-DNA adducts in the tissues of smokers (Phillips - of nicotine as well as a greater presence of lung carcino and Venitt 2012). gens, such as 4-(methylnitrosamino)-1-(3-pyridyl)-1-bu- If the DNA adducts produced by tobacco smoke tanone (NNK), in carriers of the gene variants described carcinogens and their metabolites evade repair systems above (Le Marchand et al. 2008). The increased uptake - and remain, they can cause miscoding during DNA rep of nicotine, which was confirmed by measurement of its lication when bypass DNA polymerase enzymes direct the metabolite cotinine in a similar study based on the Euro - placement of an incorrect nucleobase opposite the adduct pean Prospective Investigation into Cancer and Nutrition (USDHHS 2010). This can result in a permanent mutation - (EPIC) cohort, is a surrogate for the uptake of carcino - in the DNA sequence. If this mutation occurs in an impor gens and toxicants in cigarette smoke (Timofeeva et al. or in a KRAS, tant section of a cellular oncogene such as 2011; Yuan et al. 2011a, 2012) . Thus, carriers of the gene , the result can be TP53 tumor suppressor gene such as variants smoke their cigarettes more intensely and are an alteration of the normal growth control mechanisms, exposed to higher levels of NNK and other carcinogens in leading to uncontrolled proliferation, further mutations, smoke, thereby increasing their risk of lung cancer. and cancer. Multiple studies, using state-of-the-art meth- Modern DNA-sequencing methods allow scien - ods, have shown that thousands of mutations are present tists to carry out detailed investigations of mutations in in the DNA of lung tumors from smokers, including in human cancers. Because there are multiple carcinogens KRAS critical growth regulatory genes, most frequently in cigarette smoke and multiple DNA adducts in the lungs and TP53 . These genes are discussed in more detail below of smokers, one would expect to find many mutations (Greenman et al. 2007; Ding et al. 2008a; Lee et al. 2010c; within critical genes in the lung tumors from smokers. Pleasance et al. 2010). Sequencing studies are consistent with this expectation. Some constituents of tobacco smoke or their metab - For example, when Greenman and colleagues (2007) olites may bind directly to cellular receptors, leading to investigated mutations in the coding exons of more than activation of protein kinases, growth receptors, and other 500 protein kinase genes, they found that lung cancers pathways, which can contribute to carcinogenesis (Chen were among those with the most somatic mutations (4.21 et al. 2011b). Cigarette smoke contains substances that ). The authors attributed this finding to per megabase 149 Cancer

179 Surgeon General’s Report TP53 , new frequent muta- HRAS recurrent exposure to exogenous mutagens (Greenman et , and PIK3CA , CDKN2A , . In all, 11 of the al. 2007). Another investigation sequenced 188 primary tions were found in FBXW7 and NOTCH1 NOTCH1 were predicted 28 mutations (39%) identified in lung adenocarcinomas; altogether, 247 megabases of to truncate the gene product, suggesting that - NOTCH1 tumor DNA sequence were analyzed and 1,013 nonsynon may function as a tumor suppressor gene rather than as ymous somatic mutations in 163 of the 188 tumors were identified, including 915 point mutations, 12 dinucleotide an oncogene in this tumor type. Moreover, a similar study of 78 additional tumors reported that 30% of the cases mutations, 29 insertions, and 57 deletions (Ding et al. 2008a). Twenty-six significantly mutated genes were iden- harbored mutations in genes that regulate squamous ), TP63 , and tified, including oncogenes and tumor suppressor genes , NOTCH1 differentiation (including IRF6 implicating such dysregulation as a major driver of car - commonly found to be mutated in lung cancer, such as , and others. Mutations STK11 cinogenesis in head and neck squamous cell carcinoma KRAS TP53 , , CDKN2A , (Stransky et al. 2011). were most common in TP53 and KRAS. The results of these studies are consistent with those More recently, a report on complete exome and reported in the 2010 Surgeon General’s report and with genome sequences of 183 lung adenocarcinomas revealed a mean exonic somatic mutation rate of 12.0 events per information found in the COSMIC (Catalogue of Somatic Mutations in Cancer) database (Wellcome Trust anger megabase (Imielinski et al. 2012). Analysis of nucleotide - context-specific mutation signatures grouped the sample Institute 2012), which stores and displays somatic muta set into distinct clusters that correlated with smoking tions in genes associated with cancer, such as TP53 and history and alterations of reported lung adenocarcinoma Collectively KRAS. , the available results of late-generation genes. Elsewhere, Pleasance and colleagues (2010) sequencing studies, as well as the extensive databases on mutations, are completely consistent with KRAS and TP53 - sequenced a small-cell lung cancer cell line; these inves the induction of multiple mutations in critical growth tigators identified 22,190 somatic substitutions, including 134 in coding exons. They found that G → T transver - control genes by metabolically activated carcinogens of smoke, although other processes downstream G sions were the most common (34%), followed by cigarette → A from exposure to carcinogens could also contribute. transitions (21%) and A → transitions (19%). These G Epigenetic changes, defined as nonsequence DNA results are similar to data that have been obtained by - changes, are also an integral part of cancer progres analysis of the TP53 gene, which is discussed later in sion. Gene promoter hypermethylation is an epigenetic this overview. Elsewhere, a case report focused on a non- - change, involving extensive methylation at the 5-posi small-cell lung cancer (NSCLC) from a 51-year-old patient tion of C in CpG islands within the promoter region, and, who had smoked 25 cigarettes per day for 15 years prior often, extending into exon 1 of regulatory genes (Jones - to excision of the tumor, which yielded a poorly differen and Baylin 2002). In lung cancer, more than 750 genes tiated sample with 95% tumor content, most likely an are inactivated by gene promoter hypermethylation, and adenocarcinoma (Lee et al. 2010c). In this patient, single new genes are still being identified through genomewide C nucleotide variants were common, mostly at G → base screening approaches (Selemat 2012). The end result of - G pairs, frequently → T transversions; these were statisti this process can be the loss of gene transcription and, cally distinct from germline mutations. More than 50,000 therefore, the silencing of gene function. Comparison single nucleotide variants were observed, approximately - of DNA methylation profiles between lung adenocarci 17.7 mutations per megabase. At least eight genes in the nomas of current and never smokers, using a genome- EGFR-RAS-RAF-MEK-ERK pathway were either mutated wide platform, showed only modest differences between or amplified. the groups, and it identified only as significantly LGALS4 In another investigation, whole-exome sequenc - hypermethylated and downregulated in smokers (Selamat ing and gene copy number analyses were used to study - et al. 2012). Analysis of the DNA methylation data identi 32 primary head and neck squamous cell carcinomas fied two tumor subgroups, one of which showed increased (Agrawal et al. 2011). Tumors from patients with a his - DNA methylation and was significantly associated with tory of tobacco use had more mutations than did tumors mutation and, to a lesser extent, with smoking. KRAS from patients who did not use tobacco, and tumors that Promoter methylation of several genes, including P16, were negative for human papilloma virus (HPV) had more occurs early in tumor formation. One study of head and mutations than did HPV-positive tumors. Six of the genes P16 neck cancer found that methylation was significantly that were mutated in multiple tumors were assessed in 1 and positively associated with pack-years of smoking and - up to 88 additional head and neck squamous cell carci was an independent risk factor for overall survival, being nomas. In addition to previously described mutations in 1 Pack-years = the number of years of smoking multiplied by the number of packs of cigarettes smoked per day. 150 Chapter 6

180 The Health Consequences of Smoking —50 Years of Progress This carcinogen induces a high incidence of esophageal significantly associated with shorter survival in patients with early resectable adenocarcinomas (Ai et al. 2003). tumors in rats (Yuan et al. 2009, 2011a,b). promoter hypermethylation also cor Smokers experience proinflammatory changes in - In that study, P16 their lungs. Inflammation is intimately associated with related significantly with a history of alcohol consumption k B and tumor promotion (Malkinson or tobacco use in head and neck cancer. Other genes, such activation of NF- as BRMS1 2005; Smith et al. 2006; Lee et al. 2008), and many studies and RASSF1A, may be more frequently methyl- in laboratory animals demonstrate that anti-inflammatory ated in various tumor types from smokers. In a study by agents can decrease tobacco carcinogen-induced lung Tessema and colleagues (2009), the f requency of methyla- TNFRSF10C tion of tumorigenesis (Hecht et al. 2009). In addition, chronic BHLHB5 was significantly BOLL , , and obstructive pulmonary disease, particularly emphysema, higher in adenocarcinomas from never smokers than in MGMT is an independent risk factor for lung cancer in smok- those from smokers. Methylation of genes, such as and → ers. This association further implicates a strong role for AGT promoter hypermethylation, may increase G A transition mutations at CpG sites within the TP53 gene inflammation in lung cancer (Turner et al. 2007). In one study, the tumor-promoting activity of cigarette smoke in NSCLC. was examined in mouse models of lung tumorigenesis - These data in aggregate support the pathways illus (Takahashi et al. 2010); here, exposure to smoke after trated in Figure 6.4. The contribution of specific tobacco treatment of A/J mice with NNK increased the multiplicity - smoke carcinogens to lung cancer (and also to esopha LA2 of lung tumors. Similar results were obtained in geal cancer) has been investigated in several nested KRAS mice harboring a mutation in KRAS codon 12 identical to case-control studies as well. In these studies, the car - that caused by NNK. I ) was required for (IKK β B kinase k β cinogens or their metabolites were quantified in stored NF- - B activation and played a critical role in tumor pro k urine samples that were collected from smokers years motion in this system, most likely through the induction For example, using or decades before cancer developed. of inflammation and related phenomena (Takahashi et al. frozen urine samples collected during the 1980s from 2010). These studies amplify and extend earlier observa - more than 18,000 smokers in Shanghai, China, scientists tions demonstrating the tumor-promoting activity of ciga- have found that specific metabolite levels were associated rette smoke. with an increased risk of lung or esophageal cancer, even after correction for the number of years of smoking and (Yuan et al. 2009, number of cigarettes smoked per day 2011a,b). Thus, significantly elevated risks for lung can- Summary cer were associated with increased levels of the NNK metabolites’ total NNAL [4-(methylnitrosamino)-1-(3- Understanding of the mechanisms by which smok - pyridyl)-1-butanol and its glucuronides] and the PAH - ing causes cancer continues to advance. An overall frame metabolite phenanthrene tetraol. The strongest ele - work for the causation of cancer by tobacco smoking was vated risk was for esophageal cancer in individuals with set out in the 2010 Surgeon General’s report. The utility of the highest levels of the tobacco-specific carcinogen that framework is supported by new experimental findings N -nitrosonornicotine and its glucuronides in their urine. ′ as well as by ongoing studies of smokers in the population. Changing Cigarettes and Risk for Lung Cancer Over Time type (e.g., filtered or unfiltered) and across brands of the Cigarette smoking is the predominant cause of lung cancer in the United States, and lung cancer is same type (IARC 2004; Burns et al. 2008; World Health the country’s leading cause of cancer death (USDHHS Organization [WHO] 2008b). Over past decades, multiple 2004). Cigarette smoke, which contains multiple car - substantive changes in the design and composition of cigarettes have altered the chemistry of tobacco smoke cinogens (Hoffmann and Hoffmann 1997; IARC 2004; raising the question as to whether lung cancer risks have USDHHS 2004; Rodgman and Perfetti 2009), is composed changed in response (Hoffmann and Hoffmann 1997; Rod of gases and particles with a distribution of size that result - in substantial deposition in the lung when the smoke is gman and Perfetti 2009). This section reviews evidence inhaled (Stratton et al. 2001; Gower and Hammond 2007). relevant to this question. The composition of tobacco smoke varies with cigarette 151 Cancer

181 Surgeon General’s Report - as changing nicotine yields and the marketing of vari This section focuses on lung cancer because it is ous types of cigarettes. This section does not explore the the cancer most related to cigarette smoking (USDHHS implications of these changes for diseases other than 2004). Substantial data are available, both over time and from many countries, on the occurrence of lung cancer, lung cancer. both generally and by histologic type. The topic of lung cancer in relation to smoking has been addressed in depth in several past reports of the Surgeon General. These Changes in Cigarettes Over the reports have focused on levels of machine-measured Past Several Decades tar and nicotine in relation to risk and have considered whether changes in design and characteristics that have Since the 1950s, cigarettes have undergone changes lowered the tar yield of cigarettes have also reduced in their design and composition (Hoffmann and Hoffmann the risk of diseases caused by smoking (USDHHS 1981, 1997; NCI 2001). The most prominent changes have been 2004). The 2004 Surgeon General’s report on the health the addition of filters and the use of ventilation holes in consequences of smoking concluded that no substantive the filters to lower machine-measured tar and nicotine reduction in the risk of disease was associated with using yields. Figure 6.5 shows the rapid rise in the use of fil - cigarettes with low levels of tar, as measured by machine. tered cigarettes that followed the heavy marketing of such This and earlier reports clearly document that machine- The marketing effort prom - cigarettes in the mid-1950s. measured tar yields have little relationship to the doses ised a lower risk product to smokers who had become actually received by smokers because of the phenomenon concerned about the disease risks of smoking (Brandt of compensation. This section focuses mainly on whether 2007). This shift to filters continued and today almost the changes in the design and composition of cigarettes all manufactured cigarettes currently consumed in the over time that paralleled the reduction in tar yields (by United States are filtered (Hoffmann and Hoffmann 1997; machine measurement) may have altered—and possibly NCI 2001). Figure 6.6 shows the move to cigarettes with even increased—the risk of lung cancer associated with lower tar yields, beginning with a shift from brands with cigarette smoking. The analysis is limited to cigarette more than 20 milligrams (mg) of machine-measured tar design issues and does not consider other issues, such Market share and total annual cigarette sales of filtered and unfiltered cigarettes in the United States, Figure 6.5 1925–1993 National Cancer Institute 2001; data from Maxwell 1994. Source: 152 Chapter 6

182 The Health Consequences of Smoking —50 Years of Progress Figure 6.6 Market share of total cigarettes sold per year, by tar yield (milligrams [mgs] of tar by Federal Trade Commission method), United States, 1967–1990 National Cancer Institute 2001; data from Maxwell 1994. Source: Note: Tar levels for given years are derived from Federal Trade Commission reports (for years 1967–1990). Sales data by brand are from Maxwell (1994). Brand-specific market shares are summed by tar level of the brand in the given year to generate the market share for cigarettes with given tar yields. researchers in the past did not have access to information to lower tar-yielding brands in the late 1960s and early 1970s (NCI 2001). By 1990, about two-thirds of cigarettes about the nature and extent of these and other changes in sold had either medium (11−15 mg), low (6−10 mg), or cigarettes because they were handled as trade secrets and, therefore, not disclosed by the industry. very low (1−5 mg) yields of tar. The principal mechanism underlying the lower yields of machine-measured tar was the increase in the number and the size of ventilation Changes in Design, Curing, and Composition holes in the filter, thereby diluting the smoke entering Although smokers may perceive cigarettes as very the machine (Hoffmann and Hoffmann 1997; NCI 2001). simple devices: chopped-up tobacco rolled in paper, per - Although these changes reduced tar delivery as measured - haps with a filter attached to the end, the reality, how by the U.S. Federal Trade Commission’s (FTC’s) protocol, ever, is that cigarettes are highly engineered products which did not reflect how smokers actually smoke; they (Hoffmann and Hoffmann 1997; Rodgman and Perfetti did not reduce the risks of disease and premature mortal - 2009; Proctor 2011). The design features of cigarettes can ity in smokers (NCI 2001; USDHHS 2004). have significant effects on the composition of the tobacco Epidemiologic evaluations of risk and a ssessments of smoke and perhaps its toxicity. Over time, changes to ciga - smoke chemistry during the decades in which these sub- rettes have become progressively more extensive and more stantial changes occurred tended to treat all cigarettes as complex, further complicating the efforts of researchers to if they were equivalent, both over time and across brands. understand their health implications (Hoffmann and Hoff- The exception was that these evaluations did consider the mann 1997; NCI 2001; O’Connor et al. 2008; O’Connor machine-measured tar and nicotine yields and whether - and Hurley 2008; WHO 2008b). Many factors can influ they were filtered. However, the design and composition ence the chemistry of tobacco smoke: (1) the geographic of cigarettes changed substantially in other ways, even as location where the tobacco is grown (which can alter the they were continuously redesigned to deliver ever lower heavy metal content of smoke, for example) (IARC 2004, machine-measured yields of substances. Unfortunately, 153 Cancer

183 Surgeon General’s Report review of the extensive literature describing the effect of 2007); (2) agricultural practices (which can influence isolated changes on smoke composition is beyond the levels of nitrates and pesticides, but also polonium and scope of this section. heavy metal content as well) (Hoffmann and Hoffmann Beyond the data held by the manufacturers, the 1997; Rodgman and Perfetti 2009); (3) how the tobacco details on differences in the design and composition of is cured and processed (which can influence tobacco-spe- cigarettes across U.S. brands are not available in a sys- cific nitrosamine levels and other factors) (Hoffmann and tematic form. Complete and representative information is Hoffmann 1997; NCI 2001; Peele et al. 2001; O’Connor also not available over time on the composition of smoke et al. 2008; O’Connor and Hurley 2008); (4) the blend of - generated by individual brands or on the changes in man tobacco used; (5) the use of reconstituted tobacco sheet ufacturing practices for different brands. Longitudinal - and puffed tobacco (tobacco expanded through an indus data on brands marketed in the United States are limited trial process) (Hoffmann and Hoffmann 1997; IARC 2004, to data—using FTC’s protocol—on machine-measured 2007; O’Connor et al. 2008; O’Connor and Hurley 2008; yields of the tar, nicotine, and carbon monoxide produced. Rodgman and Perfetti 2009); (6) the engineering char - Without this information, the research and public health acteristics of the manufacturing process (Hoffmann and communities have been unable to fully assess the poten- Hoffmann 1997; O’Connor et al. 2008; O’Connor and Hur - tial effects of changes in the design and composition of ley 2008; Rodgman and Perfetti 2009); (7) the additives cigarettes on smokers’ exposures over time to toxicants in used in tobacco; and (8) the pattern of puffing the smoker cigarette smoke. Nevertheless, the limited data that exist uses to generate the smoke (which can alter the quantity allow for some assessment of likely changes in smoke tox - of smoke generated and the relative composition of its icity following changes that have been made in cigarettes. constituents) (WHO 2007, 2008b; Burns et al. 2008). Cigarettes in Australia, Canada, and the United King- Differences Across Brands in Toxicant Yields dom are made primarily of flue-cured tobacco, but most brands sold in the United States use a blend of air-cured Of the 7,000 or more constituents in tobacco and tobaccos (Hoffmann and Hoffmann 1997; IARC 2004; tobacco smoke, 69 have been identified as carcinogens WHO 2008b; Rodgman and Perfetti 2009). In addition, - (USDHHS 2010). The complexity and expense of mea substantial amounts of reconstituted tobacco sheet and suring multiple constituents for all the different brands puffed tobacco are added to the blend. The soil in which under multiple sets of machine parameters have led the tobacco is grown, the agricultural practices used, and tobacco industry scientists to suggest that constituent - the methods of curing and processing the tobacco also dif yields can be benchmarked and reliably predicted from fer across brands and have changed over time (Hoffmann machine-measured tar yields (Counts et al. 2004, 2005, and Hoffmann 1997; Peele et al. 2001; IARC 2004; Rodg - 2006; Morton and Laffoon 2008). This concept is based man and Perfetti 2009). Flavoring agents; processing aids, on the assumed relationship between the total mass of such as humectants; chemicals intended to alter the pH of - smoke and its nicotine content, as measured by a smok the smoke; and other agents are added to tobacco as part ing machine. However, the mass of smoke generated by of the manufacturing process. a smoking machine using any fixed protocol bears little - Approaches used to alter the processes of generat relationship to the amount of smoke inhaled by a smoker ing smoke may involve the cut size of the reconstituted or to the differences between brands in smoke exposure tobacco sheet, filter ventilation, the density of the tobacco (Jarvis et al. 2001; NCI 2001). A more appropriate method in the rod, the composition and design of the filter mate- for examining the variation in constituent yields across rial, the porosity of the cigarette paper, and other factors brands is to examine these yields after they have been nor - (Hoffmann and Hoffmann 1997; O’Connor et al. 2008; malized per mg of tar or per mg of nicotine to characterize O’Connor and Hurley 2008; Rodgman and Perfetti 2009). the variation that might be experienced for a given level The pattern the smoker uses to puff the cigarette is super - of nicotine intake. imposed on all of its intrinsic characteristics. This pat - Nicotine is the principal addictive constituent tern varies among smokers and can change with different sought by the smoker and the ratio of tar to nicotine is types of cigarettes smoked by the same smoker; it can also relatively constant across brands. When the Massachu - change systematically across smokers in response to cer - setts Benchmark Study data on yields for a 1999 sample tain design features, most notably filters and ventilation of U.S. brands of cigarettes are normalized per mg of tar (NCI 2001; WHO 2008b). Rodgman and Perfetti (2009), or per mg of nicotine, the ability of tar yields to predict - O’Connor and colleagues (2008), and O’Connor and Hur the variation in yields of other constituents is poor (Har - - ley (2008) have reviewed the impact of many of these fac ris 2001, 2004). In fact, the normalized yields of several tors on the composition of tobacco smoke, but a detailed 154 Chapter 6

184 —50 Years of Progress The Health Consequences of Smoking constituents are higher for cigarettes with low machine- paper, and using other techniques (O’Connor et al. 2008; measured tar yields than for those whose machine-mea - O’Connor and Hurley 2008; WHO 2008b; Rodgman and sured tar yields are high (Harris 2004). Perfetti 2009). Data are not available for all U.S. brands Table 6.2 presents the variability in the yields of a over time, but Hoffmann and Hoffmann (1997) published variety of constituents across brands, normalized per data for a prominent cigarette brand, measured repeatedly mg of tar or per mg of nicotine, from the Massachusetts ]P a from 1959–1995, that showed a modest decline in B[ Benchmark Study sample of U.S. cigarettes in 1999. In levels in smoke over that period. this table, the coefficient of variation across brands (which ]P, levels of a In contrast to the decline in levels of B[ represents the standard deviation of the measurements tobacco-specific nitrosamines, specifically NNK, increased across brands normalized to the mean value of that con - dramatically in the previously referenced brand from stituent for all brands) is divided by the mean standard 1978–1995 (Hoffmann and Hoffmann 1997). This increase deviation of replicate measurements for that constituent. was due in part to the increased nitrate levels in the This formulation expresses the variation of constituents - tobacco used in cigarettes even before the curing (Hoff across brands in relation to the precision with which the mann and Hoffmann 1997; Ding et al. 2008b; O’Connor et constituent can be measured. Table 6.2 demonstrates that - al. 2008; O’Connor and Hurley 2008; Rodgman and Per - for many of the toxicants measured, the variation in con fetti 2009) and to changes in curing practices that have stituents across brands, normalized per mg of tar or per increased the presence of oxides of nitrogen and nitrate mg of nicotine, is many times higher than can be explained ion and the latter’s reaction products during curing, with by the variability of the measurement. Clearly, at least in the resultant formation of tobacco-specific nitrosamines terms of constituent yields from machine-generated ciga - from the nicotine in the leaf (Hoffmann and Hoffmann rette smoke, smoke from all cigarettes is not uniform in 1997; NCI 2001; Peele et al. 2001; IARC 2004; Ding et al. composition. This variability is likely not limited to 1999, 2008b; O’Connor et al. 2008; O’Connor and Hurley 2008). when the cigarettes were sampled, or to have remained constant over time. Furthermore, normalized constituent Differences in Toxicant Yields Across Countries yields in Canadian and Australian cigarette brands and a Relatively more evidence is available for differences sample of international blended cigarette brands manu - in toxicant yields from comparisons of international factured by Philip Morris International have demonstrated brands of cigarettes. Of particular note, the use of burley - similar variability (WHO 2008b). In addition, when bio - tobacco in U.S.-style blended cigarettes contributes sub markers of exposure to specific toxicants are assessed, the stantially to the differences in tobacco-specific nitrosa - data show considerable variability in their levels among mines between U.S.-style cigarettes and those of Canada smokers, particularly in heavy smokers (Joseph et al. and Australia (Burns et al. 2008; Ding et al. 2008b; WHO 2005); this finding is consistent with variation in exposure 2008b), where most brands contain mainly unblended, due to differences in smoke composition across brands flue-cured tobacco. Datasets are available for some smoke and to inherent variability among smokers. constituents that have been measured for major brands in the Canadian and Australian markets (WHO 2008b) Changes in Tobacco-Specific Nitrosamine and and for a selection of international brands of blended a Benzo[ ]pyrene Levels Over Time cigarettes manufactured by Philip Morris (Counts et al. Because only limited longitudinal data are available 2004, 2005). for toxicant yields, changes in these yields over time are Several other differences between Canadian and difficult to characterize accurately for all brands. How- Australian brands were found, although cigarettes in both ever, for one major U.S. brand, some data are available for countries are made with unblended, flue-cured tobacco. two of the major toxicants: benzo[ ]P) and a ]pyrene (B[ a Differences in the levels of cadmium and lead between the N ́ -nitrosonornicotine the tobacco-specific nitrosamines ( brands are notable. Figure 6.7 presents the mean yields of [NNN] and NNK). some toxic constituents for the major Canadian and Aus - a B[ ]P, one of the earliest identified carcinogens tralian brands sampled in late 2000 to early 2001. The yields in cigarette smoke, is a typical carcinogenic PAH and is are normalized per mg of nicotine and expressed as a ratio often used as a surrogate index for the PAHs as a group. to the mean yields for an international sample of brands Efforts to reduce the levels of this carcinogen in smoke manufactured by Philip Morris. The data for the Canadian have included increasing the proportion of tobacco in brands are presented for all brands and for brands other the cigarette rod that is made up of reconstituted sheet, than those with high NNN levels (U.S.-style and Gauloise changing the tobacco blend, increasing the porosity of the cigarettes). The expected differences between flue-cured 155 Cancer

185 Surgeon General’s Report Table 6.2 Ratio of brand coefficient of variation to replicate measurement coefficient of nicotine and tar variation per milligram (mg), per Massachusetts Machine Smoking Protocol, in rank order Per mg nicotine Constituent Constituent Per mg tar 8.85 NNN NNN 6.84 8.18 NAT 6.18 NAT 7.45 NAB NAB 5.25 Mercury 6.28 Isoprene 5.00 Isoprene 4.79 Mercury 6.07 Benzene Benzene 4.10 4.86 Acetone 4.36 Toluene 3.72 3.64 4.33 Acetone Toluene 3.63 Propionaldehyde 4.30 HCN 3.59 HCN 4.21 Nitric oxide 3.59 Methyl ethyl ketone 1,3-Butadiene 4.19 Acetaldehyde Propionaldehyde 3.47 4.12 1,3-Butadiene 4.11 Acetaldehyde 3.43 3.35 4.11 NNK Acrolein 3.34 Nitric oxide 3.97 Methyl ethyl ketone 3.30 Phenol 3.78 Acrylonitrile 3.18 3.76 3-Aminobiphenyl m + p-Cresol 3.12 NNK 3.49 Acrolein 2.91 Acrylonitrile 3.40 4-Aminobiphenyl 2.86 B[ a ]P 3.35 m + p-Cresol 2.79 2-Aminonaphthalene 3.23 Ammonia 3-Aminobiphenyl 2.45 Phenol 3.18 Hydroquinone 3.14 2.45 1-Aminonaphthalene 2.77 4-Aminobiphenyl Styrene 2.32 2.59 Hydroquinone 2.27 2-Aminonaphthalene 2.09 Ammonia Styrene 2.24 Crotonaldehyde 2.03 2.03 Cadmium 1.93 1.80 Butyraldehyde 1-Aminonaphthalene Formaldehyde Crotonaldehyde 1.93 1.78 Pyridine 1.75 1.90 Catechol 1.67 Butyraldehyde 1.73 Formaldehyde a ]P Cadmium 1.66 B[ 1.46 1.44 Catechol Pyridine 1.62 Lead Lead 1.42 1.61 Arsenic 1.46 Quinoline 1.29 Quinoline 1.45 Arsenic 1.28 Source: Unpublished data from the 1999 Massachusetts Benchmark Study as provided by Greg Connolly, Massachusetts Department of Health. nitrosoanatabine; N- B[ a ]P = benzo[ a ]pyrene; HCN = hydrogen cyanide; NAB = N ′ -nitrosoanabasine; NAT = Note: nitrosonornicotine. ′- N NNK = 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNN = Chapter 6 156

186 —50 Years of Progress The Health Consequences of Smoking Figure 6.7 Ratio of the means of constituents in cigarette brands from Canada and Australia to the mean for an international sample of U.S.-style blended cigarettes manufactured by Philip Morris and sampled from late 2000 to early 2001 Counts et al. 2005; World Health Organization 2008. Reprinted with permission from World Health Organization, © 2008. Source: = NNK Note: N = nitrogen oxides. NOx -nitrosonornicotine; ′ NNN = 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; and blended cigarettes are evident (Ding et al. 2008b); These differences in tobacco-specific nitrosamine the flue-cured cigarettes from Australia and Canada levels in smoke translate to different exposures among smokers. Mouth-level exposures to NNN and NNK and have much lower levels of tobacco-specific nitrosamines urinary measures of NNAL—a metabolite of NNK—are (notably NNN and NNK) and substantially higher levels ]P. Australian and Canadian brands, however, differ a higher among smokers in the United States than in smok- of B[ ers in Australia and Canada (Ashley et al. 2010), demon markedly from blended cigarettes in a number of other - - strating that the observed differences in the composition toxicants, with lower levels of oxides of nitrogen; 1-ami nonapthalene; 1,3-butadiene; and isoprene. Canadian, but of smoke result in substantive differences in exposure to not Australian, cigarettes have higher levels of catechol, tobacco-specific nitrosamines. phenol, and cresols. These differences may reflect the use Low-Tar Cigarettes Do Not Reduce Risk of tobacco grown for use in cigarettes in different regions of Lung Cancer of Canada and Australia. Figure 6.8 shows the differences in NNN and NNK Early efforts to alter the risks of cigarettes focused between Australian brands and a blended version of the on reducing the yields of tar and nicotine as measured by Marlboro brand designed for the Australian market (Burns machine-smoking methods. As a result, machine-mea - et al. 2008; WHO 2008b). The levels of NNN and NNK in the sured yields of tar and nicotine declined by more than blended-tobacco cigarette from Marlboro are much higher 60% from the 1960s to 1990 (Hoffmann and Hoffmann than those for even the highest level brand reported to the 1997; NCI 2001). Much of that reduction was accom- Australian regulatory authorities. plished initially by adding filters and later by ventilating 157 Cancer

187 Surgeon General’s Report ́-nitrosonornicotine (NNN) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone N Mean and range of Figure 6.8 (NNK) yields per milligram [mg] of nicotine for brands reported to the Australian government, con- trasted with the levels of NNN and NNK reported for a Philip Morris Marlboro brand cigarette identified as an Australian brand, in 1999 Source: World Health Organization 2008b. Reprinted with permission from World Health Organization, © 2008. = hard pack; HP = nanogram. ng = king size; KS Note: = filter; F as to whether there might have been an increasing risk of - the filter to dilute the smoke coming through it, thus low lung cancer over time from smoking cigarettes that could ering the machine-measured yields of tar and nicotine so have contributed to this delay. the newer products could be marketed as being less risky to health (NCI 2001). But to compensate for the reduced Epidemiologic studies are a key source of evidence yields, smokers changed the way they smoked these ciga - for assessing whether the risk of lung cancer associated rettes, resulting in no meaningful reduction in either the with smoking has changed over time. Particularly infor - mative is the comparison by Thun and Heath (1997) of total dose of smoke received or in the risks of diseases two prospective cohort studies of the risk of smoking caused by smoking (NCI 2001; USDHHS 2004). Changes in patterns included increasing the volume and velocity of conducted by the ACS. Each study, conducted more than 20 years apart, followed more than 1 million men and puffs, increasing the duration of puffing, and shortening the intervals between puffs (NCI 2001). However, the pro - women. The Cancer Prevention Study I (CPS-I) began in tocol for smoking by machines was not changed. 1959, and the Cancer Prevention Study II (CPS-II) began - in 1982. The more than two decades between the stud Overall Death Rates for Lung Cancer Indicate ies saw substantial changes in the design and composition Increased Risk of Smoking in Recent Decades of cigarettes and in the brands of cigarettes that Ameri- cans smoked. The decline in machine-measured yields of In the United States, the prevalence of smoking toxicants in cigarettes between these two studies led to an among males has declined since at least the 1950s, but expectation that the risk of lung cancer death for smokers age-adjusted death rates for lung cancer among men did would likely be lower in the CPS-II. The authors compared not begin to decline until approximately 1990 (Wingo death rates from lung cancer in the first 6 years of follow- et al. 1999). Among women, the comparable death rates - up for each study among the subsamples of never and cur peaked around 2003 and significantly declined (Jemal rent smokers at enrollment. The risks were found to be - et al. 2013), likely due to considerable success in reduc higher in CPS-II (Thun and Heath 1997). Figure 6.9 pres- ing the prevalence of smoking among women. The long ents the results from these analyses for men and women delay between decreases in the prevalence of smoking and current smokers and never smokers based on 786,387 changes in death rates for lung cancer raises the question Chapter 6 158

188 —50 Years of Progress The Health Consequences of Smoking advance in age and calendar year) to estimate the expected CPS-I and 711,363 CPS-II participants. The risk for never occurrence of death from lung cancer in the absence of smokers (as measured by the death rate from lung cancer) any change in the risk imposed by smoking. Using birth- went essentially unchanged during the interval between cohort-specific data on smoking developed by Harris the two studies, but the risk for smokers increased dra- (1983) and a multistage carcinogenesis model similar to matically, with a proportionately greater increase among women smokers. The increase in risk of death from lung that developed by Whittemore (1988), Swartz (1992) pre- dicted overall age-adjusted trends in lung cancer mortality cancer remained after controlling for measured differ - for White men from 1970–1985. The author estimated that ences in duration and intensity (number of cigarettes smoked per day) between the smokers in the two studies. a 12% decline in rates should have occurred during this The 40-year follow-up of the British Doctors’ Study interval, based on the assumption of a constant effect over from 1951–1991 presents similar evidence. During the time. However, this estimated decline contrasted sharply second 20 years of follow-up, the risk of death from lung with the observed 26% increase in lung cancer death rates cancer was greater than during the first 20 years (Doll et during the interval (Swartz 1992). To predict death rates al. 1994); this increase over time was limited to smok - - for lung cancer over time by birth cohort, Tolley and col leagues (1991) used an updated set of birth-cohort-spe - ers and former smokers. Among never smokers, rates cific estimates for smoking prevalence and a risk model of lung cancer mortality were relatively constant across developed by Peto (1986) that was based on data from the calendar years (Thun et al. 2006, 2008), suggesting that British Doctors’ Study (Doll et al. 1994). These authors the changes observed in the relative risk (RR) of smoking were unlikely to have resulted from changes in population estimated that overall lung cancer mortality should have demographics or in other risk factors for lung cancer in the started to decline in the early 1980s for White men and in general population. the mid-1990s for White women. Instead, observed lung Models of risk based on smoking patterns have been cancer mortality continued to rise throughout the 1980s, applied to data on smoking prevalence for birth cohorts peaking in the early 1990s for White men (Wingo et al. (i.e., sets of individuals born during specified calendar 1999) and 2003 for women generally (Jemal et al. 2013). years and for whom rates can be examined as the cohorts A similar approach, using risk models developed from Death rates from all lung cancers, by smoking status, Cancer Prevention Study I (CPS-I) and Cancer Figure 6.9 Prevention Study II (CPS-II), 1959–1965 and 1982–1988 Thun and Heath 1997. Reprinted with permission from Elsevier, © 1997. Source: All data are age adjusted. Data for male and female smokers are also adjusted for duration of smoking and number of cigarettes Note: smoked per day. Each data point represents the mortality from the 6-year interval specified by the study. 159 Cancer

189 Surgeon General’s Report the CPS-I data and birth-cohort-specific data on smoking Some participants in CPS-I may have reported initiating cigarette smoking at the time at which they first used prevalence from the National Health Interview Survey, - demonstrated a systematic trend of increasing underesti tobacco of any type, or they may have otherwise overes - mation of observed death rates for lung cancer across all timated their duration of cigarette smoking, leading to a - birth cohorts with advancing calendar years (NCI 2001). longer reported duration of cigarette smoking than actu ally occurred. The resulting misclassification, with a bias Estimates of smoking behaviors for birth cohorts toward reporting a longer duration of smoking, could lead that incorporate changes in the number of cigarettes smoked per day were developed for NCI’s Cancer Inter - to a reduced magnitude of the estimated effect of duration vention and Surveillance Modeling Network (CISNET) of smoking on risk of lung cancer death in risk models (Anderson et al. 2012). These estimates are more detailed based on CPS I data. Because a much larger fraction of - those who developed lung cancer in CPS-II took up smok than previous data on smoking behaviors for birth cohorts and include estimates of the intensity and duration of ing after 1914, the effect of overreporting the duration of smoking would be lower in CPS-II, the magnitude of the smoking for 5-year birth cohorts from 1900–1984. For each calendar year, these estimates provide rates of smok- estimated duration effect would increase, and the risk of ing initiation; prevalence of current and former smoking; smoking would appear to have increased between the two studies, with adjustment for differences in reported dura - - and distributions of the duration of smoking, the dura tion of smoking. tion of abstinence, and the number of cigarettes smoked per day for current and former smokers. These estimates The study used the CISNET smoking rates and risk for smoking behavior were combined with risk models for models based on CPS-I (Burns et al. 2011b) and attempted current, former, and never smokers derived from 12-year to minimize the contribution of overreporting of smoking duration and early smoking by eliminating birth cohorts follow-up data from the CPS-I (Knoke et al. 2004, 2008) to estimate birth-cohort-specific lung cancer death rates born before 1915—the period during which overreporting from 1960–2000 (Burns et al. 2011b). The resulting esti - was most likely. In addition, the potential for underestima - tion of the increase in duration over time to produce the mates were compared with observed U.S. national lung observed progressive underestimation of the U.S. birth- cancer death rates for the same birth cohorts. The compar - cohort-specific death rates for lung cancer with advancing ison showed a progressively increasing underestimation calendar time was examined by iteratively increasing the of U.S. national lung cancer death rates across all birth duration term and examining the fit of the resulting esti- cohorts as calendar years advanced from the 1960s to 2000 mates to the observed U.S. death rates. Although increas- (Burns et al. 2011b). This underestimation was eliminated when a term that increased the risk of smoking, based on ing the duration term increased the estimated rates as the estimated duration of smoking after 1972, was added anticipated, the pattern of a progressive change in risk to the risk model. These analyses suggest that estimates - remained even as calendar years advanced, with an over estimated actual risk giving way to an underestimated risk of smoking-related lung cancer deaths that are based on observations in the 1960s underestimate the current risks as calendar years advanced. Thus, an increasing effect of - duration on risk of death from lung cancer did not explain of smoking, implying that the risk of death from lung can the progressive underestimation of mortality from lung cer associated with smoking may have increased over the cancer, whereas a term increasing the risk of cigarette past several decades—that is, during the same decades in which changes in the design of cigarettes were made. smoking over time did. Overreporting in CPS-I also may have resulted in an Considering the increase in risk of death from lung overestimation of the number of cigarettes smoked early cancer seen from CPS-I to CPS-II, Thun and Heath (1997) in life, but the contribution of cigarettes smoked per day recognized the possibility that the risk of death from lung cancer observed in CPS-I might underestimate the contri to risk of lung cancer is much smaller than the contribu- - tion of duration (Flanders et al. 2003; Knoke et al. 2004), - butions of (a) amount smoked and (b) duration of smok ing due to overreporting in the CPS-I data of the duration and the exponent for the cigarettes-per-day term in the CPS-I risk equations is close to one (Knoke et al. 2004, of smoking and the number of cigarettes smoked early in 2008). As a result, any underestimation of lifetime num - life. Among White men, the transition from other forms ber of cigarettes smoked per day due to overreporting of of tobacco use (e.g., cigars and pipes) to cigarettes began smoking early in life is expected to be modest and could be - largely after 1914, because cigarette smoking was uncom mon before that year (Burns et al. 1997). Because lung approximated by a constant that would be incorporated in - cancer is a disease of older ages, much of the lung can the risk equations when they are adjusted for the healthy cer mortality experience in CPS-I occurred among men population selection bias (Pinsky et al. 2007) required for who were well past their adolescence by 1914, and yet such estimates (Tolley et al. 1991; Burns et al. 2011b). many of them reported initiating smoking at early ages. 160 Chapter 6

190 —50 Years of Progress The Health Consequences of Smoking be real and not a consequence of changing diagnos- - To further assess changes in the risk of lung can cer from smoking over time, Thun and colleagues (2013) tic practices. Notably, the increase in adenocarcinoma of the lung extended their analyses by comparing the lung cancer risk has been accompanied by an increase in the estimated associated with smoking observed in five contemporary cohorts (2000–2010) with risks observed in CPS-I (1959– RR for this type of lung cancer associated with cigarette 1965) and CPS-II (1982–1988). For never smokers, rates smoking. Early in the investigation of the lung cancer of death from lung cancer remained constant across time epidemic, the most common histologic type of lung can - cer in men was squamous cell carcinoma, and the RR of among men and increased only slightly among women. squamous cell carcinoma associated with smoking was Among females 55 years of age and older at baseline, the substantially higher than that for adenocarcinoma (Wu- RR for lung cancer comparing current smokers to never smokers progressively increased from 2.73 in CPS-I to Williams and Samet 1994; USDHHS 2004). Kreyberg 12.65 in CPS-II to 25.66 for the 2000–2010 cohorts. Corre - (1962) even debated whether adenocarcinoma was asso - ciated with cigarette smoking, because of the low RR sponding RRs for current male smokers were 12.22, 23.81, and 24.97, respectively. Compared with their counterparts and because adenocarcinoma is the most common type - in CPS-I and CPS-II, both men and women in the contem of lung cancer among women who have never smoked. porary cohorts were at greater risk for lung cancer despite As the incidence of lung adenocarcinoma increased over smoking fewer cigarettes per day. Duration of smoking time, the RRs of this type of lung cancer associated with smoking also increased (USDHHS 2001), suggesting that increased substantially across the study time periods for a new, or at least a substantially enhanced, risk of develop - women. In comparison, duration of smoking changed only ing adenocarcinoma of the lung occurred in smokers. In modestly for men across the studies and actually declined slightly between CPS-II and the 2000–2010 cohorts. a comparison of data from CPS-I and CPS-II, Thun and Thun and colleagues (2013) also stratified their colleagues (1997) found that the RR for adenocarcinoma increased in smokers from 4.6 for men and 1.5 for women analyses by smoking intensity (i.e., number of cigarettes (per data from CPS-I, conducted 1959–1965) to 19.0 for smoked per day) and duration of smoking for all three men and 8.1 for women (per data from CPS-II, conducted study periods. Within each stratum of smoking intensity and duration of smoking, the RR estimates increased over 1982–1988), but that the age-adjusted death rates for adenocarcinoma of the lung among never smokers were time for women. For men, RR estimates increased over essentially unchanged over the period. Furthermore, risk time within each stratum of smoking intensity, but a con- - for lung cancer of all tissue types among never smokers sistent pattern was not evident for each stratum of smok remained constant over the same interval (Thun et al. ing duration. The authors concluded that the risk of lung cancer from smoking has continued to increase among 2006, 2008). women but among men has plateaued at the very high Trends across calendar years in age-standardized incidence rates of lung cancer have also varied by tumor levels observed in the 1980s. type. Figure 6.10 presents trends in age-standardized inci- Trends in most other tobacco-related cancers have not been examined in detail, although Baris and col - dence rates in the United States from 1973–2010 for lung cancer by gender and histologic type using data from NCI’s leagues (2009) reported an increase in the incidence of SEER Program. Among men, the decline in the incidence bladder cancer over the past several decades. rate of squamous cell carcinoma started well ahead of the decline for incidence rates for adenocarcinoma; similar . Rates of squamous cell and trends are seen for women Changes Over Time in the Types small cell carcinoma have been declining in men since the of Lung Cancer Associated With early- to mid-1980s, but rates of adenocarcinoma did not peak until the 1990s (Travis et al. 1996; Wingo et al. 1999; Smoking Devesa et al. 2005). Age-standardized rates in women Adenocarcinoma of the lung has been increasing reflect their later uptake of smoking, resulting in a later in the United States since the 1970s (Travis et al. 1996; year of peak smoking-induced rates of lung cancer, and Wingo et al. 1999), as manifested in rising incidence rates the patterns are more difficult to interpret. However, rates and an increasing proportion of all lung cancers that are of squamous cell carcinoma leveled off among women adenocarcinomas (Wingo et al. 1999; Devesa et al. 2005). around 1990, but their rates of adenocarcinoma continued Theoretically, this increase could be due to changes over to increase through the 1990s (Wingo et al. 1999; Devesa time in the classification of tumors, but an analysis by et al. 2005). The recent trends in rates for the NSCLCs Charloux and colleagues (1997) found the increase to have been affected by trends in diagnostic practice, 161 Cancer

191 Surgeon General’s Report Figure 6.10 Standardized incidence of lung cancer, by gender and histology (age adjusted to 2000 U.S. population), 1973–2010 A. Males 40 30 Adenocarcinoma 20 Squamous cell carcinoma 10 Other NSCLC Age-adjusted incidence per 100,000 Small cell carcinoma Large cell carcinoma 0 2010 2005 1990 2000 1985 1975 1970 1980 1995 Year of diagnosis B. Females 40 30 Adenocarcinoma 20 Other NSCLC 10 Squamous cell carcinoma Age-adjusted incidence per 100,000 Small cell carcinoma Large cell carcinoma 0 1980 2005 2000 1995 1990 1985 1970 1975 2010 Year of diagnosis Surveillance, Epidemiology, and End Results (SEER) Program, public use data. Source: Note: Other non-small-cell-lung carcinoma (NSCLC) includes code 8046 from the SEER Registry, as well as others. In the most recent years (2001–2010), most of the “Other NSCLC” were 8046. Before 2001, most “Other NSCLC” were coded as 8010 “Carcinoma, NOS.” Around 2004 there were changes in how lung cancers were coded in the SEER Registry data (Travis et al. 2004, 2011; Johnson et al. 2007). There were also advances in diagnosis and treatment around 2004 (erlotinib or gefitinib for patients with EGFR mutations, bevacizumab for patients with non-squamous NSCLC) that make accurate histologic classification important (Langer et al. 2010; Kulesza et al. 2011; Conde et al. 2013). 162 Chapter 6

192 —50 Years of Progress The Health Consequences of Smoking reflecting treatment approaches that are targeted by his - different, as it must be, with the earliest cohorts having tologic type. There has been a trend to avoid nonspecific only the older age ranges and the more recent cohorts classification and to designate lung cancers as adenocar only the younger age ranges. - cinoma and squamous cell carcinoma (Langer et al. 2010; - Data from the SEER Program do not contain infor Travis et al. 2011; Conde et al. 2013). mation about smoking status at the individual level, but the birth-cohort rates for the different histologic types Interpreting age-standardized rates of lung cancer is presented in the figures result from a steadily progress difficult because of variations in the prevalence of smok- - ing, in the distribution of duration of smoking, and in ing mixture of current, former, and never-smoking behav - the distribution of the duration of abstinence in the U.S. iors that are specific for each cohort as it moves forward population over the past several decades. For that reason, in time. Therefore, differences in the proportion of lung - rates of lung cancer by histologic type have also been cancers due to a specific histologic type are not due to dif examined by birth cohorts. This approach examines out - ferences by histology in overall smoking behaviors, given comes as the population born during the selected calendar that these behaviors are the same for all of the histologic years initiates and quits smoking over time (and ages, as types in any given calendar year. Differences by histologic type within a cohort can reflect differences in the relation - well). These two smoking behaviors have been found to differ substantially across sequential birth cohorts for the ship of age to histologic type, differences in the rate of decline in risk after smoking cessation for the different U.S. population (Burns et al. 1997, 2011b). histologies, or variation in the exposures over time in the Zheng and colleagues (1994) found that birth- agents causing the different types of lung cancer. cohort-specific rates of lung cancer by histologic type Effects due to aging, such as those that might be - across calendar years in the Connecticut Tumor Reg manifested if the durations of smoking required to pro - istry data demonstrated a clear birth-cohort pattern for duce squamous cell carcinoma and adenocarcinoma are increased rates of adenocarcinoma; that is, there were identifiable differences in rates by cohort. These changes - different, would likely reveal themselves in a similar fash paralleled gender and generational changes in smoking ion across all cohorts as those cohorts reach the appro- rather than advances in diagnostic procedures (Thun et priate ages, but Figures 6.11 and 6.12 do not indicate a al. 1997a). In this Connecticut study, the birth-cohort consistent pattern with age. trends for squamous cell carcinoma were consistent with The time course of reduction in excess risk of lung cancer after cessation of smoking likely differs for the dif- changes in smoking prevalence by birth cohort over time, ferent histologic types. For example, some data suggest but rates of adenocarcinoma by birth cohort progres - sively increased for both men and women in a manner that excess risks for squamous cell and small cell lung that was not consistent with changes in smoking preva- cancers may decline more rapidly after cessation than do lence by birth cohort (Zheng et al. 1994). This increase excess risks for adenocarcinoma (Kenfield et al. 2008). As calendar years have advanced, the U.S. population in the was consistent with an increase over time in the risk of adenocarcinoma associated with smoking due to changes age groups at substantial risk for lung cancer (i.e., those in the design of cigarettes, including the introduction of over 50 years of age) is composed of an increasing fraction of former smokers, and those former smokers have had filters and low-tar cigarettes (Zheng et al. 1994; Thun et longer durations of abstinence. al. 1997a). The potential effect of a Figures 6.11 and 6.12 present incidence rates for slower decline in risk for adenocarcinoma raises the pos- sibility that the decline in squamous cell carcinoma and lung cancer by histologic type based on 5-year birth cohort the increase in adenocarcinoma over time may be a result data from the SEER Program. Although the proportion of of a relatively more rapid decline in risk for squamous cell lung cancer that is adenocarcinoma is somewhat higher for women across all birth cohorts, a trend is found in carcinoma, leaving an increasing fraction of lung cancer as adenocarcinoma. However, if the increasing proportion - which adenocarcinoma represents an increasing propor of lung cancer that is adenocarcinoma was in fact due to tion of lung cancer when sequential cohorts are examined this effect (of a less rapid decline in the excess risk for for both men and women. Data in Figures 6.11 and 6.12 are combined in Figure 6.13 to present mean values for adenocarcinoma following cessation), then the greatest shift would be in the earliest birth cohorts, among whom the proportions of all lung cancers with a designated his- the effects of differences in risk with abstinence would be tologic type that were adenocarcinoma for those cohorts most evident. Figures 6.11 and 6.12 show the opposite with data available. The mean values demonstrate a sub - pattern—the greatest increase in the proportion of lung stantial increase in the proportion of lung cancer that is adenocarcinoma when moving from the earliest to the cancer that is adenocarcinoma occurs in the more recent birth cohorts who are younger in age and have less cumu more recent cohorts. An important caveat in interpret - - lative abstinence. ing these means is that the age range for each cohort is 163 Cancer

193 Surgeon General’s Report Figure 6.11 Incidence of lung cancer among U.S. men from various birth cohorts, by histologic type (adenocarci - noma, squamous cell carcinoma, and small and large cell carcinoma) and year of diagnosis, 1975–2000 164 Chapter 6

194 The Health Consequences of Smoking —50 Years of Progress Figure 6.11 Continued Cancer 165

195 Surgeon General’s Report Figure 6.11 Continued 166 Chapter 6

196 The Health Consequences of Smoking —50 Years of Progress Figure 6.11 Continued Cancer 167

197 Surgeon General’s Report Figure 6.11 Continued 168 Chapter 6

198 —50 Years of Progress The Health Consequences of Smoking Continued Figure 6.11 Source: Surveillance, Epidemiology, and End Results Program, public use data. 169 Cancer

199 Surgeon General’s Report Figure 6.12 Incidence of lung cancer among U.S. women from various birth cohorts, by histologic type (adenocarci - noma, squamous cell carcinoma, and small and large cell carcinoma) and year of diagnosis, 1975–2000 170 Chapter 6

200 The Health Consequences of Smoking —50 Years of Progress Figure 6.12 Continued Cancer 171

201 Surgeon General’s Report Figure 6.12 Continued 172 Chapter 6

202 The Health Consequences of Smoking —50 Years of Progress Figure 6.12 Continued Cancer 173

203 Surgeon General’s Report Figure 6.12 Continued Source: Surveillance, Epidemiology, and End Results Program, public use data. 174 Chapter 6

204 The Health Consequences of Smoking —50 Years of Progress Figure 6.13 Unweighted mean percentage of all lung cancers that were adenocarcinoma, by gender and birth cohort for the available calendar years, United States, 1890–1955 Surveillance, Epidemiology, and End Results Program, public use data. Source: 175 Cancer

205 Surgeon General’s Report The birth-cohort pattern observed in Figures 6.11 population selection bias and differences between rates of - and 6.12 suggests that changes in the design and com incidence and mortality and then were scaled, based on position of cigarettes may be a factor that is driving the the fraction of lung cancers of the appropriate histologic type in the SEER Program data for the first years available increase in rates of adenocarcinoma (Charloux et al. 1997; Thun et al. 1997a; NCI 2001). Risk of lung cancer reflects (1973–1975). The predicted rates for squamous cell carcinoma cumulative exposure to cigarette smoke, and if a change in and adenocarcinoma by 5-year birth cohort were com- the design or composition of cigarettes increases the risk pared with the rates observed in data from the SEER of lung cancer from smoking, then the onset of increasing Program for the same cohorts during the calendar years - risk begins at the time when the change is made. Each suc ceeding cohort would have a larger fraction of its cumula - 1973–2000. For squamous cell carcinoma, the predicted tive smoking exposure from the new cigarettes, as existing rates closely matched the rates from the SEER Program, brands are refashioned and smokers switch to brands with suggesting that much of the variability in the incidence greater risk characteristics. This increased risk becomes rates of squamous cell carcinoma over the past several stronger in successive birth cohorts, particularly if use - decades can be explained by changes in the rates of smok of the newer, more hazardous product is more common ing prevalence and cessation. In contrast, the predicted among younger than older smokers. Among older indi rates for adenocarcinoma did not match data in the SEER - viduals from the earlier birth cohorts, rates of lung cancer Program, and the differences between predicted rates and will continue to be dominated by the substantial con - those of the SEER Program varied systematically by birth cohort. When a term increasing the risk for adenocarci - tribution of their past smoking, and an increase in risk noma with duration of smoking after 1950 was added to resulting from a more recently changed cigarette product will make a relatively modest proportional contribution the risk model for current and former smokers (to simu- to the pre-existing and already substantial risk for these late an increasing risk over time associated with a change in the design of cigarettes), the predicted rates matched cohorts. As more recent birth cohorts are examined, the the rates from the SEER Program. Thus, these analyses onset of increasing risk due to a change in product design will begin at an earlier age because members of the cohort suggest that increasing risk of lung cancer over time may be associated with changes in the design or composition of will begin smoking the newer products at a younger age. The increment in risk with the use of the newer products cigarettes. The analyses also raise the possibility that the reflects a larger proportion of the total risk for the cohort, increase in overall lung cancer mortality from smoking may reflect an increase in the risk of developing adeno - simply because the duration of smoking preceding the carcinoma from smoking, with little change in the risk of shift to a more dangerous type of cigarette is shorter and developing squamous cell carcinoma. thus the risk for that earlier period as a fraction of total risk is smaller. Such an effect could explain the progres - Some researchers have suggested alternative expla - sive increase in the proportion of lung cancers that are nations for the increase in lung adenocarcinoma. Based on birth-cohort analyses of data from the SEER Program adenocarcinomas across sequential cohorts, as shown in Figure 6.13. and differences in the temporal trends in the incidence - Differences in the prevalence of current and for of squamous cell lung cancer and adenocarcinoma of the lung, Chen and colleagues (2007b,c, 2009) suggested an mer smoking and differences in the distribution of the effect of air pollution, and specifically nitrogen oxides, duration of smoking and the duration of abstinence from - as the cause for the trends in adenocarcinoma. However, smoking vary markedly across birth cohorts and con because among never smokers both lung cancer mortality tribute to differences in risks of lung cancer. To account and the incidence of adenocarcinoma do not seem to have for these differences in the examination of rising rates of - adenocarcinoma, birth-cohort-specific smoking behaviors changed over time and because the risk of adenocarci noma among smokers has increased, changes in cigarette - have been used to model changes in the rates of lung can cer of different histologic types (Burns et al. 2011a), as smoking are a more likely cause of the temporal trends than air pollution. was done for overall lung cancer mortality and incidence Changes in the demographics of smokers are rates. Risk models derived from CPS-I were applied to another potential explanation. Over time, the poorer and the smoking behaviors of birth cohorts. These behaviors less-educated segments of the population have become a include rates of smoking initiation, prevalence of current progressively greater fraction of U.S. smokers (see Chap and former smoking, and distributions of the duration of - ter 13, “Patterns of Tobacco Use Among U.S. Youth, Young smoking, duration of abstinence, and number of cigarettes smoked per day for current and former smokers (Burns et Adults, and Adults”). Within birth cohorts, an increas - al. 2011a). The resulting rates were adjusted for a healthy ing proportion of smokers come from population groups 176 Chapter 6

206 The Health Consequences of Smoking —50 Years of Progress rising proportion of lung cancer that was adenocarcinoma characterized by less education and lower income and ces - (Devesa et al. 2005). When examined at the national level, sation rates are lower in these groups as well, compared however, the rates of increase of adenocarcinoma and the with those having more education and higher incomes. - patterns of the shift to adenocarcinoma as the most com Occupational and environmental exposures associated with increased lung cancer risk are also more common mon form of lung cancer varied among countries (Devesa among those with less education and lower income. As a - et al. 2005). In many countries—such as European coun - result, the effects of this demographic shift should be rela tries (Devesa et al. 2005), including Italy (Russo et al. tively uniform across cohorts, unlike the pattern observed 1997); Japan (Yoshimi et al. 2003); and Hong Kong (Tse in the figures. In addition, a demographic shift of this type - et al. 2009)—the patterns among men have roughly mim would not affect rates of icked those of U.S. men, with falling rates of squamous in the characteristics of smokers adenocarcinoma or squamous cell carcinoma or would cell carcinoma and initially rising but then falling rates of adenocarcinoma. Among women, interpretations of not affect rates of adenocarcinoma or influence rates of . changes in rates of cancer by histologic type need to con squamous cell carcinoma - sider the rising rates of smoking prevalence for women. In summary, adenocarcinoma has been increas- ing in the United States as a fraction of all lung cancers, Regardless, rates of adenocarcinoma rose faster than rates becoming the most common histologic type of lung can - of squamous cell carcinoma in most countries for which cer. Despite decreases in smoking prevalence and concom data were available (Devesa et al. 2005). - itant decreases in squamous cell carcinoma, the incidence As described previously, flue-cured cigarettes of of lung adenocarcinoma among smokers has increased the type preferred in Australia, Canada, and the United since the 1960s. Changes in the design and/or composi - Kingdom have substantially lower levels of tobacco-spe - - cific nitrosamines than do U.S.-style blended cigarettes tion of cigarettes during the 1960s have increased the lev els of tobacco-specific nitrosamines and other carcinogens and have higher levels of B[ ]P (WHO 2008b). Tobacco- a in cigarette smoke. Evidence from birth-cohort models specific nitrosamines, specifically NNN and NNK, are organ-specific carcinogens for adenocarcinoma of the and epidemiologic studies are sufficient to conclude that lung in animal models (IARC 2007; USDHHS 2010); NNK - the increased risk of lung adenocarcinoma among smok selectively induces adenocarcinoma of the lung in rats, ers is due to changes in the design and/or composition mice, and hamsters. The level of NNAL, a metabolite of of cigarettes which increased the carcinogenicity of ciga- rette smoke. NNK, in the urine of smokers has been shown to be an independent predictor of risk for lung cancer even when the analysis controls for intensity (by cotinine concentra- Evidence for a Rising Risk of Adenocarcinoma of the Lung in the United States tion) and duration of smoking (Church et al. 2009; Yuan et al. 2009). Differences Across Time in Rates of In terms of PAHs, one prospective cohort study found that a biomarker (phenanthrene tetraol) for PAH Adenocarcinoma Within the United States and Across Countries exposure was not an independent predictor of risk for lung cancer (Church et al. 2009). When the risk for lung cancer In a population, the principal determinants of risk was examined by histologic type in this study, however, for lung cancer are the prevalence of current smoking a significant association was found between NNAL in the and the distribution of the duration of smoking among urine and adenocarcinoma of the lung. The relationship current and former smokers. As described previously, between NNAL and risk for lung cancer was not significant assessing the impact of differences in population-based for all other types of lung cancer combined, and the odds smoking behaviors on rates of lung cancer is a complex ratios for adenocarcinoma and other lung cancers did not undertaking. Even so, some understanding can be gained differ significantly from each other (Church et al. 2009). by comparing rates of lung cancer in countries where - Mouth-level exposure to tobacco-specific nitro smokers have similar behaviors but smoke different types samines in smoke has been examined among smokers of cigarettes. in countries with high use of blended cigarettes (United Incidence rates of adenocarcinoma of the lung and States) and flue-cured unblended cigarettes (Australia, the proportions of adenocarcinoma in relation to all lung Canada, and the United Kingdom) (Ashley et al. 2010). cancers increased in most countries through 1995–1997 Levels of NNK exposure among Australian and Cana - (Devesa et al. 2005). These trends were particularly evident dian smokers were approximately one-third that of U.S. among women and reflected the higher risk of lung cancer smokers, and levels of NNN exposure were 85–90% lower accompanying their increasing smoking prevalence and a than the U.S. experience. Among smokers in the United 177 Cancer

207 Surgeon General’s Report Kingdom, levels of NNK exposure were 20% lower than men and women in Australia compared with the United those of smokers in the United States, and levels of NNN States. Detailed comparisons of smoking behaviors similar were approximately 50% lower (Ashley et al. 2010). to those used to model U.S. death rates are not available In England and Scotland, flue-cured cigarettes for Australia, but estimates of the prevalence of smoking show a general similarity for Australia and the United remain popular, but measures of the level of exposure to tobacco-specific nitrosamines fall between those observed States, particularly during the 1990s (White et al. 2003). in smokers in the United States and in smokers in other - Figure 6.16 presents information on adenocarci noma as a proportion of all lung cancers with a designated countries where unblended cigarettes are common histologic type, by birth cohort and gender for the United (Canada and Australia) (Ashley et al. 2010). In England (Bennett et al. 2008) and Scotland (Harkness et al. 2002), States and Australia (Burns et al. 2011a). In Australia, a incidence rates of adenocarcinoma of the lung for men modest rise occurs in the proportion of lung cancers that are adenocarcinoma across the birth cohorts for both gen - have increased only slightly, and squamous cell carcinoma ders, but the fraction remains well below 50% for men and remains the predominant lung cancer. Rates of squamous - only slightly above 50% for women. Data for the United cell carcinoma among men in those countries are declin ing consistently as smoking prevalence drops. - States show a much more dramatic increase in the pro In Canada, the incidence rate of adenocarcinoma portion of lung cancer that is adenocarcinoma, with the among men in 1995−1997 remained lower than that of proportion exceeding 60% in the most recent cohorts for White men and women. Notably, the earliest birth cohorts squamous cell carcinoma and well below the rate for White men in the United States (Figure 6.14) (Devesa et for the U.S. population, those born from 1880–1900, have proportions similar to those found in Australia. al. 2005). In contrast, rates for squamous cell carcinoma were similar for men in the United States and Canada in In summary, rates for squamous cell carcinoma of the lung have been decreasing in most countries in which this period and in women as well (Devesa et al. 2005). the prevalence of smoking has been declining. In contrast, - Based on data up to 1997, the incidence of adenocarci the incidence rate of adenocarcinoma has been rising in noma of the lung did not appear to be increasing over time the United States and has been level or increasing in other in Canada. Instead, the data suggest that squamous cell carcinoma was decreasing so that adenocarcinoma repre- countries, with the general result that adenocarcinoma - sented an increasing fraction of lung cancers over time has increased as a proportion of lung cancer in most coun tries. The magnitude of that increase has differed between (Devesa et al. 2005). In Australia, where flue-cured, unblended cigarettes the United States, where the predominant type of cigarette with low tobacco-specific nitrosamine levels are also is made of blended tobacco with relatively high levels of prominent, the rate of adenocarcinoma among men rose tobacco-specific nitrosamines, and Canada and Australia, across birth cohorts and over time and exceeded the rates where flue-cured cigarettes with lower levels of tobacco- specific nitrosamines predominate. Incidence rates of of squamous cell carcinoma for the most recent cohorts adenocarcinoma and the proportion of lung cancer that - (Blizzard and Dwyer 2002). In contrast, the rate of adeno carcinoma among men in New South Wales, Australia, is adenocarcinoma are substantially higher in the United States than in Canada and Australia. remained essentially constant between 1985 and 1997 (Figure 6.14) (Devesa et al. 2005) or rose only slightly over time. However, the rate for squamous cell carcinoma in Effects of Filter Ventilation on Deposition of Smoke in the Lung and the Toxicity of 1995−1997 among New South Wales men declined to a level approximating that of adenocarcinoma (Figure 6.14) This Smoke (Joshua et al. 2005). Similarly in South Australia, the rate One potential explanation for the rise in adenocar - of adenocarcinoma among men through 2000 was also cinoma of the lung in the United States is a change in relatively consistent over time, and the rate of squamous the pattern of smoking after ventilated filters were intro - cell carcinoma fell to the same level as adenocarcinoma duced to lower the machine-measured yields of tar and (Nguyen et al. 2003). However, in South Australia, the rate nicotine (Zheng et al. 1994; Thun et al. 1997a; Wingo et al. of adenocarcinoma increased among younger age groups. 1999). Smokers who shift to brands with nominally lower When comparing the United States and Australia, machine-measured yields with ventilated filters change the different patterns of cigarettes smoked may contribute their smoking pattern to restore their nicotine delivery to to different patterns of lung cancer. Figure 6.15 presents the level needed to sustain their addiction. As described gender- and age-specific rates of lung cancer mortality for previously, changes include increasing puff volume and the United States and Australia for 2000 (Peto et al. 2006). velocity, greater duration of puffing, and shortening the Lung cancer death rates were lower in all age groups for 178 Chapter 6

208 —50 Years of Progress The Health Consequences of Smoking Trends in incidence rates for lung cancer (age adjusted, world standard), by histologic type (squamous Figure 6.14 cell carcinoma, small cell carcinoma, and adenocarcinoma) and geographic area, 1980–1982 to 1995– 1997 179 Cancer

209 Surgeon General’s Report Figure 6.14 Continued 180 Chapter 6

210 The Health Consequences of Smoking —50 Years of Progress Continued Figure 6.14 Source: Devesa et al. 2005. Reprinted with permission from John Wiley & Sons, Inc., © 2005. Note: Incidence rates were calculated for 3-year time periods: 1980–1982 (plotted at 1981.5), 1983–1985 (plotted at 1984.5), 1986– 1988 (plotted at 1987.5), 1989–1991 (plotted at 1990.5), 1992–1994 (plotted at 1993.5), 1996 (Switzerland only; plotted at 1996), and 1995–1997 (plotted at 1996.5). Incidence rates were age-adjusted by the direct method, using the Segi world standard (Bray et al. 2002), and expressed per 100,000 person-years. 181 Cancer

211 Surgeon General’s Report Figure 6.15 Age-specific rates of lung cancer death, by gender and age group, in the United States and Australia, 2000 Source: Peto et al. 2006. 182 Chapter 6

212 The Health Consequences of Smoking —50 Years of Progress Adenocarcinoma as a percentage of designated lung cancers in U.S. White men and women and Austra Figure 6.16 - lian men and women, by various birth cohorts, 1890–1955 Source: Burns et al. 2011a. Reprinted with permission from Science & Business Media B.V., © 2011. Data for the Australian national cancer registry provided by Helen Farrugia, Director Information Systems, Cancer Epidemiol - Note: ogy Centre, The Cancer Council Victoria. 183 Cancer

213 Surgeon General’s Report flue-cured, unblended cigarettes smoked by a machine intervals between puffs (NCI 2001). In addition, smokers (Adam et al. 2010). When experimental (Rickert et al. 2007) may increase the depth of inhalation and hold the smoke - or commercial (Roemer et al. 2004) U.S.-blended ciga in their lungs longer to increase nicotine uptake. Notably, there is little difference in markers of nicotine ingestion rettes were compared with experimental, unblended, flue- between smokers of brands of cigarettes with substantially cured cigarettes (Monitor-7 Canadian reference cigarette) different machine yields (Jarvis et al. 2001; NCI 2001). in mutagenicity testing, the level of revertants per mg (the Increasing depth of inhalation and other more intense indicator of mutational strength) of the total particulate matter was lower for the unblended Canadian reference smoking patterns likely increase the deposition of smoke in the alveolar region of the lung. cigarette. For Kentucky reference cigarettes, mutagenic - Most physical models of particles disseminating in ity per mg of total particulate matter was 30–40% lower - the lung incorporate the size-dependence of particle depo for unfiltered cigarettes than for the same cigarette with a sition in the lung, but do not fully reflect the complexity filter added (Shin et al. 2009). - of smoking behavior. As a consequence, the models may Tobacco industry documents show internal com - underestimate the fraction of smoke particles retained in pany research demonstrating that increasing filter venti lation increases the mutagenicity of the resultant tar on the lung (Stratton et al. 2001; Gower and Hammond 2007; Rostami 2009), raising questions about their validity in a per-mg of tar basis (Johnson et al. 2009). The published characterizing the distribution and deposition of particles evidence produced by the industry is less clear. In a study from R.J. Reynolds, Chepiga and colleagues (2000) com - in different regions of the lung with different tobacco pared full-flavor, full-flavor low-tar, and ultralow-tar cig- products. An analysis by Gower and Hammond (2007) of models of cigarette smoke deposition that examined the arettes and reported a nonsignificant trend of increased effects of the changes in pattern of smoking after a shift to revertants per mg of tar in mutagenicity studies as the level of machine-measured tar decreased. In a study from brands with lower machine-measured yields showed that Philip Morris, Roemer and colleagues (2004) reported that puff time, inhalation depth, time holding one’s breath, and exhalation time may affect total smoke deposition. higher total yields of particulate matter were associated with a trend toward less mutagenic activity per mg of total While a shift in deposition to the alveolar level remains a particulate matter. In another study from Philip Morris, possibility, the researchers could not determine whether Patskan and colleagues (2008) compared the mutagenic the changes in patterns of smoking resulting from the use of more highly ventilated cigarettes could produce a large activity of Marlboro full flavor, Marlboro Lights, and enough shift in the location of deposition to change the Marlboro Ultra Lights, finding that mutagenic activity was higher per mg of total particulate matter for Marl- pattern of incidence of a specific histologic type of lung cancer. Although the magnitude of the potential change in Salmo- boro Ultra Lights, but this was for only some - strains used in the mutagenicity testing and for nella regional deposition in the lung remains uncertain, exist - only some runs. Thus, the evidence supports a modest ing evidence suggests that changes in the pattern of smok ing, with a shift to lower tar-yield cigarettes, will likely increase in the mutagenicity of tobacco tar as the level - increase the fraction of cigarette smoke particles depos of machine-measured tar falls; this effect may result from increased ventilation. ited in the alveolar region of the lung. This shift may also These data should be interpreted with caution for have played a role in increasing the risk of adenocarci - noma of the lung over time. several reasons. Mutagenicity is generally used as only a screening test, is often poorly associated with carcinoge - The introduction of ventilated filters, or changes in nicity in humans, and has not been quantitatively asso - the design and composition of cigarettes that accompa- ciated with differences in human risk. In addition, most - nied their introduction, may have increased the carcino of the studies described previously compared smoke gen- genicity of cigarette smoke. Given the dilution of smoke erated under standardized machine-testing protocols. In by filter ventilation and the compensation for that dilution actual use, smokers change their patterns of smoking, by smokers when these cigarettes are used, comparisons compensating for the design changes that result in lower of the toxicity of cigarettes on a per-cigarette basis can be yields of machine-measured tar and nicotine. This com - misleading, making comparisons on the basis of “per mg tar” or “per mg total particulate matter” more useful. - pensatory smoking behavior makes comparisons of ciga rettes with very different machine-tested yields difficult - The level of filter ventilation alters the composi - Inter tion of tobacco smoke. In general, based on the to interpret relative to carcinogenicity in humans when the smoke for the different cigarettes is generated using a national Organization for Standardization protocol and under more intense smoking parameters, higher levels single, standardized, machine-smoking protocol. of ventilation result in more complete combustion in 184 Chapter 6

214 —50 Years of Progress The Health Consequences of Smoking Existing evidence about changes in the patterns of contemporary cohorts show that lung cancer risk con - smoking cigarettes with low yields of tar and high venti- tinued to rise through 2000–2010. Modeling of risks of lung cancer from smoking behaviors suggests that risk lation supports a shift in the deposition of smoke in the lung toward the alveolar region; this shift likely contrib - estimates based on the smoking experience in the 1960s utes to the observed increase in adenocarcinoma of the underestimated the current incidence of lung cancer. In lung. Research has not clarified whether the magnitude of addition, the incidence of adenocarcinoma of the lung and this shift in lung deposition, by itself, is great enough to the proportion of lung cancer that is adenocarcinoma has explain the dramatic increase in adenocarcinoma observed increased dramatically during the past several decades. This shift from squamous cell carcinoma to adenocarci - in the United States. The mutagenicity of tobacco tar from cigarettes with lower yields of machine-measured tar is noma is confined to smokers, because neither the overall risk of lung cancer nor the risk of adenocarcinoma has trending upward. However, the trend is modest in size, and difficulties in extrapolating results from mutagenic - - changed over time among never smokers. The rate of squa mous cell carcinoma of the lung has declined in the United ity testing to risk for humans make it difficult to know States since the 1980s and is well-predicted by declines in whether these changes contribute to increasing the risk - smoking behaviors, but the rate of adenocarcinoma con of lung cancer. tinued to rise for an additional 10–15 years before either leveling off or beginning to decline. Birth-cohort-specific analyses of trends in overall mortality from lung cancer Evidence Synthesis and the incidence of type-specific lung cancer suggest that increases in diagnostic accuracy, differences by tumor The design and composition of cigarettes have type in the time course of excess risk reduction with ces- changed substantively since the first major wave of evi - sation, and underestimation of the effect of intensity and dence linking smoking to lung cancer in the 1950s. duration of smoking in the studies that defined risk in the Although the details of these changes are only partially 1960s do not explain the observed trends. In contrast, a understood, changes in design—notably the addition of change in the risk of the cigarettes smoked over time does ventilated filters—have clearly changed the pattern of explain the increase in risk. A shift in the demographic smoking, including more intense puffing. In addition, composition of smokers toward those groups with less changes in the composition of cigarettes have resulted income and education may contribute to the increased in incompletely characterized alterations in the chemi- risk of lung cancer among smokers, but this shift does not cal composition of cigarette smoke. Documented changes likely explain the increase in adenocarcinoma or the differ - include increases in tobacco-specific nitrosamines and ence in the rates of incidence of squamous cell carcinoma decreases in PAHs in the smoke of U.S. cigarettes. Sub - and adenocarcinoma. stantial differences between U.S. cigarettes and those of Most countries have experienced increases in the many other nations include the use of blended tobacco proportion of all lung cancer that is adenocarcinoma, in U.S. cigarettes and the use of unblended, flue-cured but substantial differences are found in the extent of tobacco in cigarettes in Australia, Canada, and the United this increase when comparing the United States, where Kingdom. The United States has somewhat preceded most blended cigarettes are used, with Australia and Canada, other developed countries in the adoption of filtered and where unblended cigarettes are used. Adenocarcinoma in low-yield, machine-tested cigarettes, but U.S. products the United States has increased more steeply, represents a are also used widely in most countries. These changes much higher fraction of lung cancer, and has higher abso - raise a question of whether rates of lung cancer have been lute incidence rates than those of Australia or Canada. altered by the changes in the design and composition of Compared with unblended cigarettes, U.S.-style blended cigarettes—changes that were accompanied by an initial cigarettes have dramatically higher levels of tobacco- belief that lower yields of machine-tested tar might signal specific nitrosamines—an organ-specific carcinogen - a lower risk for lung cancer. In fact, the risk of lung can of adenocarcinoma of the lung in animals. Exposure to cer in the United States may have increased as a result of tobacco-specific nitrosamines is also much higher among such changes. U.S. smokers than among their counterparts in Austra - Comparison of results of CPS-I and CPS-II—two lia and Canada. Levels of a metabolite of NNK, a tobacco- large epidemiologic studies conducted 20 years apart by specific nitrosamine, are an independent risk predictor ACS—demonstrated an increased risk of death from lung for the occurrence of lung cancer after controlling for the cancer from smoking across the 20-year interval between intensity and duration of smoking. the studies. For female smokers, the results from the 185 Cancer

215 Surgeon General’s Report est reduction in the large burden of mortality from lung Compensatory changes in the patterns of puffing cancer would result in saving substantial numbers of lives and inhaling smoke by smokers switching to cigarettes with low yields of toxicants may increase the deposition over time. of smoke particles in the alveolar region of the lung. This The evidence reviewed suggests that differences in the design and composition of cigarettes may contribute is supported by modeling of particle deposition in the to differences in smoking-related risks of lung cancer in - lung that suggests this effect likely increases the deposi different populations and different geographic locations. tion of particles in the alveolar region. Increased alveolar Data also suggest that epidemiologic studies treating deposition and increasing tobacco-specific nitrosamine all cigarettes as having identical risks, or using single levels over time may have combined to increase the risk for adenocarcinoma. biomarkers of exposure to quantify actual exposure to the multiple carcinogens in cigarette smoke, should be undertaken with some caution. The number of cigarettes smoked per day, measures of cotinine in biologic samples, Conclusions and other measures of total smoke exposure will remain - useful for estimating total smoke exposure and popula The evidence is sufficient to conclude that the risk 1. - tion risk. However, the potential for differences in prod - of developing adenocarcinoma of the lung from ciga ucts to yield differences in risk suggests that a broader rette smoking has increased since the 1960s. array of biomarkers of exposure should be used to examine whether differences in the toxicity and composition of a 2. The evidence is sufficient to conclude that the given total exposure to smoke may also play an important increased risk of adenocarcinoma of the lung in role in determining differences in risks. - smokers results from changes in the design and com The changing risk for lung cancer associated with position of cigarettes since the 1950s. cigarettes over time also has implications for the surveil - lance of tobacco products. Monitoring tobacco products 3. The evidence is not sufficient to specify which design needs to go beyond tracking the most obvious changes, changes are responsible for the increased risk of ade- such as the addition of a filter, to assess the characteristics nocarcinoma, but there is suggestive evidence that - of the tobacco in the cigarette, how the product is manu ventilated filters and increased levels of tobacco-spe - factured, how it is likely to be smoked, the design of the cific nitrosamines have played a role. cigarette, and its performance under a variety of smok - ing patterns. The absence of such information for past and 4. The evidence shows that the decline of squamous cell current tobacco products limits the ability to more fully carcinoma follows the trend of declining smoking study the effects of changes in the design and composition prevalence. of cigarettes on risks of disease. The availability of such information could help in the assessment of potential dif - ferences in risks going forward. Implications Finally, the rise in the risk of adenocarcinoma of the lung from smoking was unanticipated. This experience, The evidence presented has multiple implications. - like that of cigarettes with purportedly low yields of toxi Above all, if the risk of lung cancer has increased with cants, indicates that changes to cigarettes should undergo changes in the design and composition of cigarettes, careful, evidence-based assessments as such changes are then the potential exists to reverse that increase in risk being considered. through changes in design and composition. Even a mod - Chapter 6 186

216 —50 Years of Progress The Health Consequences of Smoking Liver Cancer disease, differences in smoking and in consumption of In many parts of the world, liver cancer remains a leading cause of cancer mortality. Primary liver cancer, alcohol, and perhaps hormonal differences. the great majority of which is hepatocellular carcinoma The association between smoking and HCC is com - (HCC), generally presents at an advanced stage with lim- plicated by the potential for confounding with the causal ited treatment options and a poor prognosis. Although factors of consumption of alcohol and HBV and HCV infection. For example, people who drink alcohol are worldwide liver cancer is the sixth most common cancer more likely to be smokers than people who do not drink in terms of incidence, it represents the third most com- mon cause of cancer-related death (Ferlay et al. 2010). alcohol (Dawson 2000). In addition, most HCV infections A number of strong risk factors for HCC have been worldwide are acquired by injecting drugs, and the preva - lence of smoking is very high among injection drug users identified, including infection with the hepatitis B or C (Marshall et al. 2011). In regions of the world with a high viruses (HBV, HCV), exposure to aflatoxins, and alcohol- incidence of HCC, HBV infection is generally acquired associated cirrhosis (London and McGlynn 2006). The perinatally or during early childhood. However, in other incidence of liver cancer varies geographically worldwide, regions, HBV may be more commonly acquired through with rates generally consistent with the regional preva- lence of the primary viral etiologic factors (Nordenstedt parenteral or sexual transmission; these behaviors may also be associated with smoking. Hence, the potential et al. 2010). Globally, Asia and sub-Saharan Africa—with confounders must be examined carefully when assess - endemic HBV infection and common dietary exposure ing the association between smoking and HCC. However, to aflatoxins—have the highest incidence of HCC. Rates considerable epidemiologic evidence, including data from of HCC appear to have stabilized or started to decline in several Asian countries, where widespread vaccination studies in which measures have been taken to address potential confounding, indicates that smokers are at an against HBV and reduction of HBV cofactors have occurred increased risk for liver cancer (IARC 2004). during the past few decades (Yuen et al. 2009). HCV infection has been the primary etiologic agent for HCC in various countries having substantial incidence of HCC (London and McGlynn 2006). Conclusions of Previous Surgeon - Historically, the United States has had a low inci General’s Reports dence of liver cancer and low death rates for the disease. However, rates of HCC have been increasing in the United The Surgeon General’s report on smoking cessation States over the last two decades (Altekruse et al. 2009; (USDHHS 1990) noted an association between smoking El-Serag 2011). In recent years, Whites and Blacks, par - and HCC that persisted after controlling for potentially ticularly those 50−59 years of age, have experienced the confounding lifestyle factors, including consumption of largest annual percentage increases in rates of HCC; rates alcohol. The report also noted that HBV infections may of HCC among Asians/Pacific Islanders have been stable modify the effects of smoking on the risk of liver cancer. (O’Connor et al. 2010). The increased rates of HCC in The Surgeon General’s report on women and smoking the United States appear to be largely a consequence of - (USDHHS 2001) concluded that smoking may be a con chronic HCV infection (El-Serag 2004). However, obesity, tributing factor to the development of liver cancer. The diabetes, and associated nonalcoholic fatty liver disease, Surgeon General’s report on the health consequences of and the substantial burden of chronic HBV infection smoking (USDHHS 2004) noted a consistent association among foreign-born Asians may also be potential con - between smoking and HCC after controlling for poten - tributors to the increasing incidence of HCC (Larsson tially confounding factors, but it called for further consid - and Wolk 2007; Starley et al. 2010). In addition to viral eration of the history of viral hepatitis and consumption of hepatitis, cirrhosis from consumption of alcohol rep - alcohol. Overall, the 2004 report concluded that although resents an important cause of HCC worldwide (London - the data were suggestive of an association between smok and McGlynn 2006). HCC is more common among men ing and liver cancer, further evidence was required to clas - than women, which likely reflects gender differences in sify smoking as a cause of liver cancer. exposure to viral hepatitis and rates of progression of that 187 Cancer

217 Surgeon General’s Report the previous Surgeon General’s reports, the IARC (2004) Biologic Basis monograph on smoking and liver cancer, and the asso- ciated meta-analysis (Lee et al. 2009). The epidemiologic Circulating carcinogens from tobacco smoke are data came from a wide range of studies in both low- and metabolized in the liver, exposing the liver to many high-incidence countries (Tables 6.3 S ). For S and 6.4 absorbed carcinogens. Experimental studies have iden - many studies, the outcome was defined as HCC and was tified several constituents of tobacco smoke (e.g., - based on clinical, radiographic, laboratory (alpha-fetopro N -nitrosodimethylamine, 4-aminobiphenyl) as liver car - tein levels), or pathologic criteria. A minority of studies cinogens (IARC 2004). Limited human data on smoke- relied on linkage to cancer or mortality registries, often related carcinogens have suggested increased levels of using primary liver cancer as the outcome defined by the - 4-aminobiphenyl and PAH adducts in HCC tissues com coding of cancer diagnoses from the International Clas- pared with normal liver tissues (Wang et al. 1998; Chen et sification of Disease for Oncology from or causes of death al. 2002). Therefore, long-term exposure to carcinogens the International Classification of Diseases . Some studies in smoke may lead to cellular damage in the liver and were unable to distinguish between HCC and intrahepatic - contribute to the development of cancer. Cigarette smok cholangiocarcinoma; however, none of these studies were ing may also contribute to liver carcinogenesis through from geographic regions where intrahepatic cholangio - - the development of liver fibrosis (Dev et al. 2006; Mal carcinoma would likely represent a substantial portion lat et al. 2008; Altamirano and Bataller 2010). Similar to of primary liver cancers. Studies that did not explicitly their effects on other fibrogenic conditions (e.g., cardiac, differentiate between primary and secondary liver cancer renal, or pancreatic diseases), components of smoke may (and therefore may have included cancers with a differ - induce pro-inflammatory cytokines, oxidative stress path - ent primary site that had metastasized to the liver) were ways, and direct fibrogenic mediators (e.g., transforming excluded from the analysis. Quantitative analyses included growth factor- β 1, angiotensin II) (Altamirano and Bataller all studies that reported sufficient information to abstract 2010). Smoking has also been recognized as a risk factor or calculate an effect estimate and 95% confidence inter - for primary biliary cirrhosis, which itself can progress to val (CI); these analyses were stratified by study design HCC (Zein et al. 2006; Corpechot et al. 2012; Smyk et al. (case-control or cohort). 2012). Although their results have been inconsistent, sev - This review focused on evaluations of the separate - eral epidemiologic studies have demonstrated that smok effects observed in current smokers, ever smokers, and ing substantially increases the risk for progression from former smokers in comparisons with never smokers or chronic liver disease to HCC (Tsukuma et al. 1993; Mar - nonsmokers; studies with a reference group other than rero et al. 2005; Fujita et al. 2006). Further clarification those never smokers or nonsmokers were excluded (e.g., is needed of the mechanistic and epidemiologic effects comparing heavy smokers with light smokers). The quan- of smoking in relation to potential etiologic agents that titative analyses excluded all studies that compared liver can influence these pathways (chronic inflammation and/ cancer cases with controls who had chronic viral hepati - or oxidative stress associated with HCV infection, obesity, tis, cirrhosis, or other chronic liver disease. Finally, the or diabetes). review separately examined the effects of smoking on HCC in studies that controlled for confounding by the main etiologic factors (HBV, HCV, and consumption of alcohol) Epidemiologic Evidence for HCC in the region under study. Assessment of viral hepatitis status was considered adequate for inclusion in Since the 2004 report of the Surgeon General, 90 the quantitative analysis if the study reported on serologi - additional studies have been published or identified that - cal measurement of HBV surface antigen (HBsAg) or anti report on the association between smoking and liver can - bodies to HCV (anti-HCV) as indicators of chronic HBV or cer. IARC (2004) concluded that there was sufficient evi - HCV infection, respectively. dence of a causal association between cigarette smoking Overall, 113 studies—including 59 case-control and liver cancer. Subsequently, Lee and colleagues (2009) S S ) and 54 cohort studies (Table 6.4 (Table 6.3 )—pro- published a meta-analysis that was based on the studies vided data on smoking and primary liver cancer. These considered in the 2004 IARC report. studies, taken together, offered substantial heterogene - Studies for the current review were compiled by ity in design, study population, assessment of smoking searching the MEDLINE database (from January 1966 exposure, and the reporting of risk estimates. Many stud - to December 2012) using the medical subject headings ies, however, were limited by having few HCC cases and “tobacco,” “smoking,” “liver neoplasms,” or “hepatocel- reported nonsignificant increases in risk associated with lular carcinoma” and by examining references cited in Chapter 6 188

218 The Health Consequences of Smoking —50 Years of Progress evaluated dose response did not formally test for trends; various measures of smoking. Furthermore, many studies however, a substantial proportion of these studies were did not adequately control for potential confounding by major causal factors such as consumption of alcohol or not adequately powered to address such relationships. HBV or HCV infection. In their meta-analysis, Lee and colleagues (2009) sum- In an analysis combining data from 31 studies (12 marized data from 7 studies with evaluable estimates and reported a significant dose-response trend showing case-control and 19 cohort) that reported sufficient infor - - increased risk for liver cancer with higher number of ciga mation to estimate risk for HCC in current smokers com- rettes smoked. However, this effect was notably less appar pared with nonsmokers (Figure 6.17), the overall estimate - for RR was 1.7 (95% CI, 1.5–1.9). The relationship between ent among case-control studies that used hospital-based current smoking and HCC was similar in cohort studies instead of population-based control groups. (overall RR = 1.7; 95% CI, 1.5–1.9) and case-control stud Because of concern for residual confounding of - smoking effects by coinfection with viral hepatitis, the ies (RR = 1.6; 95% CI, 1.2–2.1). When 11 studies (6 case- association between smoking and HCC was evaluated in control and 5 cohort) that controlled for confounding by - - the primary etiologic factors (e.g., HBV, HCV, consump the present review among persons who did not have evi dence for chronic viral hepatitis. In an analysis combin tion of alcohol) were analyzed (Figure 6.18), the RR (1.6; - ing data from 13 studies (9 case-control and 4 cohort) 95% CI, 1.2–2.0) was similar to that in the overall analysis. that estimated risk among persons who were negative for Among these studies that directly addressed confounding, the relationship between current smoking and HCC was markers of chronic HBV or HCV infection (Figure 6.21), stronger in cohort studies (RR = 2.2; 95% CI, 1.4–3.3) the risk of HCC among current or ever smokers was sig - than in case-control studies (odds ratio [OR] = 1.2; 95% nificantly increased (RR = 1.8; 95% CI, 1.2–2.7) in a com- parison with never smokers. After excluding a study that CI, 0.9–1.5). Overall, these findings are similar to those in reported markedly increased risk among persons who the meta-analysis performed by Lee and colleagues (2009) - in association with the 2004 IARC report, which reported were negative for HBV and HCV (Jeng et al. 2009), the esti mated risk was attenuated but still significant (RR = 1.3; a 51% increased risk for liver cancer for current smokers 95% CI, 1.0–1.8). Finally, when the analysis was restricted compared with never smokers (meta-RR = 1.51; 95% CI, to the 3 studies that included only persons negative for 1.37–1.67). The findings of the IARC (2004) review and - both HBsAg and anti-HCV and also adjusted for consump the current review are similar, except that the present tion of alcohol (Kuper et al. 2000; Yuan et al. 2004; Koh et review includes a greater number of studies (31 vs. 20) and includes studies that reported results for only one gen - al. 2011), the RR was 1.7 (95% CI, 1.2–2.5). The present review did not identify any studies that der. Both the present review and the IARC analysis defined directly evaluated the effects of interventions aimed at current smoking as reported at entry into the cohort or at smoking cessation on subsequent risk for liver cancer. the time of diagnosis of liver cancer. Among 23 studies with the requisite data available from Among 26 studies (18 case-control and 8 cohort) the publication (11 case-control and 12 cohort) (Figure with evaluable comparisons between ever smokers 6.22), the risk for liver cancer among persons identified as and never smokers (Figure 6.19), the risk for HCC was increased among ever smokers (RR = 1.4; 95% CI, 1.3– former smokers relative to never smokers was lower (RR = 1.6), with comparable estimates of the magnitude of effect 1.4; 95% CI, 1.1–1.7) than for current smokers (RR = 1.7, observed in case-control studies (RR = 1.4; 95% CI, 1.1– 95% CI 1.5–1.9). - Despite substantial geographic variation in the inci 1.7) and cohort studies (RR = 1.5; 95% CI, 1.3–1.7). In dence of HCC and the distribution of etiologic factors, the 4 studies that adjusted for exposure to the primary smoking was consistently related to increased risk for etiologic agents (Figure 6.20), the magnitude of risk was HCC in all geographic regions, although the magnitude notably higher among ever smokers (RR = 1.7; 95% CI, of the association was not as strong in studies conducted 1.4–2.2) compared to the magnitude of risk among ever in European countries. Among 35 studies conducted in smokers in studies (Figure 6.19). ), the RR for HCC among cur Among 33 case-control studies that evaluated dose- - S Asian countries (Table 6.3 rent or ever smokers was 1.5 (95% CI, 1.4–1.6). response relationships between smoking (e.g., increasing intensity, pack-years, or duration) and HCC, only 6 (18%) In countries in sub-Saharan Africa, the present data reported a statistically significant trend. Among 26 cohort analysis was limited to case-control studies that evaluated ever smoking. The number of cases of HCC in these stud - studies that evaluated these relationships, 10 (38%) reported a significant dose-response effect of smoking ies ranged from 46–240, and all of them adjusted for HBV or HCV infection and consumption of alcohol. Each study intensity on increased risk for HCC, and 2 (8%) reported an inverse dose-response relationship. Many studies that suggested an association between smoking and HCC, but 189 Cancer

219 Surgeon General’s Report Figure 6.17 Estimated risk for liver cancer in current smokers compared with nonsmokers Weights are from random effects analysis. = 675 hepatitis B virus; HBV Note: = effect size; CI = confidence interval; ES HCV = hepatitis C virus. 190 Chapter 6

220 —50 Years of Progress The Health Consequences of Smoking Figure 6.18 Estimated risk for hepatocellular carcinoma in current smokers compared with nonsmokers among studies that controlled for confounding by primary etiological factors (viral hepatitis, consumption of alcohol) HBV Note: Weights are from random effects analysis. CI = confidence interval; ES = effect size; = 683 hepatitis B virus; HCV = hepatitis C virus 684. 191 Cancer

221 Surgeon General’s Report Figure 6.19 Estimated risk for hepatocellular carcinoma in ever smokers compared with never smokers = 690 hepatitis B surface antigen. HBsAg = effect size; ES Weights are from random effects analysis. CI = confidence interval; Note: 192 Chapter 6

222 The Health Consequences of Smoking —50 Years of Progress Figure 6.20 Estimated risk for hepatocellular carcinoma in ever smokers compared with never smokers among studies that controlled for confounding by primary etiological factors (viral hepatitis, consumption of alcohol) Weights are from random effects analysis. Note: CI = effect size. = confidence interval; ES none of them were statistically significant—likely because et al. 1988; Gelatti et al. 2005; Franceschi et al. 2006). In of the limited number of cases. Overall, the RR from the 2 cohort studies from Sweden, the risk estimate in 1 study three studies with data available (Kew et al. 1990; Olubuy- among females was less than 1.0 (RR = 0.7; 95% CI, 0.2– ide and Bamgboye 1990; Soliman et al. 2010) for countries 2.0) (Nordlund et al. 1997). But, the other study observed in Africa was 1.7 (95% CI, 1.1–2.5). cancer among a increased rates of mortality from liver Eight studies evaluated current or ever smoking and cohort of men and a significant dose-response associa - risk for HCC in the United States (Stemhagen et al. 1983; tion with increased smoking (Carstensen et al. 1987). In Austin and Cole 1986; Hsing et al. 1990; McLaughlin et al. a Europe-wide cohort study, Trichopoulos and colleagues Marrero et al. 2005; Zhu et 1995; Hassan et al. 2002, 2009; (2011) rigorously characterized the smoking behavior, al. 2007). Veterans of the armed services were substantially alcohol consumption, diet, and viral hepatitis status of a - half-million people. Overall, the RR for HCC among cur overrepresented in these studies. The overall RR estimate rent smokers compared to never smokers was 4.6 (95% in an analysis that combined current and ever smoking CI, 1.9–10.9), and the RR was notably higher among males was 1.8 (95% CI, 1.3–2.5), and substantial heterogeneity (5.4; 95% CI, 1.7–16.8) than among females (1.7; 95% CI, in estimated risk was not found by study design. 0.3–8.5). In addition, the authors estimated that smok- Among the 14 studies reviewed from countries ing contributed to nearly one-half of the number of cases in Europe, 11 were case-control studies, largely from southern Europe, and 3 were cohort studies. Substantial of HCC, exceeding the proportion of HCC attributable to - HBV, HCV, or consumption of alcohol. Finally, in a quan heterogeneity was observed in these studies. In a series titative analysis for the present review from 5 evaluable - of case-control studies from Greece, smoking was con studies in Europe, the RR for HCC among current or ever sistently associated with HCC, but the associations were smokers (La Vecchia et al. 1988; Goritsas et al. 1995; Nord - more pronounced (and statistically significant) among HBV-negative persons (Trichopoulos et al. 1980, 1987b; lund et al. 1997; Farker et al. 2003; Franceschi et al. 2006) - was 1.4 (95% CI, 1.0–2.3). Tzonou et al. 1991; Goritsas et al. 1995). After adjust ing for HBV and HCV infection, a study from Greece by Similar to the experience in Greece, several studies Kuper and colleagues (2000) demonstrated a 1.5- and 1.6- from other regions suggested a higher risk of liver cancer with smoking among HBV-negative persons than among - fold nonsignificant increase in risk of HCC among per those who were HBV positive (Lam et al. 1982; Yu et al. sons smoking fewer than or at least 40 cigarettes per day, 1991a; Chen et al. 2008). Some other studies, however, respectively. Elsewhere, 4 case-control studies from Italy reported null findings (Filippazzo et al. 1985; La Vecchia failed to find any difference in this risk by HBV status (Kew 193 Cancer

223 Surgeon General’s Report Figure 6.21 Estimated risk for hepatocellular carcinoma among persons without evidence for chronic viral hepatitis infection for current or ever smokers compared with never smokers = hepatitis B virus core 13; HBc13 = confidence interval; Notes: Weights are from random effects analysis. CI = effect size; ES HBsAg = hepatitis B surface antigen; HBV = hepatitis B virus; HCV = hepatitis C virus. 194 Chapter 6

224 The Health Consequences of Smoking —50 Years of Progress Figure 6.22 Estimated risk for hepatocellular carcinoma in former smokers compared with never smokers Notes: Weights are from random effects analysis. = hepatitis HCV CI = confidence interval; ES = effect size; HBV = hepatitis B virus; C virus. 195 Cancer

225 Surgeon General’s Report et al. 1985; Mohamed et al. 1992; Evans et al. 2002). And chronic HBV or HCV infection, the risk for HCC from smoking remained significantly increased. yet, according to eight studies published in 2000 or later, Data combined from 26 studies indicated a 40% - smokers with chronic HBV or HCV infection have a sub stantially higher risk for HCC than those who do not have increased risk of HCC from ever smoking (Figure 6.19). chronic hepatitis infection (Mori et al. 2000; Wang et al. Furthermore, the effect of ever smoking on risk of liver 2003; Jee et al. 2004a; Franceschi et al. 2006; Fujita et al. cancer was strengthened in the studies that addressed pri - 2006; Hassan et al. 2008; Jeng et al. 2009; Soliman et al. mary confounding factors. Risk for liver cancer was signif- - icantly increased in former smokers compared with never 2010). Formal evaluations of interactions between smok ing and HBV or HCV infections have been reported infre - - smokers, although risk for former smokers was attenu ated relative to risk for current smokers. While heteroge- quently from these studies. neity was observed in studies that evaluated dose-response Although the present review focuses on HCC, which represents a substantial majority of primary liver cancer, associations, meta-analysis of a limited number of studies a meta-analysis by Wenbin and colleagues (2013) reported with data that could be combined suggested that increased smoking intensity increases the risk for liver cancer. on the association between smoking with gallbladder The finding of increased risk for liver cancer from cancer. In an analysis of data from 1,158 cases across 11 - studies (all but 1 were case-control), smokers had a sig smoking was generally consistent regardless of geography nificantly increased risk for gallbladder cancer (RR = 1.5; or study design. The greatest number of studies originated - from Asia, and quantitative analysis from this region indi 95% CI, 1.1–1.9) compared with nonsmokers. cated a 50% increased risk of liver cancer from smoking. The estimated risk for liver cancer associated with smok - ing increased to 70–80% in studies from Africa and the Evidence Synthesis - United States. Greater heterogeneity was observed in stud ies from Europe than elsewhere. Several hospital-based Overall, a substantial body of evidence documents case-control studies from southern Europe reported null the association between smoking and primary liver can- - or nonsignificant associations and the overall relation cer. The role of the liver as a primary site for metabolism ship between smoking and liver cancer was thus notably of several recognized carcinogens provides strong biologic smaller in Europe. plausibility for a causal association between smoking and Modification of the effect of smoking on risk for liver HCC. In epidemiologic studies from various geographic cancer by viral hepatitis has been suggested, although regions and with different designs, findings demonstrate a formal statistical evaluation remains limited. Stronger consistent but nonuniform association between smoking associations between smoking and HCC among persons and primary liver cancer. In 2004, IARC classified smoking who are negative for HBV infection have been observed in - as a cause of HCC. In the meta-analysis by Lee and col studies conducted on selected populations in Europe and leagues (2009), which updated the evidence considered in China. In contrast, most studies from diverse regions— the 2004 IARC report, the overall OR showed a moderate such as Asia, Egypt, Europe, and the United States—have association, with an estimated 50% increased risk of liver found greater risks for liver cancer from smoking among cancer associated with current smoking. persons with chronic HBV or HCV infections. In the expanded meta-analysis included in this report, 113 studies were identified that reported data on the risk of liver cancer from smoking. In the primary analysis, which focused on studies of HCC that compared Conclusion current and never smokers, the overall estimate from 31 - 1. - The evidence is sufficient to infer a causal relation studies with evaluable data indicated that current smok ship between smoking and hepatocellular carcinoma. ing increases risk for HCC by approximately 70% (Figure 6.17). Although confounding by consumption of alcohol and HBV or HCV infection status may bias the findings of some studies, controlling for these risk factors does Implications not fully account for the effects seen. In 11 higher quality studies that adjusted adequately for potential confounding The burden of liver cancer is increasing in many factors, risk of HCC from smoking was moderated only regions of the world, notably due to HCV-related cases of slightly (60% increased risk) (Figure 6.18). Importantly, HCC occurring in more developed countries. Among such when analyses of data were restricted to persons without persons, smoking also increases risk and consequently 196 Chapter 6

226 —50 Years of Progress The Health Consequences of Smoking expected to decline. However, if smoking increases in - incidence and death rates related to liver cancer may con - these low- and middle-income countries, then the poten - tinue to grow substantially in the more developed coun tries with rising HCC. In high-burden regions of the world - tial for reducing liver cancer from these preventive inter where vaccination against HBV or reductions in exposure ventions will not be fully realized. to aflatoxin are being achieved, rates of liver cancer are Colorectal Cancer Colorectal cancer—that is, cancer of the colon or - tory drugs (NSAIDs). Aspirin use of 10–20 years is associ ated with a decreased risk of colorectal cancer mortality rectum—is the third most common type of cancer in the (Flossmann and Rothwell 2007), and short-term or cur United States and also ranks third as a cause of cancer - - rent use of hormone replacement therapy (HRT) reduces deaths among men and women in the United States (Sie risk in women (Rossouw et al. 2002). In addition, higher gel et al. 2013). For 2013, the ACS projected 102,480 new - cases of cancer of the colon and 40,340 new cases of can - levels of vitamin D may protect against adenomatous pol cer of the rectum as well as 51,710 deaths from the two yps and incidence, recurrence, and death from colorectal - cancers combined (Siegel et al. 2013). In the mid-1990s, cancer (Ng et al. 2009; Giovannucci 2010). Calcium sup plementation reduces the risk of recurrent polyps (Baron the lifetime probability of developing colorectal cancer et al. 1999). was estimated to be 5.6% in the United States (Howlader et al. 2013). The hypothesis that prolonged cigarette smoking Worldwide, incidence and death rates for colorec- may increase the risk of colorectal cancer gained support tal cancer vary more than 10-fold among countries. The in the mid-1990s when epidemiologic studies, particularly highest rates occur in Australia/New Zealand, Japan, cohort studies, showed a high incidence of adenomatous polyps and/or colorectal cancer in long-term smokers North America, and Western Europe, and the lowest rates - are seen in countries with developing economies, particu (Giovannucci et al. 1994a,b). Initially, there was concern that this observed association reflected uncontrolled larly in Africa and Asia (Parkin et al. 1999). Studies show confounding factors, such as lifestyle characteristics, as that among immigrants moving from low- to high-inci- well as differences in risk between colon and rectal can- dence countries, rates increase within one generation to cer, which are often combined in epidemiologic studies. the approximate rates of the new country, suggesting a - Subsequent studies suggested a stronger relationship strong role for environmental agents (Thomas and Kara - between smoking and rectal cancer than between smok gas 1987). Risk also varies substantially even within coun- ing and colon cancer (Terry et al. 2002b; Wei et al. 2004). tries. For example, in a study by Wei and colleagues (2009) of a middle-aged cohort of U.S. women, risk to age 70 var - This difference was confirmed in two meta-analyses that were limited to prospective cohort studies (Liang et al. ied up to 10-fold based on lifestyle factors. 2009; Tsoi et al. 2009) and one that included both case- An increased risk of colorectal cancer has been linked to a variety of risk factors, including physical inac - control and cohort study data (Botteri et al. 2008a). In the latter systematic review, Botteri and colleagues searched tivity (Wolin et al. 2009); obesity (Renehan et al. 2008); low calcium levels (Cho et al. 2004); and alcohol intake (Thun the literature through May 2008 and evaluated data from et al. 1997). Risk for colorectal cancer also increases for six studies that compared the association of smoking and persons with a family history of colorectal cancer or pol- colon cancer separately from smoking and rectal cancer yps (Fuchs et al. 1994). Finally, a high-meat diet and a diet mortality. The RRs of ever smokers and current smokers low in vegetables, fruits, or folate (World Cancer Research were significantly higher for rectal cancer mortality than for colon cancer (rectal cancer: ever vs. never smoker, Fund/American Institute for Cancer Research 2007) have RR = 1.4 [1.2–1.7], current vs. never smoker, RR 1.6 = been implicated. [1.3–1.8], colon cancer: ever vs. never smoker, RR = 1.2 - Conversely, several factors are consistently associ [1.0–1.4], current vs. never smoker, RR = 1.2 [1.1–1.3]) ated with a reduced risk of colorectal cancer, including - (Botteri et al. 2008a). the use of aspirin and other nonsteroidal anti-inflamma 197 Cancer

227 Surgeon General’s Report diagnosed in current smokers. Approximately 85% of Conclusions from Previous Surgeon p53 colorectal cancers show inactivating mutations of the General’s Reports tumor suppressor gene on chromosome 17p, resulting in loss of the ability to arrest cell growth and/or produce Until the 2001 Surgeon General’s report on women apoptosis; these mutations are important at a late stage in and smoking (USDHHS 2001), the reports of the Surgeon malignant transformation (Hollstein et al. 1991). Clonal General on smoking had not considered the relationship expansion of colorectal tumors containing mutant p53 of smoking with cancers of the colon and rectum. The genes gains a selective survival advantage for these tumors 2001 Surgeon General’s report concluded that “Women and they become increasingly invasive and metastatic. who smoke may have increased risk for ... colorectal Cigarette smoke contains many carcinogens, PAHs, cancer” (p. 231). IARC reported in 2004 that “There is heterocyclic aromatic amines, and N -nitrosamines (Hoff- some evidence from prospective cohort studies and case- mann and Hoffmann 1997) that can reach the large bowel control studies that the risk of colorectal cancer is via the circulatory system (Giovannucci and Martinez increased among tobacco smokers,” but noted that “Inad- 1996). One study documented that DNA adducts to metab- equate adjustment for various potential confounders could olites of B[ a ]P, a potent PAH, in colonic mucosa occur account for some of the small increase in risk that appears more frequently and at higher concentrations in smokers to be associated with smoking” (p. 1183). The 2004 Sur - than in nonsmokers (Alexandrov et al. 1996); this study geon General’s report, after reviewing extensive evidence, provides direct evidence that tobacco carcinogens bind concluded that the evidence is suggestive but not suffi - to DNA in the human colonic epithelium. Moreover, DNA cient to infer a causal relationship between smoking and adduction levels in the colonic epithelium were found in colorectal adenomatous polyps and colorectal cancer. one study to be higher in tumor tissue from persons with colorectal cancer than from control subjects (Pfohl-Lesz - kowicz et al. 1995). Biologic Basis Other genes known to be important in colorectal cancer include mismatch repair genes associated with Most cancers of the colon and rectum are adeno- - the hereditary familial syndrome, nonpolyposis colorec carcinomas. These tumors typically develop from clonal tal cancer, or sporadic cases of colorectal cancer (Liu et expansions of mutated cells through a series of histo- al. 1995; Thibodeau et al. 1998). One study associated pathologic stages—from single crypt lesions to benign cigarette smoking with a mismatch repair deficiency tumors (adenomatous polyps) to metastatic carcinomas— in colorectal cancer, as reflected by a sixfold increase in that take place over a span of 20–40 years (Fearon and the risk of microsatellite instability in tumors in current Vogelstein 1990). The number and order of genetic and smokers compared with nonsmokers (Yang et al. 2000). epigenetic changes in tumor suppressor genes (such as Elsewhere, in a large case-control study of incident colon RAS ) deter - P53 , and DCC , APC ) and oncogenes (such as - cancer, Curtin and colleagues (2009b) evaluated base exci mine the probability of tumor progression (Fearon and sion repair and observed a twofold increase in the risk of Vogelstein 1990). On the basis of the observation that tumor mutations in current and former smokers. More mutations of the APC gene on chromosome 5q are found generally, research continues to provide insight into path - as frequently in small adenomatous polyps as in cancers, ways by which smoking could increase risk for colorectal the loss of normal APC function is considered an early cancer (Campbell et al. 2009). (and possibly initiating) event in colorectal tumorigenesis To date, the association between cigarette smoking (Powell et al. 1992; Morin et al. 1997). Products of the APC and colorectal cancer has not been found to be modified by gene influence cell proliferation, adhesion, migration, and - polymorphisms of genes that are important in the detoxi apoptosis. Activating mutations in codons 12 and 13 of the fication of carcinogens found in tobacco smoke, includ - RAS oncogene are important in the progression of ade - GSTM1 ing , GSTT1 , and (Gertig et al. 1998; Slattery NAT2 nomas but are not directly involved in malignant trans - et al. 1998). Studies of colorectal adenomas have found formations in the bowel (Bos 1989; Ohnishi et al. 1997). - no modification of the risk of cigarette smoking by poly KRAS does have a role in advanced colorectal However, , morphisms of GSTM1 NAT2 , or cytochrome P4501A1, cancer (Fearon 2011). In addition, some studies suggest an enzyme important in the activation of PAHs (Lin et mutations KRAS that smokers develop adenomas without al. 1995; Potter et al. 1999). However, when researchers (Wark et al. 2006). Slattery and colleagues (2000) related examined only adenomas that were 1 centimeter (cm) smoking to microsatellite instability (a genetic marker) in geno- or larger, current smokers with the GSTM1 null colon tumors, and Curtin and colleagues (2009a) showed null type were at a higher risk than those without the microsatellite instability in rectal tumors that were genotype (Lin et al. 1995). Furthermore, some evidence Chapter 6 198

228 —50 Years of Progress The Health Consequences of Smoking the same prospective cohort, then the results from the suggests an increased risk of colorectal cancer and genotype advanced polyps in smokers with GST1 null longest follow-up are used unless otherwise stated. (Ates et al. 2005). Overall, a meta-analysis of 12 studies did not show any that evaluated polymorphisms in GSTM1 significant interaction with smoking and risk (Raimondi Epidemiologic Evidence et al. 2009). Combined data from 7 of the 12 studies indi - medium-metab- cated that smokers with mEH3 low- or Adenomatous Polyps genotypes had a slightly lower risk of colorectal olizer Botteri and colleagues (2008b) used rigorous search high-metabolizer mEH3 adenoma than smokers with and data extraction techniques to synthesize the evi - genotypes. None of the other common genetic polymor - dence for an association between smoking and the risk phisms involved in metabolizing tobacco carcinogens of adenomatous polyps. Among articles published from modified the risk of colorectal adenoma or cancer. 1988–2007, they evaluated 125 in detail; these studies Animal models of the carcinogenicity of tobacco were conducted in countries around the world. Combined in the colon and rectum have been limited to date and data from 33 studies found that current smokers had a have not included studies in which the route of exposure significantly increased risk of adenomas (RR = 2.14; 95% was inhalation. In inbred male Syrian hamsters, adeno - CI, 1.86–2.46) (Figure 6.23). Among current smokers, the carcinomas of the colon have been produced by intrar - pooled RR estimates were somewhat greater (RR = 2.02; ]P (Wang et al. 1985), and in vivo a ectal instillation of B[ 95% CI, 0.62–6.56) for larger adenomas (≥10 millimeters mutational assay studies found that oral administration of [mm]) and those classified as high risk (RR = 2.04; 95% B[ Mouse) induced a ]P to the lacZ transgenic mouse (Muta CI, 1.56–2.66). In addition, in a comparison with never a higher frequency of mutation in the colon than in the smokers in 27 studies, former smokers had a significantly other organs tested (Autrup et al. 1978; Hakura et al. increased risk of adenomas (RR = 1.47; 95% CI, 1.29–1.67) 1998, 1999; Kosinska et al. 1999). Finally, in vitro studies (Figure 6.24). Finally, for every additional 10 pack-years have shown that both rat and human colonic epithelium of smoking, ever smokers had a 13% increase in risk of a ]P (Autrup in cell cultures can enzymatically activate B[ adenomatous polyps (95% CI, 9–18%). An evaluation for et al. 1978). publication bias by Botteri and colleagues (2008b) showed no indication of such bias for the reporting of results about current smokers, but there was evidence for reports Description of the Literature related to former and ever smokers. Review Colon and Rectal Cancer The published studies on cigarette smoking and Table 2.27 of the 2004 Surgeon General’s report pre- colorectal adenomatous polyps and cancer cited in this sec - sented data from cohort studies of incidence and mortality tion were identified by updating through December 2009 for colon and rectal cancer among men and women in the the search of the MEDLINE database from 1966 through United States (USDHHS 2004). Data published through July 2000 that was used in the 2004 Surgeon General’s 2000 and summarized in the 2004 Surgeon General’s report. The headings “tobacco,” “smoking,” “colorectal report consistently indicated that current smokers had adenomas,” “colorectal neoplasms,” “colonic neoplasms,” an increased risk of colon cancer (the RRs ranged from and “rectal neoplasms” were used in the newer search. In 1.2–1.4) and of rectal cancer (RRs ranged from 1.4–2.0), addition, this more recent search included examination of regardless of the number or types of covariates for which the Web of Science and Embase, also through December there was adjustment. 2009. Since the 1960s, the association between cigarette S summarizes the 19 prospective cohort Table 6.5 smoking and colorectal adenomas and cancer has been studies on smoking and the incidence of colorectal evaluated in many prospective and case-control stud- cancer that were published from 2002–2009. In the first ies; the present review extends work summarized in the study listed, Terry and colleagues (2002b) followed 89,835 2004 Surgeon General’s report and focuses on published Canadian women for a mean of 10.6 years and confirmed studies that excluded cigar and pipe smokers, identified 363 cases of colon cancer and 164 of rectal cancer. The lifetime nonsmokers, and distinguished current smokers RR for rectal cancer for women with a smoking duration from former smokers. If multiple reports resulted from 2 ; for women of 30–39 years was 1.52 (95% CI, 1.01−1.26) 2 The RR does not fall within the CI. The information presented here appears just as it does on page 481 of Terry and colleagues ( 2002 b). 199 Cancer

229 Surgeon General’s Report Figure 6.23 Forest plot of relative risk for colorectal adenoma for current smokers versus never smokers Source: Adapted from Botteri et al. 2008b, with permission from Elsevier, © 2008. Note: Partial endoscopy group is composed of studies in which some or all controls underwent partial colon examination. Full = confidence interval. colonoscopy group is composed of studies in which all controls underwent complete colon examination. CI a Estimates for males only. b Estimates for distal colon. c Estimates for proximal colon. d Estimates for rectum. e Estimates for women only. 200 Chapter 6

230 The Health Consequences of Smoking —50 Years of Progress Figure 6.24 Forest plot of relative risk for adenomatous polyps for former smokers versus never smokers Source: Adapted from Botteri et al. 2008b, with permission from Elsevier, © 2008. Note: Partial endoscopy group is composed of studies in which some or all controls underwent partial colon examination. Full colonoscopy group is composed of studies in which all controls underwent complete colon examination. CI = confidence interval. a Estimates for distal colon. b Estimates for proximal colon. c Estimates for rectum. d Estimates for males only. 201 Cancer

231 Surgeon General’s Report 453 cases of rectal cancer. The risk of colon cancer was with duration of 40 or more years, the RR was 2.27 (95% elevated among former smokers but not current smokers, CI, 1.06–4.87). Tiemersma and colleagues (2002) followed 36,000 while there were no significant findings for rectal cancer. - In the United States, Berndt and colleagues (2006) fol Dutch men and women who were 20–59 years of age at - enrollment. At the end of follow-up (8.5 years), the inves - lowed 22,887 participants in the Campaign Against Can tigators confirmed 102 cases of colorectal cancer. The cer and Heart Disease (CLUE II) cohort from Washington County, Maryland, and confirmed 250 cases of colorectal relationship between smoking and risk for colorectal cancer. Compared with never smokers, ever smokers in cancer was null among current smokers but significant among two groups of former smokers (durations of 16−30 the CLUE II cohort had an increased risk of colorectal can - cer that failed to reach statistical significance (RR = 1.23; and >30 years). In a U.S.-based study, Limburg and col- leagues (2003) followed 34,467 women who were 55–69 95% CI, 0.91–1.66). This analysis adjusted for age and years of age at baseline. The study confirmed 869 cases of gender but not for other risk factors for colorectal cancer. colorectal cancer; duration of smoking was significantly In Korea, Kim and colleagues (2006), who followed a related to risk of colorectal cancer incidence. cohort of 14,103 men and women, confirmed 100 cases of Per Table 6.5 S - colorectal cancer. These investigators found that duration , Otani and colleagues (2003) fol of smoking was significantly related to risk of colorectal lowed 90,004 Japanese men and women who were 40–69 years of age at enrollment. When the analysis was limited cancer: for those who had smoked more than 45 years, the RR was 2.35 (95% CI, 1.16–4.74) in a comparison with to invasive cases, there was a significant increase in risk never smokers. Also in Asia, Akhter and colleagues (2007) among current smokers (RR = 1.6; 95% CI, 1.1−2.1) that followed a cohort of 25,279 Japanese men (40−64 years of was comparable to results when the analysis included all cases of invasive and noninvasive colon and rectal cancers. age at baseline) for a mean of 7 years and identified 188 cases of colorectal cancer. These researchers observed a In Japan, Shimizu and colleagues (2003), who fol - significant increase in risk among former smokers and a lowed 29,051 men and women for 8 years, confirmed 181 statistically insignificant, modestly increased risk among cases of colon cancer and 95 of rectal cancer. Among men, no trend was revealed between the risk of colon cancer current smokers. Both age at initiation and duration of smoking were related to risk. In the United States, Paskett and lifetime smoking (in pack-years), but for rectal can - - and colleagues (2007) analyzed data from 146,877 partici cer, the risk was significantly greater with more than 20 pack-years (RR = 2.44; 95% CI, 1.12–5.30) than it was for pants in the Women’s Health Initiative (WHI). After nearly nonsmokers. In The Netherlands, a study by van der Hel 8 years of follow-up, the study confirmed 1,075 cases of and colleagues (2003a), which followed a cohort of 27,222 colon cancer and 176 cases of rectal cancer. The study did not find a significant relationship between smoking and women, identified 249 cases of colorectal cancer. Ever smoking was similarly related (but not significantly) to risk of colon cancer, but current smokers had a signifi - colon cancer (RR = 1.36; 95% CI, 0.97–1.92) and to rectal cantly elevated risk of rectal cancer (RR = 1.95; 95% CI, 1.10–3.47). Duration of smoking was associated with risk cancer (RR = 1.31; 95% CI, 0.76–2.25). Wakai and colleagues (2003), who followed a Japa- of colon cancer (p-trend = 0.03) and rectal cancer (p-trend nese cohort of 25,260 men and 34,619 women for an = 0.05). average of 7.6 years, confirmed 408 cases of colon can Among a cohort of Chinese men and women in - cer and 204 cases of rectal cancer. Among both men and Singapore, Tsong and colleagues (2007) confirmed 516 cases of colon cancer and 329 cases of rectal cancer dur - women, there was no relationship between years of smok- ing a mean follow-up of 11 years. In this cohort, both ing and risk of colon cancer or rectal cancer. In the United Kingdom, the Oxford Vegetarian Study, which followed current and former smoking were related to risk of rectal a cohort of 11,140 vegetarians (Sanjoaquin et al. 2004), cancer but not to risk of colon cancer. Similarly, age at confirmed 95 cases of colorectal cancer and found that risk initiation and duration of smoking were related to risk of rectal cancer but not to risk of colon cancer. In the United was elevated among both former and current smokers. In States, a study by Driver and colleagues (2007) reported - Europe, The Netherlands Cohort Study on Diet and Can cer followed 58,279 men and 62,573 women (Lüchtenborg on follow-up results for male physicians in the Physicians’ Health Study; after 20 years of follow-up, there were 381 - et al. 2005); during the last 5.0 years of the 7.3-year fol confirmed cases of colon cancer and 104 confirmed cases low-up, the study identified 661 cases of colorectal cancer. of rectal cancer. Overall, ever smoking was related to risk The risk of colorectal cancer was elevated among former smokers (RR = 1.30; 95% CI, 1.03–1.65) but not current of colorectal cancer (RR = 1.42; 95% CI, 1.17–1.72). In smokers. In Asia, Yun and colleagues (2005) followed the addition, current smokers who smoked two packs per day Korean National Health Insurance Corporation cohort of had an increased risk of colon cancer (RR = 1.53; 95% CI, 733,134 men and identified 417 cases of colon cancer and 1.02–2.29) and rectal cancer (RR = 1.92; 95% CI, 1.01– 202 Chapter 6

232 —50 Years of Progress The Health Consequences of Smoking cohort and 20 case-control) that were published from - 3.66). In Maryland, Hooker and colleagues (2008) evalu ated incidence of rectal cancer in two cohorts of residents 1980–2008 and further characterized the association of smoking with colorectal cancer. Drawing on 47 of those from that state’s Washington County. In the cohort that was followed from 1963 to 1978, there was a significant studies, the authors found that former smokers had an increased risk of colorectal cancer (RR = 1.17; 95% CI, increase in risk of rectal cancer among current male 1.11–1.22) in comparison with never smokers. In addition, smokers but not among their female counterparts. The RR based on 25 of the studies, ever smokers had an increased for rectal cancer in the cohort followed from 1975 to 1994 risk of colorectal cancer (RR = 1.18; 95% CI, 1.11–1.25) ranged from 1.57 to 1.92 for current and former smokers, compared with never smokers. This meta-analysis also but only the RR for former female smokers (1.87; 95% CI, evaluated risk for colorectal cancer mortality; based on 1.02−3.45) reached significance. Also in the United States, Hannan and colleagues 14 and 12 studies, respectively, current smokers (RR = 1.28; 95% CI, 1.15–1.42) and former smokers (RR = 1.23; (2009) studied 184,187 men and women as part of the 95% CI, 1.14–1.32) had an increased risk of mortality Nutrition cohort of the CPS-II. After 13 years of follow-up, the study confirmed 1,962 cases of colorectal cancer. Cur - from colorectal cancer in a comparison with never smok- rent smokers had an increased risk of colorectal cancer ers (Botteri et al. 2008a). The increased mortality could (RR = 1.27; 95% CI, 1.06–1.52), as did former smokers reflect a higher incidence of colorectal cancer in smokers (RR = 1.23; 95% CI, 1.11–1.36). Among current smokers, or an unfavorable effect on the disease’s natural history. the RR was greatest for those with a long duration of smok - ing. RR was comparable between men and women. Finally, a study by Gram and colleagues (2009) followed 68,160 Evidence Synthesis women in Norway and confirmed 425 cases of colorectal cancer. Duration of smoking was significantly related to Taken as a whole, the results of the studies summa- overall risk of colorectal cancer, but when individual sites S –6.7 S , which come from millions of rized in Tables 6.5 - were evaluated, sparse data limited the power to find sig person-years of follow-up, confirm the findings of three nificant associations. Increasing pack-years smoked was meta-analyses for colorectal cancer (Botteri et al. 2008a; related to increased risk of colorectal cancer. - Liang et al. 2009; Tsoi et al. 2009). The individual stud summarizes 16 case-control studies pub- S Table 6.6 - ies have addressed cancers of the colon and rectum sep lished from 2001–2008; here the findings are mixed, with arately, as well as the combined outcome of colorectal only a few studies reporting significant increases in risk cancer. Mechanistic understanding at present supports associated with various measures of smoking. The stud- the handling of the combined outcome in synthesizing ies were carried out in diverse locations, including Asia, the evidence. North America, and Europe. Sample sizes ranged up to Although adjustments for covariates differed to 2,000 cases and adjustments were made for a variety of some extent across the studies included in the meta-anal - risk factors. yses, longer duration of smoking was consistently associ - S Table 6.7 presents details on nine cohort studies ated with increased risk of colorectal cancer. In addition, that reported mortality data for either colorectal can - there was no evidence of heterogeneity of effect when the cer overall or separately for colon and rectal cancer. The prospective cohort studies were combined in the three cohort studies of mortality also came from North America, separate meta-analyses (Botteri et al. 2008a; Liang et al. Asia, and Europe. In several studies, risk for death from 2009; Tsoi et al. 2009). colorectal cancer was significantly increased; for example, These epidemiologic data must be placed in the in two studies among women in the United States—the context of our growing understanding of the biologic Nurses’ Health Study (NHS) (Kenfield et al. 2008) and the etiology of colorectal cancers; researchers now have - Iowa Women’s Health Study (Limburg et al. 2003)—cur excellent insights into the sequence of genetic changes rent smokers have an approximate 60% increased risk of taking place from normal cells to a polyp to malignancy. colorectal cancer mortality. Several of these studies sum - The evidence now points strongly to an effect of smok - also observed significant increases S marized in Table 6.7 ing in increasing the formation of polyps, the precursor in risk based on number of cigarettes smoked per day or of colorectal cancer, and possibly on the development of total pack-years. malignancy (Botteri et al. 2008a,b; Liang et al. 2009; Tsoi et Most of these studies were summarized in the three - al. 2009). Furthermore, for colorectal cancer, recent find separate meta-analyses referenced earlier in this chapter ings from prospective cohort studies suggest that long- (Botteri et al. 2008a; Liang et al. 2009; Tsoi et al. 2009). term cigarette smoking is associated with increased risk of Notably, the meta-analysis by Botteri and colleagues both incidence and mortality in men as well as women. In (2008a) combined data from 53 studies (33 prospective 203 Cancer

233 Surgeon General’s Report statistically significant increase in risk for colon or rectal some studies, the risk of incidence and mortality tended to increase with longer duration of smoking and younger cancer associated with smoking, regardless of the set of covariates for which there was adjustment. Furthermore, age at smoking initiation and to decrease with a younger age at successful cessation and a greater number of years among the prospective cohort studies, many controlled since that took place, but the effects of these factors (age for physical activity, use of alcohol, and other potential at starting or quitting and duration of smoking or time risk factors. since cessation) cannot be readily separated because of Cumulative findings from large prospective cohort - studies show an increased risk of colon and rectal can their inherent correlation. cer after smoking for two or more decades. The evidence The aggregate epidemiologic evidence supports the - suggests that smoking acts in the early stages of carcino hypothesis of Giovannucci and colleagues (1994a,b) and genesis, as shown by its association with adenoma, the ele- of Giovannucci and Martínez (1996) that a latent period vated risk for most smokers, and the associated risk with of several decades is necessary for cigarette smoking to duration of smoking. The temporal pattern of the effects increase either the incidence of colorectal cancer or mor - of smoking, with continuing increase in risk, particularly tality from that disease and that cigarette smoking likely plays a role in early carcinogenesis in both the colon and for rectal cancer and for mortality among current smok - rectum. This combined hypothesis is further supported by ers, suggests that smoking may also act in the later stages - the consistent association between smoking and adeno of carcinogenesis. mas, which represents the starting point for colorectal cancer, with a doubling of risk among current smokers (Botteri et al. 2008b). Studies with null findings but only Conclusion limited follow-up of long-term smokers are not informa- tive for testing the hypothesis that a lengthy duration of 1. The evidence is sufficient to infer a causal relation - - smoking is needed to increase the risk of colorectal can ship between smoking and colorectal adenomatous cer. Analyses of available studies show little indication of polyps and colorectal cancer. publication bias. There is also no indication of significant heterogeneity of effect among study results. In assessing whether cigarette smoking plays a Implications - causal role in colorectal cancer, nutrition and other fac tors such as physical activity and screening histories for The aggregate evidence indicates that cigarette colorectal cancer must be considered because they may - smoking may be a modifiable factor that can cause colorec confound the association. Not all of the studies to date tal cancer. Accordingly, clinicians and public health per - have controlled for risk factors for colorectal cancer that sonnel should include both current and former smoking may also be associated with smoking, such as physical as potential risk factors for this disease. inactivity. However, indirect evidence against confound - ing comes from the consistent finding of a small but Prostate Cancer dence may be due to trends in prostate-specific antigen Among American men, prostate cancer is the most (PSA) testing. commonly diagnosed cancer and the second leading cause To date, several risk factors for prostate cancer have of cancer death. In 2013, 238,590 American men were been identified with certainty; these risk factors cannot be expected to be diagnosed with prostate cancer and 29,720 were expected to die from this disease (Siegel et al. 2013). modified: Since the mid-1990s, death rates for prostate cancer have been declining, but incidence rates have fluctuated (Siegel • Age . The risk of prostate cancer increases with age. et al. 2013). The decline in death rates has been attrib- • Race. Prostate cancer incidence and death rates are uted to the combination of earlier detection and advances in the treatment of men who are in advanced stages of highest among African American men and lowest - among Asian men. the disease (Etzioni et al. 2008); the fluctuation in inci Chapter 6 204

234 —50 Years of Progress The Health Consequences of Smoking • not reduce risk in men who were not selected for exposure . Men who have a father or brother Family history to selenium or smoking status. diagnosed with prostate cancer are twice as likely to be diagnosed with prostate cancer as those with unaffected fathers and brothers. Conclusions from Previous Surgeon Unlike the case in breast and colon cancer, research General’s Reports has not yet identified the inherited mutations in genes that consistently explain the strong family associations The relationship between smoking and risk for found in prostate cancer, but some studies have discovered prostate cancer was first addressed in the 2004 Surgeon a small number of common variants across the genome General’s report on the health consequences of smok - that are associated with the risk for this disease (Eeles et ing. That report drew two conclusions: (1) the evidence al. 2008, 2009; Thomas et al. 2008). is suggestive of no causal relationship between smoking Biologic pathways influencing prostate cancer and risk for prostate cancer; and (2) the evidence for mor - involve hormones and growth factors. Androgens and their tality, although not consistent across all studies, suggests signaling pathways are necessary for the development of a higher mortality rate from prostate cancer in smokers prostate cancer. Support for the role of these pathways is than in nonsmokers (USDHHS 2004, p. 26). based on results of two trials showing that drugs inhibit - α -reductases, the enzymes that convert testosterone ing 5- to the more androgenic dihydrotestosterone, reduce the Biologic Basis risk of prostate cancer (Thompson et al. 2003; Andriole - et al. 2010). In epidemiologic studies, however, circulat Zu and Giovannucci (2009) outlined several pos- ing levels of androgens have not been associated with the sibilities for increased mortality from prostate cancer, risk of prostate cancer (Roddam et al. 2008a). Growth fac- including mutations in genes associated with the can - tors are also important: for example, results from cohort cer’s progression caused by carcinogenic constituents of studies have consistently associated circulating levels of cigarette smoke and the effects of smoking on levels of sex insulin-like growth factor-1 with increased risk for pros - steroid hormones, angiogenesis, and DNA methylation. tate cancer (Roddam et al. 2008b). Research on pathways Regarding carcinogenicity and methylation, for example, may provide insights into etiologic factors. - -transferase pi expression, via hyper S loss of glutathione In terms of modifiable risk factors, obesity is associ - - methylation of its gene promoter region early in the natu ated with an increased risk of death from prostate cancer ral history of prostate cancer (Nakayama et al. 2003) may (Calle et al. 2003), but evidence for an association between render prostate cancer cells susceptible to DNA damage risk for incident prostate cancer and physical inactivity as well as other kinds of damage caused by electrophiles - is not consistent (Friedenreich and Thune 2001). Drink from cigarette smoke (e.g., PAHs) (Roberts et al. 2003). ing alcohol does not appear to be an important factor for In terms of hormones, compared with men who do not prostate cancer incidence or mortality (Velicer et al. 2006; - smoke, men who currently smoke have higher circulat Gong et al. 2009; Chao et al. 2010). Some studies have ing levels of androstenedione—a weak androgen that is a found a higher risk of prostate cancer or advanced disease precursor to testosterone and estradiol—and higher lev - with a higher intake of energy (calories), processed meat, els of total and free testosterone (Dai et al. 1988; Field et - dairy foods, and calcium, as well as lower intake of toma al. 1994; Muller et al. 2003; Shiels et al. 2009). On the toes and cruciferous vegetables (Giovannucci et al. 2007; other hand, former and never smokers have similar levels World Cancer Research Fund 2007). Regarding preven- of total and free testosterone (Shiels et al. 2009). Because tion, two studies found reduced risk of prostate cancer as androgens are necessary for the development of prostate a secondary endpoint. In one study, persons who had skin cancer, this pattern is consistent with the observation in cancer and lived in areas with low levels of selenium in the some epidemiologic studies that current but not former soil received selenium supplements (Clark et al. 1998); in smoking is associated with risk of death from prostate the other study, men who were current or former smok - cancer. As for estradiols, some studies have found that ers received vitamin E (Alpha-Tocopherol 1994). However, men who smoke have higher total and free levels of this in a subsequent trial designed to test the hypothesis that hormone than men who do not smoke (Barrett-Connor supplementation with these agents would reduce the risk and Khaw 1987; Shiels et al. 2009). The role of estrogens of prostate cancer, Lippman and colleagues (2009) found in human prostate carcinogenesis is not clear. that supplementation with selenium or with vitamin E did 205 Cancer

235 Surgeon General’s Report cancer (RR = 1.04; 95% CI, 0.87–1.24) among current Description of the Literature - smokers. The elevated risk was significant in data strati Review fied by amount smoked (cigarettes per day: RR = 1.22; 95% CI, 1.01–1.46; pack-years of smoking: RR = 1.11; To further examine the association between cigarette 95% CI, 1.01–1.22). Increased risk of deaths from prostate smoking and the risk for prostate cancer incidence, case cancer was also found among current smokers (RR = 1.14; fatality (prostate-cancer-specific mortality), and mortality 95% CI, 1.06–1.19) (Huncharek et al. 2010). from all other causes, epidemiologic studies were iden - Twenty-one of the 35 prospective studies reviewed in tified through reviews of the reference lists in the 2004 Table 6.8 did not support a positive association between S Surgeon General’s report on the health consequences of cigarette smoking and risk (incidence) of prostate cancer. smoking; published meta-analyses, expert reviews, and Four of the 35 studies supported positive associations research articles; and through searches of the National (Whittemore et al. 1984; Hiatt et al. 1994; Adami et al. Library of Medicine’s PubMed service for research articles 1996; Cerhan et al. 1997), and 10 produced either null published after the 2004 report. The PubMed search terms associations or findings that appeared to indicate inverse used were “smoking,” “cigarettes,” “tobacco,” “prostate associations. Beyond the studies summarized in Table cancer,” “prostate neoplasms,” “prostatic neoplasia,” and - S 6.8 , a nested case-control study by Heikkilä and col “prostate tumor.” The last PubMed search was performed leagues (1999) did not reveal a baseline difference in the April 15, 2010, for studies dating back to 2000. Case- prevalence of current smoking between incident prostate control studies were not considered because they do not cancer cases and controls. In another study, in a compari - directly address factors determining incidence or provide son with the general population, Malila and colleagues data about mortality. (2006) found a higher than expected incidence rate of - prostate cancer in the placebo arm of the Alpha-Tocoph erol, Beta-Carotene Trial of male smokers (median level of Epidemiologic Evidence smoking at randomization: 20 cigarettes/day for 36 years): the standardized incidence ratio here was 1.20 (95% CI, Incidence and Mortality 1.06–1.35) (Malila et al. 2006). In contrast to the lack of a consistent association More than 30 prospective studies have investigated described above between smoking and incidence of pros - the link between smoking and incidence of prostate can- tate cancer, 12 prospective studies (Hammond and Horn cer or death from that disease; Table 6.8 summarizes the S 1958; Akiba and Hirayama 1990; Hsing et al. 1991; Tverdal findings from studies that reported rates, risks, or RRs - et al. 1993; Adami et al. 1996; Coughlin et al. 1996; Rodri of prostate cancer associated with cigarette smoking. Of guez et al. 1997; Giovannucci et al. 2007; Rohrmann et note, Table 6.8 presents updated findings from 8 studies S al. 2007; Batty et al. 2008; Watters et al. 2009; Weinmann that have examined five cohorts over time (see notes a–f et al. 2010) of the 20 such studies that evaluated prostate in Table 6.8 ). Epidemiologic studies of the association S cancer mortality in Table 6.8 S supported a modest-to- between cigarette smoking and prostate cancer incidence - moderate positive association with smoking. In an inves and mortality have been reviewed previously (Colditz S , a prospective cohort tigation not included in Table 6.8 1996; Lumey 1996; Hickey et al. 2001; Levi and La 2001; study by Eichholzer and colleagues (1999) that used non - Zu and Giovannucci 2009), including an Australian con - - smokers with normal levels of vitamin E as a compari sensus conference report (Colditz 1996). To date, the asso - son group reported a higher risk of prostate cancer death ciation between cigarette smoking and prostate cancer has - among men who smoked and had a low plasma concentra not been found to be modified by polymorphisms of genes - tion of vitamin E (RR = 3.26; 95% CI, 1.27–8.35). In con that are important in the detoxification of carcinogens trast, no difference in risk was found among male smokers GSTM1, GSTT1 found in tobacco smoke, including , and who had a normal level of vitamin E. (Gertig et al. 1998; Slattery et al. 1998). However, NAT2 Unlike associations between smoking and other some studies indicate association of xenobiotic metabo - types of cancer such as neoplasms of the lung, the risk lism gene SNPs with colorectal cancer and smoking (Nisa of prostate cancer death does not appear to rise with an et al. 2010; Koh et al. 2011; Osawa et al. 2012; Fu et al. increasing number of cigarettes smoked per day, dura - Meta-analyses of prospective studies (Huncharek et 2013). tion of smoking, or total pack-years. However, current al. 2010) and case-control studies (Lumey 1996) have also or recent smoking (Figure 6.25), rather than smoking been conducted. In the pooled analysis of data from 24 in the distant past or a cumulative smoking history, may cohort studies, Huncharek and colleagues (2010) reported influence prostate cancer mortality. For example, among some evidence of increased risk for incident prostate Chapter 6 206

236 The Health Consequences of Smoking —50 Years of Progress Prospective cohort studies on the association between current cigarette smoking and prostate cancer Figure 6.25 mortality Includes studies reporting a relative risk and 95% confidence interval for current smoking or current number of cigarettes Note: S for additional studies for which confidence intervals were not reported. smoked per day. See Table 6.8 CI = confidence interval. studies in Table 6.8 from 1986 to 1994 and noted 177 prostate cancer deaths in S reporting a positive association for smoking, the RR was larger for current smokers than 351,261 person-years. Compared with never smokers, the RR was 1.58 for current smokers at baseline, 1.73 for men for former smokers (Hsing et al. 1991; Adami et al. 1996; who had quit smoking within 10 years of baseline, and 1.04 Rodriguez et al. 1997; Giovannucci et al. 2007; Batty et for those who had quit 10 or more years before baseline. al. 2008; Watters et al. 2009; Weinmann et al. 2010) or was stronger when considering smoking status closer to In a later report from the same study, Giovannucci and the time of death from prostate cancer (Hsing et al. 1991; associates (1999) followed participants from 1986–2002 and noted 312 prostate cancer deaths in 673,706 person- Rohrmann et al. 2007). years. Using simple updating of biennially assessed smok - Two reports from Giovannucci and colleagues (1999, 2007) provide further evidence for the importance of rela ing status (rather than baseline smoking status, as in their - - 1999 report), the authors found that the RR among cur tively recent smoking. In an earlier report from the Health ), S Professionals Follow-up Study (not shown in Table 6.8 rent smokers, in a comparison with smokers who had quit within 10 years, was 1.41 (95% CI, 1.04–1.91). Giovannucci and coworkers (1999) followed participants 207 Cancer

237 Surgeon General’s Report pathologist or poorly differentiated cancer at pathologic Data from some studies do not support the hypothe - examination) are indicators of a poor prognosis. Thus, sis that the association between prostate cancer mortality studies about smoking and stage or grade of the cancer are is stronger for current smoking than for former smoking (Doll et al. 2005). Their British Doctors Study, which fol relevant for interpreting the findings of higher mortality - in the prospective studies. Cases were ascertained from a lowed physicians from 1951–2001, noted 878 prostate can - clinical setting in three studies (Hussain et al. 1992; Dani- cer deaths in 34,439 male physicians. The study recorded ell 1995; Roberts et al. 2003), from a regional cancer reg updated smoking status in 1957, 1966, 1971, 1978, and - 1991. The prostate cancer mortality rate (indirectly stan istry in one (Kobrinsky et al. 2003), and from the SEARCH - dardized for age and study year) did not differ (Table 6.8 ) cohort in the fifth (Moreira et al. 2010). All five studies S support the hypothesis that smokers diagnosed with pros- between never smokers (89.4/100,000 men per year), for - tate cancer are more likely to have advanced-stage dis - mer smokers (80.9), and current smokers (90.0). Despite ease or less-well-differentiated disease than men who have the overall lack of association among smokers, however, prostate cancer and do not smoke. In the only study that the prostate cancer mortality rate (per 100,000 men per - evaluated intensity of smoking, risk of extraprostatic dis year) increased with the number of cigarettes smoked per day by current smokers (1–14/day = 66.7; 15–24 = 99.6; ease or high-grade disease increased with number of pack- years of smoking (Roberts et al. 2003). ≥25 = 113.3), but the p for trend was not significant (0.52) S ). (Table 6.8 Progression, Case Fatality, and All-Cause were not cited in Ten of the studies in Table 6.8 S Mortality the 2004 Surgeon General’s report (Lotufo et al. 2000; Lund Nilsen et al. 2000; Putnam et al. 2000; Allen et al. Nine studies have investigated the association 2004; Doll et al. 2005; Giovannucci et al. 2007; Rohrmann between smoking and the progression of prostate cancer et al. 2007; Batty et al. 2008; Watters et al. 2009; after diagnosis, death from the disease, or death from all Weinmann et al. 2010). Of these, 7 reported on cigarette ). causes in men who have prostate cancer (Table 6.10 S smoking and prostate cancer mortality (Lotufo et al. 2000; Eight of the studies used a retrospective cohort design, Doll et al. 2005; Giovannucci et al. 2007; Rohrmann et al. while one (Gong et al. 2008) was a prospective study. Five 2007; Batty et al. 2008; Watters et al. 2009; Weinmann et studies reported on progression, defined as biochemical al. 2010); 4 of the 7 gave quantitative support for a posi- recurrence/progression/failure, local recurrence/failure, tive association between smoking (3 implicated current distant failure, or development of hormone-refractory smoking) and death from prostate cancer (Rohrmann et disease (Merrick et al. 2004; Oefelein and Resnick 2004; al. 2007; Batty et al. 2008; Watters et al. 2009; Weinmann Pickles et al. 2004; Pantarotto et al. 2007; Moreira et al. et al. 2010). Two of the 10 studies not cited in the 2004 2010). Five studies reported on case fatality (Daniell 1995; (Doll et S Surgeon General’s report but shown in Table 6.8 Pickles et al. 2004; Jager et al. 2007; Pantarotto et al. 2007; al. 2005; Giovannucci et al. 2007) were updates of studies Gong et al. 2008), and five reported on all-cause mortality included in the 2004 report (Doll et al. 1994; Giovannucci (Yu et al. 1997; Oefelein and Resnick 2004; Pickles et al. et al. 1999). The findings in the 2004 report of no associa- 2004; Jager et al. 2007; Pantarotto et al. 2007). One study tion with prostate cancer mortality in the British Doctors than prostate can reported on death from all causes other - Study (Doll et al. 1994) and of a positive association in the cer (Gong et al. 2008). Health Professionals Study (Giovannucci et al. 1999) were Of the nine studies reported in Table 6.10 S , six sug - unchanged with additional follow-up. gest that in men who have prostate cancer, smoking is associated with a higher risk of progression or death from Stage and Histologic Grade the disease; these findings were independent of smoking’s , three studies (Hussain et al. possible influence on stage or grade. Among men diag As shown in Table 6.9 S - nosed with prostate cancer, all-cause mortality appears 1992; Roberts et al. 2003; Moreira et al. 2010) investigated to be higher in smokers than in nonsmokers. In some the association between smoking and both disease stage and histologic grade at the time of diagnosis or surgical studies, many of these deaths were due to prostate can - cer because the majority of men had advanced-stage dis - treatment, while two (Daniell 1995; Kobrinsky et al. 2003) ease (Oefelein and Resnick 2004). In other studies, deaths looked at smoking and disease stage but not histologic were more likely due to other causes because the major - grade. Advanced stage (e.g., local invasion, metastasis to a ity of men had localized disease (Pickles et al. 2004; Gong regional lymph node, metastasis to bone) and high grade (e.g., a high sum of the two Gleason scores given by the et al. 2008). 208 Chapter 6

238 The Health Consequences of Smoking —50 Years of Progress 2. The evidence is suggestive of a higher risk of death Evidence Synthesis from prostate cancer in smokers than in nonsmokers. The published literature suggests that smoking, 3. In men who have prostate cancer, the evidence is sug - especially current or recent smoking, is a risk factor for gestive of a higher risk of advanced-stage disease and prostate cancer mortality but not for incidence of the less-well-differentiated cancer in smokers than in disease. Findings of a positive association with prostate nonsmokers, and—independent of stage and histo - cancer mortality and null associations with incidence are logic grade—a higher risk of disease progression. somewhat consistent across a set of prospective cohort studies (in which temporality is clear) that have been con - ducted in a number of settings and across several decades. The strength of the association between current smoking Implications and prostate cancer mortality is modest to moderate, and unlike the case with some other cancers, the strength of The biologic processes underlying the sugges - the association does not appear to depend on the number tive association between cigarette smoking and prostate of cigarettes smoked per day or pack-years of smoking. cancer mortality, case fatality, and more seriously unfa - The published literature also consistently shows vorable pathologic characteristics of the tumor require that in men who have prostate cancer, smoking is a risk further investigation, particularly because incidence is not factor for being diagnosed with disease that is already of - associated with smoking. Further research on the associa advanced stage or of high grade, and—independent of tion between smoking and the incidence of prostate cancer stage and grade—is a risk factor for progression of the is warranted because the mortality rate indicates an effect disease, including progression to death. Although these of public health significance. Additional epidemiologic - patterns of association are biologically plausible, the spe studies should address the timing of cigarette smoking cific biologic basis is unknown at this point. Alternative relative to mortality and case fatality, and laboratory-based explanations to a causal association cannot be completely - studies should address the biologic mechanisms underly excluded with confidence (Zu and Giovannucci 2009). ing the apparently worse phenotype of prostate cancer in smokers. The finding that the risk of prostate cancer mor - tality is not elevated in former smokers who quit years in the past suggests that quitting smoking may reduce pros- Conclusions tate cancer mortality. Further research is needed to refine this temporal relationship and to quantify the benefits of The evidence is suggestive of no causal relationship 1. cessation after a diagnosis of prostate cancer. between smoking and the risk of incident prostate cancer. Breast Cancer 2013). Average annual incidence rates per 100,000 women Breast cancer is the most frequently diagnosed varied substantially by race/ethnicity in 2004–2008: 77.9 type of cancer, other than nonmelanoma skin cancers, for American Indians/Alaska Natives, 92.1 for Hispanics, and the second leading cause of cancer death among 93.7 for Asians/Native Hawaiian or Other Pacific Island - women (Siegel et al. 2013). Despite an approximate 2% decrease in incidence since 1999 and a 28% decline in ers, 119.9 for Blacks, and 127.3 for non-Hispanic Whites. - breast cancer mortality since 1991 (Jemal et al. 2010a,b), Death rates per 100,000 women also varied by race/ethnic about 211,000 new cases of invasive breast cancer were ity during this period: 12.2 for Asians/Native Hawaiian or diagnosed and approximately 40,000 deaths resulted from Other Pacific Islanders, 15.1 for Hispanics, 17.2 for Ameri - breast cancer among U.S. women in 2009 (Howlader et can Indians/Alaska Natives, 22.8 for non-Hispanic Whites, al. 2013). The age adjusted incidence and death rates for and 32.0 for Blacks. 2004–2008 were 124/100,000 and 23.5/100,000 women The burden of breast cancer morbidity and mor - per year, respectively, based on the 17 geographic areas tality is high. Thus, researchers have long sought to identify modifiable etiologic factors to prevent and covered by the SEER Program of NCI (Howlader et al. 209 Cancer

239 Surgeon General’s Report cancer-causing mutations. For hereditary cancers, as control this disease. Active cigarette smoking and expo- proposed in the Knudson (1996) model, at least two allelic sure to secondhand smoke have received increasing mutations are necessary, one of which might be inherited. attention over the past two decades, as clinical stud- Endogenous and exogenous exposures can potentially ies have detected nicotine and its metabolite cotinine in affect the development and proliferation of mutant cells the breast fluid of nonlactating women (Petrakis et al. in both inherited and sporadic breast cancer and thereby 1978; Hill and Wynder 1979), and data from rodent affect breast carcinogenesis. studies have indicated that genotoxic carcinogens in The following section addresses biologic mecha- cigarette smoke can induce mammary tumors (el-Bay - nisms by which tobacco smoke, an exogenous exposure, oumy 1992). Sixty-nine known carcinogens are detect - able among the myriad chemicals in tobacco smoke - can potentially contribute to the causation of breast can (USDHHS 2004). Adipose tissue of the breast can store cer. The review in this section addresses the plausibility lipophilic carcinogens, and these can be locally activated of a causal association between risk of breast cancer and by breast epithelial cells to form DNA adducts (Phillips et active or passive smoking. The studies were identified al. 2002). The prevalence of carcinogen DNA adducts is through literature searches using the following key words: smoking and breast cancer, carcinogenesis, DNA adducts, reported to be increased in smokers and in women with epigenetic, hormones (androgens, progesterones, and breast cancer (see “DNA Adducts”). A recent report sug - estrogens), anti-estrogen hypothesis, and ovarian func - gests that nicotine leads to overexpression of cyclin D3 tion. Past Surgeon General’s reports were also reviewed: and induces neoplastic transformation and proliferation of those published in 2004 and 2006, which addressed - breast epithelial cells in vitro (Lee et al. 2010a). Thus, evi dence is accumulating for several plausible mechanisms active and passive smoking, respectively (USDHHS 2004, by which smoking may induce breast cancer; this evidence 2006), and the one in 2010, which focused on mecha- is reviewed in greater detail below. nisms by which tobacco smoke contributes to disease (USDHHS 2010). Historically, the epidemiologic evidence for an asso - ciation between breast cancer and active cigarette smok- ing and between breast cancer and exposure to secondhand DNA Adducts smoke has been inconsistent, leading to conclusions in the Cigarette smoke contains thousands of compounds past that smoking is not a risk factor for this type of can - including 69 known to be carcinogens (USDHHS 2010). cer (Palmer and Rosenberg 1993; Terry and Rohan 2002). Some of these compounds have been shown to cause However, some recent reviews have concluded that both mammary tumors in rodents (Hecht 2002). Nicotine, one active and passive smoking may increase the risk of breast - of the major constituents of tobacco smoke, has been mea cancer, although there is continuing disagreement as to sured in the nipple aspirate of female smokers (Petrakis the magnitude of effect (California Environmental Protec- et al. 1978) and smoking-related DNA adducts have been tion Agency [Cal/EPA] 2005; Collishaw et al. 2009 for the found in the DNA of epithelial cells within breast milk - Canadian Expert Panel on Tobacco Smoke and Breast Can (Thompson et al. 2002), documenting that components of cer Risk; Institute of Medicine 2012). smoke reach breast tissue. Carcinogens in tobacco smoke cause cancer by damaging DNA; this is the initiating event in tumorigenesis (Figure 6.4). Many carcinogens from Biologic Basis—Evidence for tobacco smoke are metabolically activated by the cyto - chrome P-450 (CYP) enzymes, including CYP1A1 and Potential Etiologic Mechanisms all of which are present in breast tis - CYP1B1, and by NAT2, sue. These activated metabolites bind to DNA to form DNA Breast cancer is the end result of a multistep pro - adducts that in turn can damage DNA (USDHHS 2010). cess in which some epithelial cells in the breast undergo Elevated levels of DNA adducts have been associated with a series of mutations. In doing so, these cells escape from certain types of cancer, supporting a positive association programmed cell death and then proliferate and invade - between increasing levels of DNA adducts and risk of can surrounding tissue (Armitage and Doll 1957; Fisher 1958; - cer (Phillips 2005). The degree of activation of detoxifica - Cairns 1975; Tomlinson et al. 1996). Genetic and epigen - tion enzymes—such as glutathione S transferases (GSTs), etic mutations in critical genes in cells—such as tumor uridine-5 ′ -disphosphate-glucuronosyltransferases (UGTs), suppressor genes, DNA replication and repair genes, and epoxide hydrolases, and sulfatases, which are also pres - - proto-oncogenes—can lead to the initiation of tumorigen ent in the breast—is important because these enzymes esis. Clones from these mutated cells continue to expand catalyze the excretion of the toxic metabolites, thereby and proliferate, rendering them susceptible to further potentially decreasing the formation of DNA adducts. 210 Chapter 6

240 The Health Consequences of Smoking —50 Years of Progress highest PAH-DNA adduct levels and the risk of breast can Smoking induces activity of some of these enzymes - cer, but they did not observe a dose-response relationship (USDHHS 2010). As a biomarker, smoking-related DNA adducts are an with increasing adduct levels. Furthermore, the strength integrated measure of exposure to tobacco smoke, meta of the association did not differ between active and passive - bolic activation, and delivery of the metabolite to DNA in - smokers. To date, no prospective cohort study has incor the target tissue (Groopman et al. 1995). Smoking-related porated these markers. Polymorphisms in genes encoding enzymes involved DNA adducts can be quantified in breast fluid, tissue, in the metabolic activation and detoxification of toxins, and peripheral blood cells. However, an increase in the levels of DNA adducts does not directly correspond to a such as those from exposure to cigarette smoke, could similar increase in cancer risk because other processes also affect breast carcinogenesis by either promoting or preventing the formation of DNA adducts. Firozi and col - are involved (Phillips 2005). To causally link the presence leagues (2002) observed a significant interaction between of smoking-related DNA adducts to risk of breast cancer, levels of DNA adducts in breast tissue and CYP1A1, elevated levels ideally need to be detected in breast epithe - NAT2 and polymorphisms among ever smokers. GSTM1, lial cells before the onset of the cancer and at higher levels in those individuals going on to develop cancer than in These authors also observed higher levels of DNA adducts *2/*2 those who do not. Levels of DNA adducts measured at the among smokers with combined CYP1A1*1/*2 or genotypes than among smokers with time of diagnosis or after diagnosis (e.g., in case-control GSTM1 null and or cross-sectional studies) may not reflect the etiologically polymorphisms in either genes. In addition, the frequency of smoking-related DNA adducts was higher in those with relevant time window of tumor initiation. Similarly, levels slow acetylator alleles of the NAT2 gene than in those hav- of DNA adducts in peripheral cells may not reflect what ing rapid acetylator alleles. is happening locally at a specific target site: circulating Several studies have examined the association levels of biomarkers have not always been correlated with mutations, and/or protein expres p53 between smoking, - levels at the tissue site. Several studies have evaluated the relationship sion in breast tumors; results have been mixed (Conway et al. 2002; Furberg et al. 2002; Gaudet et al. 2008; Van between smoking and the prevalence of smoking-related Emburgh et al. 2008a). Mordukhovich and colleagues DNA adducts in breast tissue (Perera et al. 1995; Li et al. (2010), who conducted a large case-control study of 859 1996; Rundle et al. 2000). These studies have confirmed cases and 1,556 controls from the Long Island Breast the presence of smoking-related DNA adducts in breast Cancer Study Project, found that women in the study tumor cells and adjacent normal epithelial cells in some, with p53-positive tumors were less likely to have been but not all, current and former smokers (Perera et al. p53 1995; Li et al. 1996; Rundle et al. 2000; Faraglia et al. exposed to cigarette smoke than women without 2003). Some case-control studies have reported high lev- mutations. This finding suggests that smoking may not significantly affect the p53 pathway. In this study, - p53 els of DNA adducts in smokers compared with nonsmok ers (Perera et al. 1995; Li et al. 1999; Conway et al. 2002; mutations were identified from DNA extracted from par - Li et al. 2002; Rundle et al. 2002). Faraglia and colleagues affin blocks and p53 protein expression was evaluated using immunohistochemistry. (2003) conducted a large, comprehensive case-control - study that included 148 breast tumor tissues and adja cent normal samples from the Long Island Breast Cancer Other Cellular Mechanisms Study Project. The arylamine 4-aminobiphenyl (4-ABP) In addition to forming DNA adducts, constituents of DNA adduct was measured using an immunoperoxidase tobacco smoke may contribute to carcinogenesis by pro- method that had been validated by mass spectrometry. moting cell growth and proliferation through the activa- The study’s authors observed a significant trend between tion of a number of receptors, such as cyclooxygenase II levels of 4-ABP DNA adducts in normal breast tissue and and prostaglandin E2, and signaling pathways, including smoking status, and they measured higher levels of DNA Akt and epidermal growth factor receptor (Narayan et al. adducts in active and passive smokers than in never smok- 2004; Miller et al. 2005; Kundu et al. 2007; Botlagunta - ers. Interestingly, mean levels of DNA adducts were sig et al. 2008; Guo et al. 2008; Connors et al. 2009; Dasgupta nificantly lower in tumor tissue than in adjacent normal et al. 2009) . Constituents of tobacco smoke may also cause tissue. cells to evade apoptosis after DNA damage by altering Elsewhere, circulating levels of PAH-DNA adducts in cellular response at the mRNA and protein levels (Con- peripheral blood mononuclear cells were assessed in two . In addition, cigarette smoke can inac - nors et al. 2009) sample sets taken 4.5 years apart from the same case-con - tivate tumor suppressor genes via genetic and epigenetic trol study (Gammon et al. 2004b). The authors observed Narayan and colleagues (2004) (Liu et al. 2010a) . changes a modest association in both sets of samples between the 211 Cancer

241 Surgeon General’s Report activation of signaling pathways via the estrogen receptor found that cigarette smoke condensate increases levels (ER), which leads to altered gene expression of GADD45 —a gene whose expression is upregulated in and increased proliferation and, in turn, the opportunity for more muta - response to DNA damage and/or growth arrest in a dose- tions. The second pathway involves the oxidative metabo dependent manner—to increase proliferation of epithelial - cells and to induce cell cycle arrest at the synthesis/gap lism of estrogen (E2/E1) to catechol estrogens and then / 2 to reactive quinone metabolites. The quinone metabolites mitosis (S/G Dasgupta and col- /M) phase. Furthermore, 2 have the ability to form depurinating DNA adducts or to leagues (2009) found that the exposure of human breast cancer cells to nicotine can contribute to epithelial- form catechols through the oxidation-reduction cycle that mesenchymal transition, a collection of changes seen in - produce reactive oxygen species causing oxidative dam age to DNA (Lavigne et al. 2001). The catechols can be more advanced cancers that is characterized by loss of O -meth- cell adhesion, increased cell mobility, and repression of inactivated by methylation mediated by catechol- E-cadherin. These mechanistic studies were conducted yltransferase, glucuronidation, and sulfation. In women, blocking the action of the ER by such agents as tamoxifen, in cell culture experiments using normal and malignant a selective estrogen receptor modulator, or by decreas- breast epithelial cell lines, but they have yet to be repli - ing estrogen production (e.g., by removing the ovaries in cated in an in vivo model. premenopausal women) has been shown to decrease the incidence of breast cancer up to 50% (Fisher et al. 1998; Hormones Parker et al. 2009). Estrogen metabolism, which occurs Estrogen’s role in the initiation, promotion, and in the liver, kidney, and other organs, including the progression of breast cancer is well established through breast, involves a complex set of pathways (Figure 6.27). preclinical data, observational studies, and clinical trials Various CYP isoforms, which are often tissue specific, are - (Yager and Davidson 2006). Studies in experimental ani responsible for the oxidation and conjugation of estrogen mal models and cultured human cells demonstrate that metabolites. One of the first steps in estrogen metabolism estradiol (E2) and estrone (E1) are carcinogenic (Yager is the oxidation of the parent estrogens (E2/E1) at the and Davidson 2006). Estrogen is thought to exert its car - 2, 4, and 16 positions of the carbon skeleton to the 2, 4, cinogenic effects primarily through two complementary and 16 hydroxylated metabolites (Yager and Liehr 1996). The first pathway involves the pathways (Figure 6.26). Figure 6.26 Pathways to estrogen carcinogenesis Yager and Davidson 2006, updated for Surgeon General’s Report. Adapted from Source: = 16 -OH E1 α 16 = 4-hydroxyestradiol; 4-OH E2 4-OH E1 = 4-hydroxyestrone; = estradiol; Note: E2 = estrone; E1 -hydroxyestrone; α = estrogen receptor. ER 212 Chapter 6

242 The Health Consequences of Smoking —50 Years of Progress Pathways involved in estrogen metabolism Figure 6.27 Source: Adapted from Ziegler et al. 2010. Davis and colleagues (1993) showed that the 16 hydroxy Observational studies have linked cigarette smok- ing to earlier age at menopause (Baron et al. 1990; estrogens exhibit strong estrogenic and mitogenic activi - Bromberger et al. 1997) and reduced bone density in ties and hypothesized that higher levels of such activities postmenopausal women (Daniell 1976; Baron et al. 2001); increase the risk for breast cancer by uncontrolled cellular - both conditions are associated with relative estrogen proliferation and by binding to the ER, thereby damag deficiency and a reduction in the risk for breast cancer. ing DNA. The 2- and 4-hydroxy metabolites also exhibit estrogenic activity and can stimulate cellular prolifera - - Smoking is also associated with decreased fertility (USD HHS 2004, 2010) and with earlier menarche in children tion. Despite being more abundant than the 4-hydroxy whose mothers were heavy smokers during pregnancy metabolite, the 2-hydroxy metabolite is much less potent (Windham et al. 2004); both conditions are known risk and shorter acting. Both the 2- and 4-hydroxy estrogen factors for breast cancer. However, as noted in the 2001, metabolites can go on to form genotoxic reactive quinone 2004, and 2010 Surgeon General’s reports, the majority of metabolites. epidemiologic studies comparing circulating endogenous In cell culture studies of granulosa cells, chorio - carcinoma cells, and placental microsomes, nicotine was ) and post- S estrogen levels in premenopausal (Table 6.11 menopausal women (Table 6.12 S ) have not found differ - - shown to directly inhibit the aromatase enzyme, result ing in reduced conversion of androgens to estrogen in a ences between smokers and nonsmokers. In several small studies, premenopausal women who smoked were found dose-dependent manner (Barbieri et al. 1986a,b). This is to have significantly elevated urinary levels of 2-hydroxy an important pathway, particularly in postmenopausal E1 or reduced levels of E1, E2, or estriol (E3) during the women among whom estrogen is generated primarily luteal phase of the menstrual cycle compared with non in peripheral tissues. In animal studies, cigarette smoke - reduced the number of oocytes, caused toxicity to ovarian smokers (MacMahon et al. 1982; Michnovicz et al. 1986, - follicles, and led to ovarian atresia (Mattison 1982; Black 1988; Westhoff et al. 1996). The clinical implications of burn et al. 1994; Miceli et al. 2005), which could affect these findings and any associated changes in breast tissue have not been investigated. estrogen production in premenopausal women. 213 Cancer

243 Surgeon General’s Report - levels of androgens (e.g., androstenedione, dihydroepian- Studies that compared the effect of HRT, an exoge drosterone sulfate, and testosterone), progesterone, and nous hormonal exposure, in smokers and nonsmokers did cortisol have been found to be higher in smokers than in observe differences by smoking status in circulating levels of estrogen and its metabolites, supporting the hypothesis nonsmokers. In postmenopausal women, these elevated that smoking increases hepatic metabolism of estrogens - levels may affect breast carcinogenesis. Missmer and col leagues (2004) associated increased levels of circulating (Jensen et al. 1985; Jensen and Christiansen 1988; Cas - androgens with increased risk for breast cancer among sidenti et al. 1990; Geisler et al. 1999). Among postmeno - pausal women who were using orally administered HRT, postmenopausal women. A meta-analysis by Law and col- - leagues (1997) found that levels of dihydroandroepiand circulating estrogen metabolites—including E1, E2, and estrone sulfate—were 40–70% lower in smokers than in osterone sulfate and androstenedione were significantly nonsmokers (Jensen et al. 1985; Jensen and Christiansen higher in postmenopausal smokers than in nonsmokers 1988; Cassidenti et al. 1990; Geisler et al. 1999). A dose- but that levels of estrogens did not differ. Finally, cigarette dependent, reciprocal increase in the binding capacity of smoking has been shown to directly affect adrenal cortical hormone levels (Baron et al. 1995). The effects of these sex-hormone-binding globulin was observed by Cassidenti hormonal changes on breast tissue are not known. - and colleagues (1990) and, importantly, differences in lev els of estrogen and its metabolites were not evident before treatment with HRT in these same women (Jensen et al. Summary 1985; Cassidenti et al. 1990). Furthermore, significant The available evidence supports biologically plau - - changes in circulating hormone levels between smok sible mechanisms, particularly for DNA adduct formation ers and nonsmokers were not observed after transdermal and unrepaired DNA mutations, by which exposure to administration of HRT, a method that bypasses estrogen tobacco smoke could cause breast cancer. However, data metabolism in the liver (Geisler et al. 1999; Mueck and are limited and a detailed mechanistic model of how expo- Seeger 2005). sure to tobacco smoke may affect risk for breast cancer Alterations in estrogen metabolism pathways have cannot yet be assembled. also been observed in pregnant women who smoked (USDHHS 2001). Several studies have found that pregnant women who smoked had lower levels of circulating E2 and Epidemiologic Evidence—Overview E3 than pregnant women who did not smoke (Targett et al. 1973; Mochizuki et al. 1984; Bernstein et al. 1989; Petri- The following sections update and expand the dou et al. 1990; Kaijser et al. 2000). However, compared reviews in previous Surgeon General’s reports on the with their nonsmoking pregnant counterparts, rates of associations between cigarette smoking and breast cancer 4-hydroxylation were increased in pregnant smokers in and between exposure to secondhand smoke and breast samples of placental tissue (Chao et al. 1981; Juchau et al. cancer. Conclusions from previous reports and recent epi - 1982), and rates of 2-hydroxylation were nonsignificantly demiologic evidence are summarized with reference to increased (Juchau et al. 1982). Smoking did not alter E2 the criteria for the assessment of causation used in this metabolism or the formation of E1, 2-hydroxyestradiol, (Hill 1965; USDHHS 2004) series of reports . The stud- and other estrogen metabolites, but 15 α -hydroxyestradiol, ies reviewed cover a lengthy period of time and include a -hydroxyestradiol were sig- α 4-hydroxyestradiol, and 7 - variety of study designs and inclusion criteria, data collec nificantly elevated (Zhu et al. 2002). Finally, Piasek and tion techniques, exposure measurements, and study end- colleagues (2001) found that levels of progesterone were points. Reports based on cohort studies prior to 2012 and lower in pregnant women who smoked than in those who case-control studies published between 2000–2011 were did not smoke. If the rate of 4-hydroxylation continues to identified in MEDLINE using key words and extended be higher after pregnancy in smokers than in nonsmok- terms. All studies that evaluated the association between ers, then smoking may increase risk for breast cancer smoking and breast cancer risk and mortality were eligi - rather than having a protective effect, as suggested by the ble for review. Combinations of the following key words anti-estrogenic hypothesis proposed by Michnovicz and were used, depending on the evidence sought: breast can- colleagues (1986). cer, breast neoplasms, tobacco smoke, cigarette smoking, Several other circulating hormones have also been active smoking, passive smoke, secondhand smoke, invol- compared between smokers and nonsmokers. In pre- untary smoke exposure, case-control study, cohort study, menopausal women, Cramer and colleagues (1994) and risk, survival, mortality, prognosis, recurrence, second - Windham and colleagues (2005) did observe higher lev - primary, genotype, polymorphism, single nucleotide poly - els of circulating follicle-stimulating hormone in smok morphisms (SNPs), NAT1, NAT2, CYP1A1 and CYP1B1, S ). Last, circulating ers than in nonsmokers (Table 6.11 Chapter 6 214

244 The Health Consequences of Smoking —50 Years of Progress - with smoking of long duration, smoking before a first full GST, GSTM1, GSTT1, GSTP1, GSTA1, SULT1A1 , MnSOD2, - term pregnancy, and passive smoking require (sic) confir , and BRCA1 . XRCC1, XPD or ERCC2, MGMT, and BRCA2 Additional studies were identified from reference lists in mation in future epidemiological studies” (p. 965). They pertinent papers. The search focused on English-language suggested that future studies and meta-analyses consider - studies that evaluated either (a) the main effects of ciga timing of exposure (e.g., age at initiation of smoking and rette smoking or passive exposure to smoke on breast smoking before first pregnancy), duration and dose (years of exposure and pack-years of smoking), sources of pas - cancer risk or mortality, or (b) the interaction of cigarette sive exposure, the overlap of active and passive exposures, - smoking or passive exposure to smoke with such risk fac tors as menopausal status, hormone receptor status, fam - potential confounders, and modification by menopausal ily history, and susceptibility genotypes. All studies that status and genetic susceptibility. - reported a main effect for smoking are identified in the IARC (2004) summarized results from 36 case-con trol studies, 8 cohort studies, and a large pooled analysis sections below on active smoking (see “Active Cigarette Smoking and Risk for Breast Cancer”) and exposure to of data from 10 cohort and 43 case-control studies, the secondhand smoke (see “Exposure to Secondhand Smoke pooled analysis having been conducted by the Collabora - tive Group on Hormonal Factors in Breast Cancer and and Risk for Breast Cancer”), regardless of whether they colleagues (2002) and based on studies having at least were one of multiple studies on the same population. 100 women with incident invasive breast cancer. The However, when multiple studies were reported for the pooled analysis was restricted to nondrinkers (38% of same population, only the most recent findings, with a few cases and 43% of controls), because alcohol was consid - exceptions noted in the analytical sections, were included ered a potentially significant confounder of the effects of in the meta-analyses presented later. smoking. Sufficient data were available to consider a wide variety of other potential confounders, including age at diagnosis, parity, age at birth of first child, breastfeeding, Active Cigarette Smoking and Risk race, country, education, family history, age at menarche, for Breast Cancer height, weight, body mass index (BMI), use of hormonal contraceptives, and menopausal status. Study site, age, Individual authors and various review panels parity, and age at first birth were included as covariates have evaluated the evidence for an association between in the final analysis of the effect of smoking on risk of active and passive cigarette smoking and breast can - breast cancer among nondrinkers. However, the analy - cer. The first systematic review of such an association sis did not consider duration or amount of smoking or was included in IARC Monograph 38 (1986). Based on a exposure to secondhand smoke. Results indicated no asso- review of 10 case-control and 8 cohort studies published ciation between active smoking and risk for breast cancer between 1959 and 1983, the 1986 IARC monograph found (RR = 1.03; 95% CI, 0.98–1.07) in women who did not “no consistent effect of smoking on the risk of breast drink alcohol. The Collaborative Group (2002) also con- cancer” (p. 298). The literature at the time was limited, trasted this result with those for all women regardless of however. Only 2 of the case-control studies (CDC 1983; alcohol intake (RR = 1.09) and statistically adjusted for Janerich et al. 1983) were population-based, rather than alcohol intake (RR = 1.05). The 2004 IARC report con - - hospital-based, and few studies adjusted for potential con cluded that: (a) the majority of epidemiologic studies founders. All but 1 cohort study (Hiatt et al. 1982) mixed - “found no association with active smoking, after control incident and decedent cases and few adequately adjusted - ling for established risk factors”; and (b) the Collabora for relevant confounders. Palmer and Rosenberg (1993) tive Group analysis of women who reported themselves to reviewed 5 cohort and 16 case-control studies (9 with be nondrinkers “confirms the lack of an increased risk of population controls, 3 with participants in a screening breast cancer associated with smoking” (p. 1183). The Cal/ program, and 4 with hospital controls), finding “little EPA reviewed many of the same studies in 2005 and came evidence to suggest that cigarette smoking materially to a different conclusion: “Considering the epidemiologi - increases risk” (p. 154). However, the authors noted that cal studies, the biology of the breast and the toxicology future investigations should consider age at initiation of of tobacco smoke constituents together, the data provide smoking because of evidence that women were begin- support for a causal association between active smoking ning to smoke at earlier ages. Terry and Rohan (2002) and elevated breast cancer risk” (p. 7-79). published a comprehensive literature review on cigarette In April 2009, the Canadian Expert Panel on Tobacco - smoking and breast cancer, concluding that “the associa Smoke and Breast Cancer Risk conducted an exten - tion between cigarette smoking and breast cancer risk sive descriptive evaluation of active cigarette smoking remains unclear” and that the observed “increased risk and exposure to secondhand smoke, paying particular 215 Cancer

245 Surgeon General’s Report durations of active smoking, ranging from 20 or more attention to the timing of these exposures (age at initial - exposure and before or during first full-term pregnancy), years to 31 or more years (Al-Delaimy et al. 2004; Reyn al. 2005). According to the duration and dose (years of exposure and number of pack- olds et al. 2004b; Gram et years of smoking), modification by menopausal status, Canadian Expert Panel, when these studies were consid - and genetic susceptibility (Collishaw et al. 2009). The ered along with three of the four older cohort studies al. 2002; Olson et al. 2005; Cui et al. 2006) (Calle (Egan et panel’s approach, to some extent, followed the suggestions et al. 1994 was excluded because it was a mortality study), of Terry and Rohan (2002) that future studies and meta- five reported an increased risk for the highest duration analyses focus more carefully on the issues of duration, category of smoking: two with borderline significance timing, genetic susceptibility, source of passive exposure, (RR = 1.15 [95% CI, 1.00–1.33]; 1.18 [95% CI, 1.00–1.38]) the overlap of passive and active exposure, and potential confounders. The evaluation included summative reviews, (Reynolds et al. 2004b; Olson et al. 2005, respectively) meta-analyses, and the most recently published stud - and three with statistical significance (RR = 1.21 [95% ies through November 2008. Pooled analyses and meta- CI, 1.01–1.45]; 1.36 [95% CI, 1.1–1.7]; and 1.50 [95% CI, 1.19–1.89]) (Al-Delaimy et al. 2004; Gram et al. 2005; Cui analyses were not performed. The evaluation paid particu- et al. 2006, respectively). However, it should be noted that - lar attention to results from the 2002 analysis by the Col laborative Group on Hormonal Factors in Breast Cancer, the result used for the Gram study is based on a subgroup the 2005 Cal/EPA report, and the 2004 and 2006 Surgeon of women who reported ever smoking for at least 20 years. General’s reports. The result for all current smokers with 25 or more years of The Canadian Expert Panel evaluated results from smoking was increased but not statistically significant (RR = 1.26; 95% CI, 0.98–1.63). Although four of these five more recent, updated analyses published for four of the - studies reported statistically significant trends across lev cohort studies and nine of the case-control studies that were included in the 2002 Collaborative Report in which els of duration (Olson did not calculate a p for trend), only duration of smoking was reported. Unlike the 2002 report, three (Gram et al. 2005; Olson et al. 2005; Cui et al. 2006) which excluded women who consumed alcohol, the actually showed unambiguous evidence of an increasing trend with duration of active smoking. The panel also Canadian panel reported risk estimates adjusted for alco- reviewed four cohort studies published after 2002 that hol intake. The four cohort studies included the NHS-I (Egan et al. 2002), the Canadian National Breast Screen - reported risk estimates by pack-years of smoking (Reyn - olds et al. 2004b; Gram et al. 2005; Olson et al. 2005; Cui ing Study (Cui et al. 2006), the CPS-II (Calle et al. 1994), et al. 2006). Among these studies, three had statistically and the Iowa Women’s Health Study (Olson et al. 2005). Three of these (Calle et al. 1994; Olson et al. 2005; Cui significant RRs ranging from 1.17–1.48 for the highest category of pack-years (Reynolds et al. 2004b; Gram et al. et al. 2006) reported significantly increased RRs, ranging from 1.18–1.50, for the longest duration of smoking (≥40 2005; Cui et al. 2006). Additionally, the panel reviewed years). Among the nine case-control studies (Rohan and 32 case-control studies in which ORs were reported for duration of active smoking and 27 in which estimates Baron 1989; Palmer et al. 1991; Smith et al. 1994; Baron were reported for pack-years. The results from these case- et al. 1996; Johnson et al. 2000; Kropp and Chang-Claude control studies were found to be inconsistent, regardless of 2002; Alberg et al. 2004; Magnusson et al. 2007; Prescott menopausal status. The Canadian Expert Panel concluded et al. 2007), five reported an increase in risk of greater than 45% for smoking durations ranging from 11 to more that the results from the cohort studies for increased risk with longer duration and higher pack-years were more than 50 years and for high cumulative levels of pack-years 3 (Rohan and Baron 1989; Palmer et al. or cigarette-years “persuasive” than those from the case-control studies and 1991; Johnson et al. 2000; Kropp and Chang-Claude 2002; “that the relationship between active smoking and breast However, results were statistically sig- Alberg et al. 2004). cancer is consistent with causality” (Collishaw et al. 2009, nificant only for postmenopausal women who reported p. 49). Johnson and colleagues (2011) summarized the more than 35 years of active smoking (OR = 1.7; 95% CI, results from the Canadian Expert Panel in a brief report. 1.1–2.7) in one study (Johnson et al. 2000). In November 2009, IARC issued a special report on The Canadian Expert Panel also evaluated three human carcinogens, including tobacco, that encompassed cohort studies published after 2002 in which the risk of - more than 150 epidemiologic studies about the associa breast cancer was significantly increased for the longest tion between tobacco smoke and breast cancer (Secretan 3 Cigarette-years: the number of years of smoking multiplied by the number of cigarettes smoked per day. 216 Chapter 6

246 The Health Consequences of Smoking —50 Years of Progress - ably identified,” and that the previous finding of a lower et al. 2009). This report updated findings and conclu sions from the 2004 IARC report and noted that two large or BRCA1 risk for breast cancer among women with cohort studies conducted after 2002 showed positive, but mutations in one study (Brunet et al. 1998) “was BRCA2 not replicated” in a later study (Couch et al. 2001) and small, statistically significant associations. These stud- ies included the California Teachers Study (hazard ratio therefore not established (USDHHS 2004, p. 312). The sections below review and quantitatively sum - [HR] = 1.13; 95% CI, 1.0–1.28) (Reynolds et al. 2004b), marize studies of cigarette smoking by study design which was also reviewed in the 2006 Surgeon General’s (cohort, case-control), and by geographic regions (North report, and the Canadian National Breast Screening Study America, Europe, Asia) that differ for smoking prevalence, (HR = 1.18; 95% CI, 1.09–1.27) (Cui et al. 2006). Based on these findings and those from previous reports, as well as well as breast cancer incidence and mortality. Table 6.13 as evidence from studies of animal and human tissues, shows selected estimates of the prevalence of smoking from the WHO Reports on the Global Tobacco Epidemic the IARC panel concluded that “there is limited evidence that tobacco smoking causes breast cancer” (Secretan (2008a, 2011) for countries represented in these reports. et al. 2009, p. 1033) and added the female breast as a new Although there is considerable variation, the prevalence of cancerous tumor site associated with exposure to tobacco smoking in women is generally similar in North America smoking. and Europe but substantially lower in Asia. In addition to these extensive reports, several The following sections include reports on the asso - reviews and meta-analyses have addressed active cigarette ciation between smoking and breast cancer risk based on cohort studies published up to 2012 (Table 6.14 S smoking alone (Khuder and Simon 2000; Khuder et al. ) and case-control studies published from 2000–2011 (Table 2001; Nagata et al. 2006; Ren et al. 2007), exposure to sec - ). A list of studies by category of exposure is provided 6.15 S ondhand smoke but not active smoking (Lee and Hamling S - in Table 6.16 . Studies based on incident cases that esti 2006; Pirie et al. 2008), both active and passive smoking mate risk of breast cancer are emphasized because studies (Morabia 2002b; Johnson 2005; Sadri and Mahjub 2007; - Iwasaki and Tsugane 2011), smoking-genotype interac- that focus on mortality may include a different mix of cor - tions (Vogl et al. 2004; Masson et al. 2005; Terry and Good relates and etiologic pathways affecting survival that alter the association with smoking (Al-Delaimy et al. 2004). As a man 2006; Ochs-Balcom et al. 2007; Ambrosone et al. 2008; Zhang et al. 2010), smoking-DNA repair marker result, studies of smoking and breast cancer mortality are interactions (Neumann et al. 2005), timing in relation to evaluated in a separate section (see “Exposure to Tobacco first pregnancy or birth of first child (Lawlor et al. 2004; Smoke and Breast Cancer Mortality”). Some studies or reviews that mix prevalent with incident cases, however, DeRoo et al. 2011b), and intrauterine exposure (Park are included (Lawlor et al. 2004; Hanaoka et al. 2005; Ha et al. 2008). et al. 2007). Conclusions from Previous Surgeon General’ Cohort Studies s Reports S The 2001 Surgeon General’s report on women and Table 6.14 presents an overview of 15 publications from the 12 cohort studies on breast cancer and active smoking concluded that “active smoking does not appear smoking published since 2000 (Manjer et al. 2000b, 2001; to appreciably affect breast cancer risk overall,” but it sug- gested that future research address both age at initiation of Egan et al. 2002; Terry et al. 2002a; Al-Delaimy et al. 2004; - Lawlor et al. 2004; Reynolds et al. 2004b; Gram et al. 2005; smoking and potential susceptibility associated with spe cific genetic polymorphisms (p. 217). The 2004 Surgeon Hanaoka et al. 2005; Olson et al. 2005; Cui et al. 2006; General’s report on the health consequences of smoking Ha et al. 2007; Lin et al. 2008; Xue et al. 2011; Luo et al. 2011b ) evaluated: (a) the influence that cigarette smoking has on Lawlor and colleagues (2004) . The study by endogenous estrogen levels due to changes in metabolism was restricted to parous women in the United Kingdom and lowered body weight; (b) the effects of early age at and combined prevalent and incident cases. The report smoking initiation, smoking-genotype interactions, and by Manjer and colleagues (2001) was based on the same - cohort as used in an earlier report by Manjer and col exposure to secondhand smoke; and (c) carcinogenic and leagues (2000b), but was restricted to women with tumor anti-estrogenic effects of smoking on breast tissues. tissue available for analysis. Consequently, Lawlor and The 2004 Surgeon General’s report concluded that colleagues (2004) and Manjer and colleagues (2001) are “evidence is suggestive of no causal relationship between active smoking and breast cancer,” that subgroups of excluded from the meta-analyses and forest plots. Addi - and women at high risk because of smoking could not be “reli- Cui Terry and colleagues (2002a) tionally, reports by 217 Cancer

247 Surgeon General’s Report Table 6.13 Age-standardized estimates of the prevalence of current cigarette smoking for selected member states of the World Health Organization (WHO), 2009 Males (%) Females (%) Member states WHO region United States 24 28 North America Canada 16 19 North America Denmark Europe 30 28 Europe 28 22 Finland Europe France 36 27 Europe Germany 33 25 Europe Italy 33 19 Europe Netherlands 28 22 Europe 28 31 Norway Europe Poland 36 25 Europe 22 30 Slovenia Europe 25 20 Sweden 21 Europe Switzerland 31 Europe 23 United Kingdom 25 Western Pacific Australia 22 19 50 Western Pacific China 2 India 11 1 Western Pacific 12 Western Pacific Japan 42 Western Pacific 6 53 Republic of Korea 47 Philippines Western Pacific 10 Source: Data for Republic of Korea and Sweden are from WHO 2008a (Appendix III, Tables 3.4b and 3.6b). Data for the other member states presented in this table are from WHO 2011 (Appendix VII, Table 7.1.0). Reprinted with permission from World Health Organization, © 2008, 2011. Note: Prevalence estimates are standardized to age distributions of the country’s current smoking. Estimates of current smoking are calculated based on cigarette smoking at the time of survey, daily or nondaily. Estimates rounded to nearest whole number. North American Studies were based on the same cohort, and colleagues (2006) the Canadian National Breast Screening Study. Although The U.S. Radiologic Technologists Study (USRTS) S Table 6.14 summarizes both studies, estimates only from (Ha et al. 2007) reported nonsignificantly increased RRs Cui and colleagues (2006) are used in the meta-analyses to for breast cancer among former smokers (RR = 1.17; 95% avoid duplication. Two reports stem from the NHS-I (base - CI, 0.99–1.38) and current smokers (RR = 1.13; 95% CI, line year 1976) (Egan et al. 2002; Xue et al. 2011); data 0.96–1.32). Although the study adjusted for the first year from the more recent report are used in the majority of the in which an individual worked as a radiologic technician, meta-analyses. Data from the NHS-II are included because either residual confounding or synergy may have occurred it is a separate premenopausal-women-only cohort with a between smoking and exposure to radiation at work, different baseline year (1989) (Al-Delaimy et al. 2004) . All because previous analyses showed an increased risk (RR = S three of these studies are summarized in Table 6.14 . Chapter 6 218

248 —50 Years of Progress The Health Consequences of Smoking - data from the NHS-II cohort (1989 baseline), based pre- 2.41; 95% CI, 1.32–4.41) associated with exposure to radi dominantly on premenopausal women, and reported a sig- ation at work among women who were employed before - nificantly increased risk for breast cancer among former 1940, when exposure to occupational radiation was poten smokers (RR = 1.18; 95% CI, 1.02–1.36) but not current tially higher for this group (Mohan et al. 2003; Sigurdson smokers (RR = 1.12; 95% CI, 0.92–1.37) in comparisons et al. 2003). Eighty-six percent of cases (781 out of 906) were ascertained by self-report, of which 20% (155 out of with never smokers over an average of 10 years of follow- 781) could not be verified against medical records (Ha et up. This study also reported a positive association between risk for breast cancer and increasing duration of smok al. 2007), but a previous report from the USRTC indicated - 99.4% agreement between self-report and breast cancers ing (p trend = 0.04) and a significantly increased risk for smoking 20 years or more (RR = 1.21; 95% CI, 1.01–1.45). ascertained from medical records (Sigurdson et al. 2003). Cui and colleagues (2006), using data from the In addition, 14% (125) of the cases were ascertained from Canadian Breast Screening Study (1980–1985 baseline, death certificates, and date of diagnosis was imputed using 40–59 years of age) reported an increased risk for breast the average survival time between 1973 and 2000 based on cancer among current smokers (RR = 1.18; 95% CI, SEER data. 1.09–1.27) but not former smokers (RR = 1.00; 95% CI, Several reports from the NHS-I cohort on smok- 0.93–1.08). This report was an update of the same cohort ing and breast cancer are based on both premenopausal and postmenopausal women (Willett et al. 1987; London from an analysis by Terry and colleagues (2002a), but at et al. 1989; Hunter et al. 1997; Collaborative Group on an average of 16 years of follow up for 4,445 cases rather Hormonal Factors in Breast Cancer et al. 2002; Egan et al. than an average of 10.6 years for 2,552 cases. Overall, few differences can be found between these two reports. The 2002; Xue et al. 2011). All but two (Egan et al. 2002; Xue 16-year follow-up study, using never smokers as the refer - et al. 2011) have been incorporated in previous reviews or ent, found significant associations between risk for breast meta-analyses. The report by Xue and colleagues (2011) cancer and the highest categories of cigarettes smoked per analyzed data for 8,772 incident cases of breast cancer in 111,440 women followed for up to 30 years (1976 base day (RR = 1.20; 95% CI, 1.00–1.44), duration of smoking - (RR = 1.50; 95% CI, 1.19–1.89), and pack-years of smok line). The RR was 1.09 (95% CI, 1.02–1.17) for current - smokers and 1.06 (95% CI, 1.01–1.11) for former smok ing (RR = 1.17; 95% CI, 1.02–1.34), as well as for smoking - ers in comparisons with never smokers. These estimates for more than 5 years before first pregnancy (RR = 1.13; 95% CI, 1.01–1.25) and for initiation of smoking between were adjusted for age, family history of breast cancer, age at menarche, height, BMI at 18 years of age, use of oral 16 and 19 years of age (RR = 1.10; 95% CI, 1.01–1.21). contraceptives, history of benign breast disease, physical Effect modification by menopausal status was not found, activity, alcohol use, age at first birth, parity, passive smok - but positive associations were stronger in women who did not report vigorous physical activity. ing at home and at work, current BMI, age at menopause, In the Iowa Women’s Health Study cohort, Olson menopausal status, and use of HRT. Risk was significantly and positively associated with increasing pack-years of and colleagues (2005) reported a significantly increased risk (vs. never smokers) for breast cancer among post - smoking (p trend = 0.001), number of cigarettes smoked per day (p trend = 0.02), and duration of smoking (p trend menopausal current smokers (RR = 1.19; 95% CI, 1.03– 1.37) but not former smokers. Increased risks were also = 0.01), particularly in the intervals between menarche and menopause and menarche and first birth (RR = 1.11; detected for age at smoking initiation (older than 18 years 95% CI, 1.07–1.15; and 1.18; 95% CI, 1.10–1.27 for every of age: RR = 1.11; 95% CI, 1.00–1.24), smoking duration 20 pack-years, respectively). The authors also reported a (≥40 years: RR = 1.18; 95% CI, 1.00–1.38), and smoking before first pregnancy (RR = 1.21; 95% CI, 1.07–1.37). significant trend (p = 0.02) toward decreasing risk with Reynolds and colleagues (2004b) used data from increasing pack-years of smoking after menopause. the California Teachers Study to evaluate the association In an earlier report from the NHS-I, Egan and col- - leagues (2002) reported results for a subset of 78,206 pre- of smoking with breast cancer. The authors detected sig - menopausal or postmenopausal women that was restricted nificantly increased risks for breast cancer among cur - to those for whom data on both active and passive smok rent smokers in comparisons with two reference groups: never smokers (RR = 1.32; 95% CI, 1.10–1.57) and women ing were collected and based on an average of 14 years of who reported no active or passive exposure to smok- follow-up (1982 baseline). The RR, based on a reference ing (RR = 1.25; 95% CI, 1.02–1.53). Results for former group of women who reported no history of active or pas- sive smoking, was 1.15 (95% CI, 0.98−1.34) and 1.17 (95% smokers, when compared with women who reported no - active or passive exposure to smoking or with never smok- CI, 1.01–1.34) for current and former smokers, respec tively. Last, Al-Delaimy and colleagues (2004) analyzed ers, were attenuated and not significant, regardless of 219 Cancer

249 Surgeon General’s Report ies provide contradictory results for the effect of smoking reference group. This study reported significant trends toward increasing risk of breast cancer with longer dura- on risk of breast cancer in postmenopausal women. tion and greater pack-years of smoking and more ciga- Last, in a companion report from the Women’s rettes smoked per day. In addition, risk of breast cancer was Health Initiative, Luo and colleagues (2011a) provided increased in women who initiated smoking before 20 years results suggesting that the risk of breast cancer is greater of age (RR = 1.17; 95% CI, 1.05–1.30) and who smoked for - in nonobese women who smoke. The RR for current smok ing was 1.25 (95% CI, 1.05–1.47) in nonobese women (BMI 5 or more years before first pregnancy (RR = 1.13; 95% <30) versus 0.96 (95% CI, 0.69–1.34) in obese women. CI, 1.00–1.28). In response to a letter by Johnson (2004), Significant trends in risk were found for age at initiation, Reynolds and colleagues (2004a) conducted additional duration and pack-years of smoking, and cigarettes per analyses to evaluate the associations for smoking duration, day in nonobese but not in obese women. The RR for 50 pack-years of smoking, and average number of cigarettes or more years of smoking was 1.62 (95% CI, 1.22–2.17) smoked per day with risk of breast cancer stratified by nulliparous women only, parous women who smoked less in nonobese women but only 0.62 (95% CI, 0.28–1.40) in than 5 years prepartum, and parous women who smoked obese women. This is one of three studies to date that have for 5 or more years prepartum. These analyses suggested a examined the interaction of smoking and body size on stronger risk effect among parous women who smoked for risk of breast cancer and the only one to formally test for statistical interaction; the other studies have been case- 5 or more years before first pregnancy for duration, pack- years, and cigarettes smoked per day (RR = 1.12, 1.28, and control. Gammon and colleagues (2004a) also reported an increased risk of breast cancer in lean women (BMI <22.3) 1.25, respectively, for highest levels) than for women who exposed to both active and passive smoking (OR = 1.76; smoked for less than 5 years before first pregnancy (RR = 95% CI, 1.06–2.92) but no association for obese women - 1.18, 1.12, and 1.11, respectively, for highest levels) com - pared with their nonsmoking counterparts. Results were (BMI >29.2) in their case-control Long Island Breast Can cer Study Project. In contrast, Band and colleagues (2002) significant for the highest levels for pack-years and ciga- found a nonsignificant inverse association in ever smokers rettes smoked per day for parous women who had smoked for 5 or more years prior to pregnancy. Risk of breast with a BMI less than 21 (RR = 0.75; 95% CI, 0.29–1.94) but an increased risk in those with a BMI 21 or greater (RR cancer was increased among nulliparous women (RR = = 1.13; 95% CI, 0.63–2.04); however, the latter result is 1.13, 1.33, and 1.37, respectively, for highest levels), but significant for only those women who reported smoking for lean, normal, overweight, and obese women combined and therefore cannot be compared with the other studies. 20 or more cigarettes per day compared with nonsmoking Luo and colleagues (2011a) speculated as to whether this nulliparous women. Luo and colleagues (2011b) reported results for - interaction could be associated with either an anti-estro 3,520 cases among 79,990 postmenopausal women fol - genic effect of smoking or with different distributions of genetic susceptibility polymorphisms in obese versus non lowed for an average of 10.3 years in the Women’s Health - obese postmenopausal women. Initiative Observational Study cohort. The RRs for former and current smokers were 1.09 (95% CI, 1.02–1.17) and 1.16 (95% CI, 1.00–1.34), respectively, when based on a European Studies reference group of never smokers. These risks increased Since 2000, three European cohort reports have about 7–8% when based on a no active/no passive expo - been published for findings on two studies of smoking and sure reference group (RR = 1.16; 95% CI, 0.98–1.38 and Gram and colleagues (2005) studied risk for breast cancer. - RR = 1.24; 95% CI, 1.00–1.54, respectively). Risk was sig the Norwegian-Swedish Cohort, a large population-based nificantly (p <0.05) and inversely associated with age at cohort (n = 102,098) in Scandinavia with up to 9 years initiation of smoking, and it was positively associated with of follow-up. Although the study detected nonsignificant cigarettes per day, duration, and pack-years of smoking. - increased risks for breast cancer among former smok The RR for 50 or more pack-years of smoking was 1.18 ers (RR = 1.15; 95% CI, 0.94–1.41) and current smok - (95% CI, 1.02–1.37), very similar to the estimate of 1.19 ers (RR = 1.17; 95% CI, 0.95–1.40), it found some strong (95% CI, 1.07–1.33) for 51 or more pack-years reported by associations with timing of smoking initiation, duration Xue and colleagues (2011) for the NHS-I. It is important to - of smoking, and smoking dose. Risk estimates for initia note, however, that the estimate for the Women’s Health tion of smoking before 15 years of age (RR = 1.48; 95% CI, Initiative (Luo et al. 2011b) is for postmenopausal women 1.03–2.13), “before/around menarche” (RR = 1.39; 95% only; the NHS-I (Xue et al. 2011) reported a significant CI, 1.03–1.87), and before first pregnancy (RR = 1.27; 95% (p = 0.02) inverse association with pack-years of smoking CI, 1.00–1.62) were all significantly associated with breast - after menopause but a strong (p <0.001) positive associa cancer among women who reported smoking for at least 20 - tion before menopause. Thus, these two large cohort stud 220 Chapter 6

250 The Health Consequences of Smoking —50 Years of Progress pausal women to conduct a formal test of interaction, but years in comparisons with never smokers. Among women the results suggest the possibility of effect modification with 20 or more years of smoking, significant increased by menopausal status. The Japan Public Health Center- risks were also reported for smoking at least 10 cigarettes per day (RR = 1.34; 95% CI, 1.06–1.70), accumulating 20 based prospective cohort study by Hanaoka and colleagues - or more pack-years of smoking (RR = 1.46; 95% CI, 1.11– (2005) also lacked statistical power, with only 180 inci 1.93), and smoking for at least 25 years (RR = 1.36; 95% dent cases among 21,805 women and a smoking preva- CI, 1.06–1.74) in comparison with never smokers. These lence of 5.7%. Moreover, the analyses appeared to mix incident morbidity data with mortality data. The RRs were results were attenuated on the order of 1–7% when ana - 1.7 (95% CI, 1.0–3.1) for current smokers and 1.1 (95% lyzed for current smokers and were no longer significant CI, 0.4–3.5) for former smokers, using a no active/no pas- except for pack-years of smoking and number of cigarettes S smoked per day, as shown in Table 6.14 Manjer sive reference group. Among premenopausal women, the . Earlier, reported results from a smaller RR was significantly increased, but imprecisely estimated and colleagues (2000b) for ever smokers (RR = 3.9; 95% CI, 1.5–9.9); the study cohort study (n = 10,902) conducted in Malmö, Sweden. found no increased risk among postmenopausal women In premenopausal and postmenopausal women combined, (RR = 1.1; 95% CI, 0.5–2.5). former smoking—but not current smoking or number of cigarettes smoked per day—was significantly associated Case-Control Studies with risk for breast cancer (RR = 1.31; 95% CI, 1.02–1.69). Since 2000, there have been 34 reports based on 30 Asian Studies case-control studies on smoking and breast cancer (Table ) S 6.15 Since 2000, studies published have included a sys- The reports provided by Metsola and colleagues . (2005) and Sillanpaa and colleagues (2005a) were based tematic review of three cohort and eight hospital-based on the same study group, and both used a no active/no case-control studies by Nagata and colleagues (2006) and a single cohort study by Lin and colleagues (2008). The passive exposure reference group. Because the report by - Sillanpaa and colleagues (2005a) adjusted for a number three cohort studies in the review by Nagata and col - of potential confounders and these adjustments made a leagues (2006) included the study by Hanaoka and col difference in the reported estimates, this report is used in leagues (2005) of middle-aged Japanese women, a study presents S of atomic bomb survivors by Goodman and colleagues the meta-analyses and forest plots. Table 6.15 studies are limited . Seven an overview of these studies (1997), and a study of breast cancer mortality by Hirayama - by either a small sample (<200 cases) with low statisti (1984, 1990). All eight case-control studies were con- (Delfino et al. 2000; Morabia et al. 2000; Alberg cal power ducted before 2000. In addition to multiple problems with or et al. 2004; Gibson et al. 2010; Kaushal et al. 2010) the design of these studies, their results are difficult to by other design features that limit interpretation, such interpret and have poor generalizability because of the (Delfino et al. 2000; Kruk 2007; as clinic-based controls low incidence of breast cancer and very low prevalence Cerne et al. 2011) or benign breast disease controls of smoking among Asian women (Table 6.13). Although (Delfino et al. 2000) . These studies vary considerably in the prevalence of smoking is very low among Chinese reporting type and detail for measures of smoking and women (2%) and low among Japanese (12%) women, it whether results are stratified by ethnicity, menopausal is high among Chinese (50%) and Japanese (42%) men status, or genetic biomarkers. (Table 6.13, based on WHO 2011). Thus, women in Asia are exposed to secondhand smoke more so than to active North American Studies cigarette smoking. The study by Lin and colleagues (2008) included Since 2000, findings on smoking and risk for breast approximately 12 years of follow-up of 34,401 women cancer have been reported across seven large population- (Table 6.14 S ). However, the study had limited power to based case-control studies with at least 1,000 cases (John- detect an association between smoking and breast cancer son et al. 2000; Innes and Byers 2001; Band et al. 2002; because of a small number of cases (n = 208) and the low Gammon et al. 2004a; Mechanic et al. 2006; Prescott - prevalence of current smoking (1.6%) and former smok et al. 2007; Slattery et al. 2008; Young et al. 2009) . The ing (5.3%). The RRs for breast cancer were 0.67 (95% CI, reports by Fink and Lash (2003) and DeRoo and colleagues 0.32–1.38) for current smokers and 1.27 (95% CI, 0.46– (2011a) are not included in this section because they dealt 3.48) for former smokers. However, when the analysis was exclusively with smoke exposure during pregnancy. Young restricted to postmenopausal women, current smokers conducted the largest case-control and colleagues (2009) had an elevated, albeit not significant, risk (RR = 1.20; study to date, with 6,235 cases and 6,533 controls (Table - 95% CI, 0.52–2.80). The study included too few premeno 221 Cancer

251 Surgeon General’s Report one of only a few with sufficient statistical power to detect ). The study was based on pooled data from two case- 6.15 S control studies in Ontario, Canada: the Ontario Women’s associations among premenopausal women. - Only two studies reported results that were strati Health Study and the Ontario Women’s Diet and Health Study. The designs of the two studies were similar, with fied by race/ethnicity. In one, Mechanic and colleagues cases ascertained through the provincial cancer registry (2006) provided data from Phases I and II of the Carolina Breast Cancer Study, a study that examined former and and controls randomly selected from a population-based current smoking among 894 African American and 1,414 listing or by random-digit dialing. A risk estimate of 1.10 non-Hispanic White women. These cases were ascertained (95% CI, 0.98–1.23) was reported for current smokers ver - through the North Carolina Central Cancer Registry, and sus women with no history of active or passive smoking. A significantly increased risk was found for older age at population-based controls (n = 2,022) were selected from smoking initiation (≥26 years vs. a no active/no passive motor vehicle and Health Care Financing Administra - group) (OR = 1.26; 95% CI, 1.03–1.55), but there were tion (now the Centers for Medicare & Medicaid Services) listings. This report serves as an update to the study by no associations at younger ages of initiation (<12 years: OR = 0.88; 95% CI, 0.59–1.31; 12–15 years: OR = 1.02; 95% Marcus and colleagues (2000), which provided age and race-adjusted estimates from Phase I. In the study by CI, 0.90–1.16; 16–20 years: OR = 1.12; 95% CI, 1.01–1.24). There was a significant risk of breast cancer for smoking Mechanic and colleagues (2006), risk for breast cancer was significantly increased in African American women who initiated more than 5 years before first birth (OR = 1.16; were former smokers (OR = 1.80; 95% CI, 1.30–2.50) or 95% CI, 1.04–1.31), and for smoking initiated after first birth (OR = 1.24; 95% CI, 1.02–1.52). These results do not who had smoked more than 20 years (OR = 1.80; 95% CI, 1.20–2.60). In contrast, risk was not significantly elevated support the hypothesis that early initiation of smoking and smoking before first birth are more strongly associ for White women who were former smokers (OR = 1.20; - ated with risk of breast cancer than are later initiation and 95% CI, 0.90–1.50) or who had smoked for more than 20 years (OR = 1.10; 95% CI, 0.90–1.50). Increased risk initiation of smoking after first birth. was not significantly associated with current smoking for Johnson and colleagues (2000), in a study in eight Canadian provinces, ascertained 2,317 cases through the either racial group. - Slattery and colleagues (2008) conducted a popula provincial tumor registries in the mid-1990s. Controls - tion-based case-control study in Arizona, Colorado, New (2,438) were randomly sampled from health plan list Mexico, and Utah. This study provided data on the risk of ings, a property assessment database, or by random-digit breast cancer associated with smoking status, pack-years - dialing. Extensive data were collected via a mailed ques of smoking, age at smoking initiation, and smoking before tionnaire on active smoking and exposure to secondhand first pregnancy. Among women with a first primary breast smoke. The analyses of cigarette smoking status used two cancer who had data for smoking, 798 were Hispanic/ reference groups: never smoker and no active/no passive American Indian and 1,527 were non-Hispanic White. - exposure. Only the no active/no passive exposure refer ent was used for age at smoking initiation, number of Cases were ascertained from state or national cancer registries (e.g., NCI’s SEER Program). Population-based cigarettes smoked per day, duration of smoking, pack- years of smoking, and number of years since quitting controls were randomly sampled, of which 924 Hispanics/ smoking. In general, risk estimates were higher when American Indians and 1,601 non-Hispanic Whites had data for smoking. Among premenopausal non-Hispanic White using the no active/no passive referent group than when using the never smoker referent group. Among premeno - women, risk for breast cancer was significantly increased among ever smokers (OR = 1.3; 95% CI, 1.0–1.7), those pausal women, adjusted estimates (using no active/no who smoked before first pregnancy (OR = 1.4; 95% CI, passive as the referent) were higher for former smokers (OR = 2.6; 95% CI, 1.3–5.3) than for current smokers 1.0–1.9), and those who accumulated more than 15 pack- years of smoking (OR = 1.6; 95% CI, 1.1–2.4). The study (OR = 1.9; 95% CI, 0.9–3.8); estimates for postmeno - pausal women were marginally higher for current smok- did not find any significant associations with breast cancer ers (OR = 1.6; 95% CI, 1.0–2.5) than for former smokers in premenopausal Hispanic and American Indian women or in postmenopausal non-Hispanic White or Hispanic/ (OR = 1.4; 95% CI, 0.9–2.1). As for other measures of smoking (using no active/no passive exposure as the ref - American Indian women. erent), premenopausal women had risk estimates at least Results from the three remaining large case-control 20% higher than postmenopausal women for current and studies are inconsistent. Gammon and colleagues (2004a), former smoking status, age at smoking initiation, num- who reported results from the Long Island Breast Cancer ber of cigarettes smoked per day, duration of smoking, Study Project for 1,356 cases and 1,383 population-based - controls, found that risk for breast cancer was not signifi and number of years since quitting smoking. The study oversampled women younger than 55 years of age, so it is cantly increased among active/current smokers using a Chapter 6 222

252 —50 Years of Progress The Health Consequences of Smoking analyzed data from the Massachusetts and Wisconsin sites no active/no passive exposure referent regardless of the (791 cases, 797 controls) for effect modification of smoking number of cigarettes per day, pack-years of smoking, age risk by at smoking initiation, or smoking before first pregnancy. NAT2 genotype. Not accounting for genotype, this study found a significantly increased risk for ever smokers Significant associations were not found in a variety of sub - groups, even after stratifying by menopausal status, BMI, (OR = 1.37; 95% CI, 1.12–1.69) and for women with more alcohol use, use of HRT, use of oral contraceptives, family than 25 pack-years of smoking (OR = 1.54; 95% CI, 0.87– history, and age at reference date. In Los Angeles, Prescott 2.71). Results for the latter variable were OR = 1.54 (95% and colleagues (2007), who conducted a case-control study CI, 0.87–2.71) for premenopausal women and OR = 1.53 (95% CI, 1.10–2.13) for postmenopausal women. In a of 1,728 cases and 441 controls, did not find significant subsequent report, Ahern and colleagues (2009) analyzed associations between risk for breast cancer and smoking status, duration of smoking, age at smoking initiation, or data from only the Massachusetts site in the Collaborative smoking before first pregnancy. In contrast, Band and col Breast Cancer Study (557 cases, 432 controls) but did not - find an association between pack-years of active smoking leagues (2002) reported significant associations between (OR = 0.90; 95% CI, 0.7–1.3 for >23 pack-years) and risk risk for breast cancer and ever smoking (OR = 1.50; 95% CI, 1.09–2.07) and smoking for at least 20 years or more of breast cancer. However, this report was focused mainly (OR = 1.60; 95% CI, 1.08–2.37) in premenopausal but not on effects or associations with passive smoking. The study by Rollison and colleagues (2008) reported postmenopausal women based on responses from 1,018 an increased risk for breast cancer among ever smokers cases and 1,025 controls who participated in a study con - (OR = 1.43, 95% CI, 1.03–1.99). The authors attempted ducted in British Columbia, Canada. There were no sig - to compare results based on a no active/no passive to a nificant associations between risk and age at smoking no active-only reference group but the sample size was initiation, but smoking before first pregnancy was sig - nificant for premenopausal women (OR = 1.51; 95% CI, too small to provide sufficient statistical power to make an evaluation. Brown and colleagues (2010) conducted a 1.07−2.13) but not for postmenopausal women. case-control study of risk factors for breast cancer among Six additional but smaller studies (<1,000 cases) that were conducted in the United States are notable for Asians (Chinese, Filipino, Japanese) who immigrated to - San Francisco-Oakland, California; Los Angeles, Califor their findings (Lash and Aschengrau 2002; Egan et al. 2003; Li et al. 2005; Rollison et al. 2008; Ahern et al. 2009; nia; or Oahu, Hawaii. Just over one-half of the women in the study (54% of cases; 58% of controls) were born in Brown et al. 2010). In one, Li and colleagues (2005) exam - Asia (China, Taiwan, Hong Kong, Macau, Japan, the Phil ined a sample of 975 cases and 1,007 controls in Wash - - ington state and found a significantly increased risk (30% ippines, Southeast Asia, the Malaysian Peninsula, Singa- in each instance) for breast cancer among ever smokers, pore, or India) as opposed to Western or Western-style countries (such as those in North America or Europe or those who smoked, those 20−39 years of age, those who started smoking before age 20, and those who smoked the nations of Australia and New Zealand). Women born in Asia and more recent migrants (<8 years) to the West had before their first full-term birth. In addition, women who a lower risk of breast cancer regardless of smoking history reported 20 or more pack-years of smoking and a history of than women born and raised in the West or a Western-style HRT involving both estrogen and progestin had increased country. The overall OR for ever smoking was 1.2 (95% CI, risk for breast cancer. The study by Lash and Aschengrau 0.9–1.6). The only significant association between smok (2002) stands out because it found a significant inverse - ing and breast cancer was for age at initiation of younger association for ever smoking (OR = 0.72; 95% CI, 0.55– than 16 years of age (OR = 2.92; 95% CI, 1.1–7.9), but this 0.95). That 2002 study conflicts, however, with a 1999 was based on a very small stratum (11 cases, 9 controls). study (OR = 2.0; 95% CI, 1.1–3.6) in the same geographic area of Cape Cod, Massachusetts, carried out by the same European Studies team (Lash and Aschengrau 1999). Both studies included deceased cases and controls for which information about Since 2000, three large (>1,000 cases) popula - smoking was collected from proxies. However, the 2002 tion based case-control studies have been conducted in study, unlike the 1999 study, did not provide information Europe: one each in Germany (Andonova et al. 2010; Rab - about the fraction of data collected from proxy respon- stein et al. 2010), Sweden (Magnusson et al. 2007), and dents. Thus, the results of the 2002 study could have been Poland (Lissowska et al. 2006). Andonova and colleagues affected by information bias. - (2010) reported results from the Gene Environment Inter In a report from the Collaborative Breast Cancer action and Breast Cancer in Germany (GENICA) study Study, a population-based study conducted in Maine, that included estimates of risk for breast cancer for for - Massachusetts, New Hampshire, and Wisconsin between mer (OR = 0.95; 95% CI, 0.75–1.19) and current (OR = 1988–1991 (Baron et al. 1996), Egan and colleagues (2003) 223 Cancer

253 Surgeon General’s Report 0.84; 95% CI, 0.66–1.06) smoking using data for 1,021 for women with no passive exposure, and it decreased a bit more for women reporting passive exposure (OR = 1.09; cases and 1,015 controls in the greater Bonn region. This report and a companion report (Rabstein et al. 2010) were 95% CI, 0.75–1.56). Interpreting the importance of restricted to those subjects with available DNA and are the differences among the various estimates is difficult further described in the section on genetic susceptibility. - because none are statistically significant and the CIs over lap. Kropp and Chang-Claude (2002) evaluated the same Magnusson and colleagues (2007) ascertained 3,345 smoking measures with a no active/no passive reference cases from six regional cancer registries and randomly - selected 3,454 controls from a governmental popula group. Their estimate for former smokers was comparable tion listing in Sweden. The study found few significant to that of Roddam and colleagues (2007) but was consid - or consistent associations between risk of breast cancer erably higher for current smokers (OR = 1.47; 95% CI, 0.99–2.20). Last, Cerne and colleagues (2011) reported and current or former smoking, duration of smoking, results from a clinic-based case-control study of breast pack-years of smoking, age at initiation of smoking, or cancer among 784 cases and 709 controls among post smoking before first full-term birth. However, data were - missing for nearly 17% of cases and 25% of controls. The menopausal Slovenian women. This report was focused higher rate of missing data in controls was due partly on the effects of HRT, but an estimate was provided for to the use of a telephone interview that did not collect smoking at least 10 cigarettes per day, adjusting for age data for alcohol consumption and perhaps other covari - and education only (OR = 1.70; 95% CI, 1.20–2.43). ates. In the Polish study, Lissowska and colleagues (2006) Notably, the reference group of nonsmokers included former smokers. did not obtain significant results for all women consid - ered together. However, among women younger than 45 Asian Studies years of age (n = 511), significantly increased risks were observed for current smoking (OR = 2.03; 95% CI, 1.40– - Two small case-control studies from Asia were pub 2.95), the highest level of duration of smoking (>20 years: lished between 2000 and 2011. For ever smoking, the OR = 2.33; 95% CI, 1.32–4.13), smoking before first preg- study conducted in Manila, the Philippines (Gibson et al. nancy (OR = 2.03; 95% CI, 1.40–2.94), and ever smoking 2010), reported an RR of 1.3 (95% CI, 0.6–2.9), and a study approached statistical significance (OR = 2.40; 95% CI, in northeast India (Kaushal et al. 2010) reported an RR of - 1.00–5.72). It is difficult to interpret some of these asso 1.15 (95% CI, 0.62–2.13). ciations due to conflicting findings across levels of these - exposures; for example, risks were also increased for dura Adjustment for Selected Covariates tion of smoking less than 10 years (OR = 1.57; 95% CI, Breast cancer is recognized as a heterogeneous dis - 1.01–2.44) and for smoking after first pregnancy (OR = ease with many associated risk factors (Hankinson and 2.40; 95% CI, 1.27–4.53). Kruk (2007) also conducted a Hunter 2001; Brinton et al. 2002; Spicer and Pike 2005; clinic-based study in Poland in which the control group Hortobagyi et al. 2006). Some of these risk factors have was characterized by a higher prevalence of smoking than complex relationships with cancer of the breast, and those in the general population. This study found some of the highest significant risks to date among women who the direction of their associations may differ according smoked 10 or more cigarettes per day. Here, the ORs were to characteristics such as breast cancer phenotype, age, menopausal status, and race/ethnicity. Established risk 2.55 (95% CI, 1.81–3.60) for premenopausal women and factors include: 1.78 (95% CI, 1.33–2.37) for postmenopausal women. - In England, Roddam and colleagues (2007) con increasing age; • ducted a population-based study of 639 cases, 36–45 years of age, with 640 age-matched controls. Significant asso - • family history of breast cancer in first-degree rela- ciations were not detected for former smokers (OR = 1.15; tives; 95% CI, 0.87–1.53) or current smokers (OR = 1.04; 95% CI, 0.79–1.36), age at initiation of smoking, duration of • increased levels of endogenous estrogen; smoking, or number of cigarettes smoked per day. Data for duration of smoking and age at smoking initiation were history of benign breast disease; • analyzed as continuous variables. Thus, the results were not combined with those from other studies in generating • mammographically dense breasts; summary estimates. The OR for former smokers, when calculated using a no passive/no active exposure reference • less frequent screening; group, was slightly lower (OR = 1.12; 95% CI, 0.72–1.73) 224 Chapter 6

254 The Health Consequences of Smoking —50 Years of Progress estrogen levels, there was an apparent positive trend with history of ionizing radiation exposure to the chest; • increasing alcohol consumption (Kendall et al. 2007). All • various reproduction-related factors—increased cohort studies described in this report adjusted for at least risk with younger age at menarche (<12 years of one reproductive factor and BMI; most of them either adjusted for or stratified on menopausal status; and all but age), older age at menopause (>54 years of age), older age at first pregnancy or live birth (>30 years . (Lawlor et al. 2004) one adjusted for alcohol consumption ) did not adjust for fam- S of age), no history of breast feeding or a short dura - Three cohort studies (Table 6.14 tion of lactation, nulliparity, and decreased risk with (Manjer et al. 2000b; Lawlor et al. 2004; Gram ily history increased number of pregnancies; et al. 2005) . The selection of covariates for adjustment varied S ). Some studies higher socioeconomic status (e.g. higher level of • across case-control studies (Table 6.15 (Delfino et al. 2000; did not adjust for reproductive factors education and/or family income); , alco- Alberg et al. 2004; Li et al. 2005; Metsola et al. 2005) use of exogenous hormones (HRT, combined estro (Delfino et al. 2000; Zheng et al. 2002b; van der hol intake • - gen/progesterone oral contraceptives); and , Hel et al. 2003b; Alberg et al. 2004; Metsola et al. 2005) (Delfino et al. 2000; van der Hel et al. 2003b; body size increased body size among postmenopausal women • Alberg et al. 2004; Metsola et al. 2005; Mechanic et al. (as determined by height, weight, BMI, waist cir - 2006; Prescott et al. 2007) , or family history (Johnson et al. cumference, waist/hip ratio). 2000; van der Hel et al. 2003b; Alberg et al. 2004; Li et al. 2005; Metsola et al. 2005; Slattery et al. 2008) . Five case- control studies did not adjust, stratify, or match on meno Studies have also demonstrated a modestly increased - pausal status, but in these studies the age range included risk for breast cancer, on the order of 25–30%, associated (Marcus with low level of physical activity (Friedenreich and Cust both premenopausal and postmenopausal women 2008) and on the order of nearly 50% with intake of 45 or et al. 2000; Lash and Aschengrau 2002; Alberg et al. 2004; Metsola et al. 2005; Magnusson et al. 2007) . Several stud- more grams of alcohol per day (Collaborative Group on Hormonal Factors in Breast Cancer et al. 2002; Baan et al. ies explored models that adjusted for multiple covariates but reported results for only the most parsimonious mod 2007). IARC (2002) has concluded that alcohol consump - - els, adjusting for covariates that changed point estimates tion is a causal risk factor for breast cancer; additionally, on the order of 5–15% (Marcus et al. 2000; van der Hel Volume 6 of the IARC Handbook on Cancer Prevention concluded that regular physical activity reduces the risk et al. 2003b; Gammon et al. 2004a; Li et al. 2005; Lis - of breast cancer. Many of these factors show a complex - sowska et al. 2006; Mechanic et al. 2006; Kruk 2007; Mag nusson et al. 2007; Rollison et al. 2008; Young et al. 2009) . pattern of association that depends on timing in relation to other exposures, specifically increased estrogen levels, Most studies with findings that were considered for inclu- sion in the meta-analyses made an effort to statistically duration of exposure, and menopause. Differences in the distributions of these factors between women who smoke - detect and adjust for confounders within the data. How and those with no history of active smoking are likely to ever, the methods for considering potential confounders vary across populations; to the extent possible, the poten - varied across studies and the basis for selecting the final, tial for confounding has been considered in individual adjusted model was not always explicit. studies and in the meta-analyses. The great majority of cohort and case-control studies published since 2000 and described in this report (Tables Meta-Analysis of Breast Cancer and 6.15 6.14 ) either adjusted for, or evaluated the S S Risk Associated with Measures of - need for adjustment of, relevant confounders. Reproduc tive factors and family history are well-established, strong Active Smoking risk factors for breast cancer (Spicer and Pike 2005) . In All available non-overlapping cohort study reports addition, since 2000 an increasing number of studies have published prior to 2012 and case-control study reports demonstrated that alcohol use and obesity are important published from 2000–2011 were included in meta-analyses - (Collaborative Group on Hor risk factors for breast cancer for this report. These timeframes were selected to identify . In a review monal Factors in Breast Cancer et al. 2002) the most recent evidence that was specifically relevant to by Kendall and colleagues (2007), the authors found that associations between risk for breast cancer and active and higher BMI is associated with increased endogenous estra- passive smoking. The older literature has been repeatedly diol levels among postmenopausal women. Although they reviewed; the majority of studies published before 2000 did not find a clear relationship between alcohol use and 225 Cancer

255 Surgeon General’s Report were either cross-sectional or case-control in design and - differences between studies are the result of random vari were not considered for inclusion in the meta-analysis. ation (precision) within studies, a random-effects model Reports from cohort studies published prior to 2000 were - assumes that between-study variation is partly due to fac tors that influence the magnitude of the effect within true evaluated for inclusion; most of these have been super - each study, resulting in a distribution of true effects across seded by subsequent reports. Table 6.16 S provides a list- studies. The fixed-effects model gives greater weight to ing of the 65 reports from case-control and cohort studies. larger, more precise studies, whereas the random-effects Twenty-six reports overlapped with results on the same study population, and of these, 11 were included in the model dampens to some degree the influence of these meta-analyses because they were either the most recent or larger studies relative to smaller ones. Additionally, the summary estimates from random-effects models gener complete reports from their study. In the case of 1 cohort - study (NHS-I) and 1 case-control study (Collaborative ally have broader CIs than those from fixed-effects models, Breast Cancer Study), 2 reports contributed to separate - making the former method intrinsically more conserva meta-analyses because they offered different measures tive (Borenstein et al. 2009). The random-effects model (NHS-I: Egan et al. 2002 and Xue et al. 2011; Collabora accounts for heterogeneity among studies, which can be - - quantified, for example, in the Q-test statistic. When het tive Breast Cancer Study: Egan et al. 2003 and Ahern et erogeneity is low, the random-effects model converges al. 2009). Three cohort studies (Mills et al. 1989b; Land et with the fixed-effects model. al. 1994; Thomas et al. 1997), which were included in the Meta-analyses were conducted in STATA 11.0 report by the Collaborative Group on Hormonal Factors in (STATA Corp., College Station, TX, USA) using the meta Breast Cancer and colleagues (2002), were excluded from (Sterne 2009) the present report because the individual estimates were STATA command . The meta-funnel STATA not published in the original reports and they were com - command was used to create funnel plots for visual assess - ment of publication bias and outliers. Between-study het- bined into an ‘other’ category for the Collaborative Report. 2 test, reported χ Four studies (1 cohort, 3 case-control) were included in erogeneity was assessed with Cochran’s as the Q-test statistic, and bias was assessed formally using only the meta-analysis of smoking before a first full-term Egger’s statistical test pregnancy or first birth (Innes and Byers 2001; Fink and (Egger et al. 1997) and Begg’s rank . Thus, correlation test (Begg and Mazumdar 1994) Lash 2003; Lawlor et al. 2004; DeRoo et al. 2011a) , with the a total of 46 separate reports were included in the initial latter calculated via the metabias STATA command. The analysis of ever smoking. The total number included in Begg test is reported to have low power when the number each subsequent meta-analysis depended on whether of studies is small. The Egger test is more powerful but (Deeks a risk estimate was reported in a study for the measure also biased and can produce false-positive results . Sensitivity analyses considered study design, et al. 2005) of smoking. RR estimates were pooled across categories of exposure to fit common definitions of ever smoking, prevalence of exposure, sample size, and measurement smoking status (former or current), duration of smoking, of exposure effect. Results for the Begg and Egger tests cigarettes smoked per day, pack-years of smoking, age at are included as a note in figures as appropriate. Summary smoking initiation, and smoking before first pregnancy. estimates from random effects models are reported for Data are provided in Table 6.16 on studies affected by S all meta-analyses. design and analysis issues, including small sample size, a Ever Smoking mixed reference group (former smokers and nonsmokers combined), inadequate covariate adjustment, use of proxy If not reported, a measure for ever smoking was subject reports, issues associated with exposure category calculated for all 46 studies by pooling available data on cutpoints, and the presence of extreme outliers. smoking status, smoking duration, cigarettes smoked The DerSimonian and Laird (1986) procedure for per day, or pack-years of smoking, with the exception of - random-effects meta-analysis was used to calculate sum four studies that provided data only for exposure before or mary estimates. The random-effects model was selected ). A meta-analysis was S during first pregnancy (Table 6.16 because the studies included in the meta-analysis showed conducted of nonoverlapping reports from all cohort stud - substantial variation in type and quality of design, time ies through 2011, as well as case-control studies published period, geographic setting, composition of population, from 2000–2011, for ever smoking, resulting in a sum- ascertainment of cases, selection of controls for case- mary estimate with significant heterogeneity (p <0.001): h - control studies, and definition and measurement of smok RR = 1.12 (95% CI, 1.07–1.17; n = 46) (Table 6.17 , Figure S ing exposure. Whereas a fixed-effects model assumes that 6.28). From visual inspection, the funnel plot in Figure all studies are estimating the same true effect and that 6.29 shows no sign of skewness, indicating that publica - Chapter 6 226

256 —50 Years of Progress The Health Consequences of Smoking Figure 6.28 Forest plot showing association between ever smoking and risk for breast cancer, based on cohort stud - ies published before 2012 and case-control studies published from 2000 to 2011 (n = 46) * = cohort study; ^ = case-control study. Meta-analysis RR = 1.12 (95% CI, 1.07–1.17); Begg z = 0.48, p = 0.63; Egger bias = Note: 0.44, p = 0.25. See Table 6.17 S ( note a ) for studies excluded. Size of square is proportional to the weights used in the meta-analysis; error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and RR = relative risk. = Million Women Study Collaborative Group; associated 95% CI. MWSCG 227 Cancer

257 Surgeon General’s Report Funnel plot for estimates in meta-analysis of ever smoking with risk for breast cancer, based on cohort Figure 6.29 studies published before 2012 and case-control studies published from 2000 to 2011 (n = 46) Note: = cohort study; = case-control study. Includes the same studies reported in Figure 6.28. p l 2000; Alberg et al. 2004; Gibson et al. 2010; Kaushal - tion bias was not a significant issue. This finding was fur et al. 2010), with less than 210 cases. ther confirmed by Begg’s rank correlation test (z = 0.48, - p = 0.63) and the Egger test (bias = 0.44, p = 0.25). Strati fication by study design revealed that the heterogeneity Two additional studies, one cohort (Vatten and 3. was due primarily to variation among the 27 case-control Kvinnsland 1990) and one case-control (Cerne et al. <0.001) than to studies (RR = 1.15; 95% CI, 1.06–1.25; p 2011), with an estimate reported for only current h variation among the 19 cohort studies (RR = 1.10; 95% CI, smokers and for which the reference group appeared = 0.793). to mix never smokers with former smokers. 1.07–1.13; p h Thirteen studies were excluded in the following sequence (some studies fell into more than one category). The summary estimate for the 12 cohort studies remaining (Table 6.17 S ) after the exclusion of the 7 stud- ies that were restricted to nondrinkers had a small sample, Six cohort studies reported in the pooled analysis 1. restricted to nondrinkers conducted by the Collab - or a mixed reference group did not change meaningfully from the overall estimate (RR = 1.10; 95% CI, 1.07–1.13; orative Group on Hormonal Factors in Breast Can- cer and colleagues (2002) and for which there were p = 0.717). For case-control studies, the RR was attenu - h no data available on smoking in the original report <0.001) ated slightly (RR = 1.13; 95% CI, 1.04–1.23; p h (van den Brandt et al. 1995; Engeland et al. 1996; when 6 were excluded that were either small (<210 cases), S Million Women Study Collaborative Group 1999). from Asia, or had a mixed reference group (Table 6.16 ). The additional exclusion of a cohort study (Nordlund et al. 1997) that adjusted only for age and place of residence Eight additional studies, three cohort (Schatzkin et 2. did not alter the summary RR for cohort studies. The fun- al. 1989; Hanaoka et al. 2005; Lin et al. 2008) and nel plot in Figure 6.29 indicates that the studies by Kruk five case-control (Delfino et al. 2000; Morabia et al. 228 Chapter 6

258 The Health Consequences of Smoking —50 Years of Progress (2007) and Lash and et al. 2000; Egan et al. 2002; Reynolds et al. 2004b; Gram are outliers. The Aschengrau (2002) Aschengrau (2002) relied et al. 2005; Hanaoka et al. 2005; Lissowska et al. 2006; case-control study by Lash and Roddam et al. 2007; Rollison et al. 2008; Ahern et al. 2009; on proxy interviews for deceased cases. Kruk (2007), which Luo et al. 2011b). Six compared estimates using the two was conducted in Poland, used clinic-based controls that were reported to have a higher percentage of smoking referent groups by smoking status (Johnson et al. 2000; (33%) than in the general population (23%). However, a Egan et al. 2002; Reynolds et al. 2004b; Hanaoka et al. comparison of self-reported prevalence of cigarette smok- 2005; Roddam et al. 2007; Luo et al. 2011b), 1 did so by - - pack-years (Ahern et al. 2009), and 2 provided compari ing and cotinine saliva samples (cutpoint for active smok ing—1.5 nanogram [ng]/milliliter [mL]) indicated that sons by duration, dose, and timing (Rollison et al. 2008; Luo et al. 2011b). Nine studies used only a no active/no true prevalence may be underestimated in Poland by 4.4% passive reference group (Delfino et al. 2000; Morabia et al. (West et al. 2007) . The removal of Kruk (2007) and Lash - and Aschengrau (2002) resulted in a summary risk esti- 2000; Kropp and Chang-Claude 2002; Lash and Aschen mate of 1.08 (95% CI, 1.03–1.13) and decreased heteroge- grau 2002; Alberg et al. 2004; Gammon et al. 2004a; Sil - = 0.340) for case-control studies, without adding neity (p lanpaa et al. 2005a; Mechanic et al. 2006; Young et al. h significant bias according to the Begg (z = 0.73, p = 0.46) 2009). As noted previously, estimates for ever smoking and Egger (bias = 0.43, p = 0.19) tests (see notes for Figure were derived for some studies by pooling other exposure measures, such as former and current smoking. Addition - 6.30). The RR for the combined case-control and cohort ally, the terminology for defining these reference groups studies (n = 30) decreased to 1.09 (95% CI, 1.06–1.12; = 0.500). In summary, the significant heterogeneity p - (no active-only, no active/no passive) varies among stud h among studies for the association between ever smoking ies, although the definitions are common. and breast cancer is attributable mainly to the study by The size of the reference group is greatly decreased Kruk (2007), which is the more extreme of the two out - when restricted to no active/no passive exposure because liers. Excluding this study substantially reduces hetero- of the high prevalence of passive smoking exposure: most geneity and results in an attenuated summary estimate. studies indicate that only about 10−20% of never smok - When taken together, these 30 studies suggest that ever ers report no passive exposure. In a study by Arheart - smoking increases the RR for breast cancer by a statisti and colleagues (2008), an estimated 28% of people who , Figure 6.30). cally significant average of 9% (Table 6.17 S reported no passive exposure were actually exposed based These 30 reports remained as the baseline to be consid - on serum cotinine levels, suggesting that the true no ered for the remaining meta-analyses. active/no passive group may be even smaller, particularly if considered in a lifetime context. No systematic analyses No Active-Only Versus No Active/No Passive have been conducted to determine whether using only a Exposure Referent Group small no active/no passive referent produces selection bias or sparse data bias (Greenland et al. 2000) as well as loss - Wells (1991) first suggested that the most appro of statistical power, or whether statistical adjustment for priate reference group would exclude women who were passive smoking exposure in assessing active smoking is exposed to passive smoke because their inclusion would as efficient as having a no active/no passive referent. One attenuate the association with active smoking. Mora - - exception may be Ahern and colleagues (2009), who esti bia and colleagues (1996) first used this criterion in an mated associations of active smoking with breast cancer - analysis of data from a case-control study in Switzer using a restricted no active/no passive exposure refer - land. Since then, other investigators have narrowed the ence group while also employing statistical adjustment definition of the reference group to women who report for passive smoking exposure. Unfortunately, it is difficult no active or passive smoking exposure. In this report, 5 to interpret the differences between the two approaches cohort studies (Egan et al. 2002; Reynolds et al. 2004b; because only 30% of participants in that study had data for Gram et al. 2005; Hanaoka et al. 2005; Luo et al. 2011b) both active and passive smoking. and 14 case-control studies (Morabia et al. 2000; Delfino In the California Teachers Study cohort (Table et al. 2000; Johnson et al. 2000; Kropp and Chang-Claude S ), the RRs for breast cancer in current smok- 6.14 2002; Lash and Aschengrau 2002; Alberg et al. 2004; Gam - ers overall were both significant and quite similar with mon et al. 2004a; Sillanpaa et al. 2005a; Lissowska et al. the two reference groups used: no active-only (“never”) 2006; Mechanic et al. 2006; Roddam et al. 2007; Rollison (RR = 1.32; 95% CI, 1.10–1.57) and no active/no passive et al. 2008; Ahern et al. 2009; Young et al. 2009) included (RR = 1.25; 95% CI, 1.02–1.53) (Reynolds et al. 2004b). results based on a no active/no passive exposure reference In contrast, ORs for ever smokers (i.e., former or cur - group. Ten studies reported results for both reference rent) in Johnson and colleagues’ (2000) population-based groups that can be compared for ever smoking (Johnson 229 Cancer

259 Surgeon General’s Report Figure 6.30 Forest plot showing association between ever smoking and risk for breast cancer, based on cohort stud - ies published before 2012 and case-control studies published from 2000 to 2011, excluding studies with design or analysis issues (n = 30) * = cohort study; ^ = case-control study. Meta-analysis RR = 1.09 (95% CI, 1.06–1.12); Begg z = 0.73, p = 0.46; Egger bias = Note: ) for studies excluded. Size of square is proportional to the weights used in the meta-analysis; 0.43, p = 0.19. See Table 6.17 S ( note c error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and RR = confidence interval; = relative risk. associated 95% CI. CI 230 Chapter 6

260 The Health Consequences of Smoking —50 Years of Progress study population to determine whether there were any Canadian case-control study were 1.0 (95% CI, 0.8–1.3) differences for other potential confounders such as race/ for premenopausal women and 1.2 (95% CI, 1.0–1.4) for postmenopausal women when based on the no active-only ethnicity, education, alcohol consumption, or reproduc- (“never”) reference group, versus 2.3 (95% CI, 1.2–4.5) tive variables. This comparison, in fact, was not made in any of the studies that used a no active/no passive expo - for premenopausal women and 1.5 (95% CI, 1.0–2.3) for postmenopausal women when based on the no active/no sure reference group. - Meta-analyses were conducted to compare 27 stud passive exposure reference group. Although these results seem to suggest a strong effect when using a no active/ ies reporting results based on a no active-only reference no passive exposure reference group, the estimates were group with 15 studies reporting estimates based on a no active/no passive exposure reference group (Table 6.16 S ), based on a restricted subgroup of women (62% of the ref - after the 13 exclusions cited previously. The number of erence group) who were able to account for and report studies was further reduced to 25 for the no active-only data for more than 90% of their lifetime residential passive and 14 for the no active/no passive exposure analyses smoking exposure. In addition, the no active/no passive with the exclusion of 3 studies (Nordlund et al. 1997; reference group consisted of only 193 women (49 pre - Lash and Aschengrau 2002; Kruk 2007) for reasons given menopausal and 144 postmenopausal women), compared above. The report by Egan and colleagues (2002) was used with 2,292 women in the no active-only reference group. because the more recent report by Xue and colleagues Only two case-control studies have compared results for measures of smoking other than ever smoking (2011) did not report results using a no active/no passive exposure reference group. The RR for the no active-only - or smoking status, but the results are difficult to inter pret because of small samples and low statistical power exposure reference group was 1.09 (95% CI, 1.06–1.13; = 0.308) (Table 6.17 , Figure 6.31). This estimate is S - (Rollison et al. 2008; Ahern et al. 2009). For cohort stud p h slightly lower than that calculated for 14 studies using a ies, Lin and colleagues (2008) compared results using no active/no passive exposure reference group (RR = 1.15; the two different definitions of reference groups (no S 95% CI, 1.09–1.21; p = 0.572) (Table 6.17 , Figure 6.32). active/no passive, no active-only) in the Japan Collab- h Nine of the studies—4 of which were large cohort stud - orative Cohort Study for Evaluation of Cancer Risk and ies (Egan et al. 2002; Reynolds et al. 2004b; Gram et al. stated there was no difference in the estimates, but they 2005; Luo et al. 2011b)—calculated estimates using both did not provide numerical evidence. Luo and colleagues reference groups. These 9 studies were included in the two (2011b) reported findings for the only cohort study to meta-analyses. Neither of these analyses was significantly date with parallel, multivariable adjusted analyses con- affected by publication or small-study bias, according to trasting no active/no passive exposure with no active-only Begg or Egger statistics (see notes for Figures 6.31 and reference groups for multiple measures. The use of a no 6.32; funnel plots not shown). These analyses suggest that active/no passive exposure reference group resulted in a - the use of a restricted no active/no passive exposure refer small but consistent increase in RR ranging from 2–10% ence group results in a small increase in estimates of the for most measures of active smoking (ever, status, age association between ever smoking and breast cancer. at initiation, duration, cigarettes smoked per day, pack- years). The strongest effect of active smoking was for Cigarette Smoking Status duration greater than 50 years, where the RR was 1.45 (95% CI, 1.06–1.98) using a no active/no passive exposure A total of 25 studies reported estimates for current group compared with 1.35 (95% CI, 1.03–1.77) using a no and former smoking; 20 used a no active-only and 5 a no active-only (“never”) reference group. The analysis sug - , S active/no passive exposure reference group (Table 6.16 gests that the use of a no active/no passive exposure ref- Figures 6.33 and 6.34). The summary estimates were simi - erence group may provide a small benefit in control for lar for current smokers (RR = 1.12; 95% CI, 1.08–1.16; confounding between active and passive smoking effects. = 0.347) and former smokers (RR = 1.09; 95% CI, 1.05– p h However, this small gain in control of confounding is at = 0.062) (Table 6.17 S 1.13; p ). Results for former smokers h the cost of statistical power. It has not been established were virtually identical for the two study designs: cohort - whether statistical adjustment for passive exposure of esti (RR = 1.09; 95% CI, 1.03–1.14; p = 0.021) and case- h mates for the risk of active smoking adequately controls = 0.354). The control (RR = 1.09; 95% CI, 1.03–1.16; p h for this confounding. Additionally, the small, restricted summary estimate for current smokers in the 11 cohort subgroup with no active/no passive exposure could dif- studies (OR = 1.14; 95% CI, 1.10–1.18; p = 0.746) was h fer systematically for other confounders or modifiers that higher than the estimate for those in the 14 case-control are not measured or adequately controlled. Luo and col- studies (OR = 1.07; 95% CI, 1.00–1.16; p = 0.209). Sen- h - leagues (2011b) did not systematically compare the sub sitivity analyses were conducted that excluded the 4 case- group of no active/no passive smokers with the rest of the control studies (Kropp and Chang-Claude 2002; Gammon 231 Cancer

261 Surgeon General’s Report Figure 6.31 Forest plot showing association between ever smoking and risk for breast cancer, based on the subset of cohort studies published before 2012 and case-control studies published from 2000 to 2011 with a no active-only referent group (n = 25) Note: * = cohort study; ^ = case-control study. Meta-analysis RR = 1.09 (95% CI, 1.06–1.13); Begg z = 0.70, p = 0.48; Egger bias = ) for studies excluded. Size of square is proportional to the weights used in the meta-analysis; note d 0.43, p = 0.34. See Table 6.17 ( S error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and associated 95% CI. CI = confidence interval; RR = relative risk. 232 Chapter 6

262 The Health Consequences of Smoking —50 Years of Progress Forest plot showing association between ever smoking and risk for breast cancer, based on the subset of Figure 6.32 cohort studies published before 2012 and case-control studies published from 2000 to 2011 with a no active/no passive exposure referent group (n = 14) * = cohort study; ^ = case-control study. Meta-analysis RR = 1.15 (95% CI, 1.09–1.21); Begg z = 0.05, p = 0.96; Egger bias = Note: 0.04, p = 0.94. See Table 6.17 S ( note e ) for studies excluded. There were nine studies with estimates reported for both a no active- only and a no active/no passive reference group (also shown in Figure 6.31). Size of square is proportional to the weights used in the meta-analysis; error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary CI = confidence interval; RR = relative risk. estimate and associated 95% CI. et al. 2004a; Sillanpaa et al. 2005a; Mechanic et al. 2006) associated with an increase in the RR for breast cancer by with estimates (Gram et al. 2005) and 1 cohort study an average of 12%, and former smoking with an increase by an average of 9%. These results are similar to those based on only a no active/no passive exposure reference - group. Excluding these studies did not meaningfully alter for ever smoking. Neither of these analyses was signifi the overall results for either current smokers (RR = 1.11; - cantly affected by publication or small-study bias accord ing to Begg or Egger statistics (see notes for Figures 6.33 95% CI, 1.07–1.16) or former smokers (RR = 1.09; 95% CI, 1.04–1.13). There was significant heterogeneity among and 6.34). the cohort studies for the association with former smok - Duration of Cigarette Smoking ing because of 1 study (Hiatt et al. 1988) with an outlying estimate (RR = 0.65; 95% CI, 0.47–0.89). The exclusion Several cohort studies support an association of this study, as well as the other 5 that were excluded, between risk for breast cancer and long duration of = 0.220) but did not eliminated the heterogeneity (p h ). The Canadian National S smoking exposure (Table 6.14 change the point estimate. The association between risk Breast Screening Study (RR = 1.50; 95% CI, 1.19–1.89; for breast cancer and former smoking may be attenuated p trend = 0.0003 for ≥40 years) (Cui et al. 2006) and the - relative to current smoking because the former associa NHS-II (RR = 1.21; 95% CI, 1.01–1.45; p trend = 0.04 tion includes women with variable lengths of time since for ≥20 years) (Al-Delaimy et al. 2004) both showed cessation. These results suggest that current smoking is 233 Cancer

263 Surgeon General’s Report Forest plot showing association between current smoking and risk for breast cancer, based on the Figure 6.33 subset of cohort studies published before 2012 and case-control studies published from 2000 to 2011 (n = 25) Note: * = cohort study; ^ = case-control study. Meta-analysis RR = 1.12 (95% CI, 1.08–1.16); Begg z = -0.75, p = 0.46; Egger bias = S ( ) for studies excluded. Size of square is proportional to the weights used in the meta-analysis; note f -0.21, p = 0.62. See Table 6.17 error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and associated 95% CI. CI = confidence interval; RR = relative risk. fewer cases than that of the report by Cui and colleagues increased risks that were significant at approximately 16 and 10 years of follow-up, respectively. (2006). The two analyses adjusted for the same covariates. An earlier analysis Although Egan and colleagues (2002) did not observe a of the Canadian cohort by Terry and colleagues (2002a) showed risk to be approximately 7% higher for 40 or more significant trend for the association between risk for trend = 0.18) years of smoking (RR = 1.61; 95% CI, 1.19–2.19; p for breast cancer and duration of smoking (p for trend = 0.009), but the 2002 report was based on 1,893 in the NHS-I, the recent updated analysis by Xue and 234 Chapter 6

264 —50 Years of Progress The Health Consequences of Smoking Figure 6.34 Forest plot showing association between former smoking and risk for breast cancer, based on the subset of cohort studies published before 2012 and case-control studies published from 2000 to 2011 (n = 25) Note: * = cohort study; ^ = case-control study. Meta-analysis RR = 1.09 (95% CI, 1.05–1.13); Begg z = 0.96, p = 0.34; Egger bias = ( S note f ) for studies excluded. Size of square is proportional to the weights used in the meta-analysis; 0.58, p = 0.19. See Table 6.17 error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and = relative risk. associated 95% CI. = confidence interval; CI RR CI, 1.03–1.77) at the highest level (≥50 years). Because all - colleagues (2011) with 30 years of follow-up found a signifi of these studies adjusted for age, it is difficult to attribute cant trend (p = 0.01). The RRs were 1.04, 1.07, and 1.15 for these trends to confounding by that variable. <20, 20–39, and 40 or more years of smoking, respectively. In response to comments posed by Luo and colleagues (2011b) reported a highly significant Johnson (2004) about analyses of the California Teachers Study (p trend = 0.0002) increased risk with duration of smoking data (Reynolds et al. 2004b), Reynolds and colleagues in the Women’s Health Initiative, with an RR of 1.35 (95% 235 Cancer

265 Surgeon General’s Report of duration used in the meta-analyses. This indicates an (2004a) presented essentially the same results for all women (RR = 1.15; 95% CI, 1.00–1.33; p trend = 0.009 increasing trend in risk with longer duration of smoking or a dose-response relationship. These results suggest that at ≥31 years of smoking duration) and for nullipa- active smoking of long duration (i.e., 20 or more years) rous women only (RR = 1.13; 95% CI, 0.84–1.52; increases risk for breast cancer by a significant average of p trend = 0.081, also at ≥31 years of duration). Two other cohort studies showed increased risks of 26% (Gram et al. 15%. This estimate may be conservative, as some studies indicate that risk continues to increase with smoking over 2005) and 18% (Olson et al. 2005), respectively, for the highest categories of smoking duration. longer periods (Cui et al. 2006; Luo et al. 2011b). - A total of 21 studies reported estimates for dura tion of smoking, after the 13 exclusions cited above Cigarettes Smoked Per Day S (Table 6.16 A and B) (Roddam et al. 2007) not included The number of cigarettes smoked per day provides - because only continuous result reported. Nineteen stud a measure of smoking intensity. In most studies, it rep - ies with data for smoking duration of 20 or more years resents the intensity of current smoking unless data are have examined the associated risk for breast cancer and available for multiple time points that can be used to were included in the meta-analysis: 7 cohort (Al-Delaimy - intensity of smok usual interpret the measure as the et al. 2004; Reynolds et al. 2004b; Gram et al. 2005; Olson ing, or intensity over time, the latter often expressed as et al. 2005; Cui et al. 2006; Luo et al. 2011b; Xue et al. pack-years of smoking. A recent study (Lubin et al. 2007) 2011) and 12 case-control studies (Johnson et al. 2000; suggests that smoking intensity, measured as cigarettes Band et al. 2002; Kropp and Chang-Claude 2002; Zheng per day, may have complex interactions with duration et al. 2002a; van der Hel et al. 2003b; Li et al. 2005; Lis - of smoking on risk of disease: high-intensity effects may sowska et al. 2006; Mechanic et al. 2006; Magnusson et al. diminish over time, while low-intensity effects increase. In 2007; Prescott et al. 2007; Rollison et al. 2008; Brown et contrast, associations of duration or pack-years of smok - - al. 2010) (Table 6.16 , Figure 6.35). The summary esti S ing with risk may involve residual confounding with age, mate (RR) for these studies was 1.16 (95% CI, 1.12–1.21; as older women will have smoked longer but will also have S = 0.318) (Table 6.17 ). The Egger test was significant, p h increased risk for breast cancer regardless of smoking. but the Begg test was not, and thus this result may be While all studies included in the present meta-analyses of influenced by publication or small-study bias (see note - duration and pack-years of smoking adjusted for age, resid for Figure 6.35). The summary estimate (RR) was 1.15 ual confounding may remain that could inflate estimates (95% CI, 1.10–1.19; p = 0.819) for the 7 cohort studies h - for longer duration or higher pack-years of smoking. Con = 0.146) for the 12 case- and 1.23 (95% CI, 1.12–1.36; p h sequently, meta-analyses were conducted for studies that ). Three case-control studies S control studies (Table 6.17 quantified risk of breast cancer with cigarettes per day, as had cutpoints that were greater than 20 years (Zheng et al. well as duration of smoking and pack-years of smoking, to 2002a; van der Hel et al. 2003b; Magnusson et al. 2007), provide an alternative measure of dose-response. and the reference group in 1 cohort (Gram et al. 2005) and - A total of 20 studies (9 cohort, 11 case-control) pro 3 case-control studies was based on no active/no passive vided a report on cigarettes per day as a measure of the exposure (Johnson et al. 2000; Kropp and Chang-Claude A and B) (Roddam et al. S intensity of smoking (Table 6.16 2002; Mechanic et al. 2006). A sensitivity analysis 2007 not included because only a continuous result was that excluded these 7 studies resulted in similar over - reported). Higher level of intensity was categorized at 20 all summary estimates for all studies (RR = 1.15; 95% cigarettes for 9 studies, at 21 for 6 studies, and at 25 for 3 = 0.43), case-control (RR = 1.21; 95% CI, 1.11–1.19; p h studies. The cutpoint at 20 is consistent with smoking one CI, 1.05–1.40), and cohort studies (RR = 1.14; 95% CI, pack of cigarettes or more per day. Two of the 20 eligible 1.10–1.19). studies were excluded from the meta-analysis because The same analyses were conducted to estimate the in 1 (Gram et al. 2005) the highest category was 10 or summary RR for less than 20 years of smoking duration more cigarettes per day and in the second (Lissowska et to compare it with the result for 20 years or more. The al. 2006) it was more than 14. Because the focus of the summary estimate for the 19 studies was 1.04 (95% CI, meta-analysis was on maximum dose, studies that have a 1.01–1.07) (Table 6.17 S ). There was no evidence of publi - maximum-dose category less than 20 have the potential to cation or small-study bias according to Begg’s or Egger’s have subjects with substantially higher levels of smoking statistics (p >0.05). There was no difference in the RR included with individuals who smoke considerably less. between case-control and cohort studies, and the estimate Results for low-level compared with high-level smok - was not attenuated with the exclusion of studies using a no ing intensity differed on the order of 2.7% for all studies, - active/no passive reference group or those that had a cut 4.7% for cohort studies, and 3.4% for case-control studies. point that differed by more than 2 years from the 20 years Chapter 6 236

266 —50 Years of Progress The Health Consequences of Smoking Figure 6.35 Forest plot showing association between 20 or more years of smoking duration and risk for breast can- cer, based on the subset of cohort studies before 2012 and case-control studies published from 2000 to 2011 (n = 19) Note: * = cohort study; ^ = case-control study. Meta-analysis RR = 1.16 (95% CI, 1.12–1.21); Begg z = 1.57, p = 0.12; Egger bias = 1.03, p = 0.02. See Table 6.17 S ( note g ) for studies excluded. There were three studies with a cutpoint differing from 20 years by more than ± 2 years: 15 or more years (Zheng et al. 2002a and van der Hel et al. 2003) and 11 or more years (Magnusson et al. 2007). Size of square is proportional to the weights used in the meta-analysis; error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and associated 95% CI. CI = confidence interval; RR = relative risk. the 10 case-control studies for estimates involving 20 or The summary estimate for the 18 studies was 1.10 (95% = 0.033). When 3 case-control = 0.031) for fewer than 20 cigarettes fewer cigarettes per day (p CI, 1.06–1.16; p h h smoked per day. Although there was no evidence of pub - studies that used a no active/no passive reference group were excluded, the overall summary estimate was reduced lication bias according to the Begg’s statistic (p = 0.103), = 0.179). the Egger statistic (p = 0.025) suggested bias was present. to 1.08 (95% CI, 1.05–1.12, p h The summary estimate was 1.13 (95% CI, 1.09–1.17; p h Pack-Years of Cigarette Smoking = 0.903) for 20 or more cigarettes per day and there was no evidence of publication or small study bias according The number of pack-years of smoking is calculated S , Figure 6.36). to Begg’s or Egger statistics (Table 6.17 as the product of intensity (i.e., cigarettes smoked per These results appear to be more heavily weighted by the day) and duration of smoking, and thus this indicator 8 cohort studies. There was significant heterogeneity for provides an index of lifetime dose of cigarette smoking. 237 Cancer

267 Surgeon General’s Report Forest plot showing association between 20 or more cigarettes/day and risk for breast cancer, based on Figure 6.36 the subset of cohort studies published before 2012 and case-control studies published from 2000 to 2011 (n = 18) * = cohort study; ^ = case-control study. Meta-analysis RR = 1.13 (95% CI, 1.09–1.17); Begg z = -0.34, p = 0.73; Egger bias = Note: note h S 0.23, p = 0.44. See Table 6.17 ( ) for studies excluded. Size of square is proportional to the weights used in the meta-analysis; error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and RR = relative risk. = confidence interval; CI associated 95% CI. Sixteen studies (6 cohort and 10 case-control) have Some investigators prefer this measure, noting that it pro - examined the association between risk for breast cancer vides greater analytic power than duration alone (Ha et al. 2007). However, in their modeling of lung cancer and cig- and pack-years of smoking and were included in the meta- analysis (Table 6.16 arette smoking, Lubin and Caporaso (2006) noted that the A and B). The summary estimate (RR) S for the 16 studies was 1.16 (95% CI, 1.11–1.21; p = 0.304) measure of pack-years mixes low-intensity smoking over h long durations with high-intensity smoking over short for 20 or more pack-years of smoking (Table 6.17 S , Figure 6.37). The Begg and Egger tests did not reveal any bias periods. Low-dose smoking over a long duration results exposure (see notes for Figure 6.37). Estimates for 20 or more pack- in increasing trends for risk estimates, termed and high-dose smoking over short peri- enhancement, years did not differ meaningfully between study types: ods produces the reverse trend, termed cohort (RR = 1.15; 95% CI, 1.10–1.19; p = 0.346) and reduced potency h (Lubin and Caporaso 2006). In addition, estimates of the case-control (RR = 1.21; 95% CI, 1.09–1.34; p = 0.314) h usual number of cigarettes smoked per day lose validity (Table 6.17 S ). After excluding 1 cohort (Gram et al. 2005) - over longer durations if smoking is punctuated by inter and 3 case-control studies with estimates based on only mittent attempts at cessation. 238 Chapter 6

268 —50 Years of Progress The Health Consequences of Smoking Figure 6.37 Forest plot showing association between 20 or more pack-years of smoking and risk for breast cancer, based on the subset of cohort studies published before 2012 and case-control studies published from 2000 to 2011 (n = 16) * = cohort study; ^ = case-control study. Meta-analysis RR = 1.16 (95% CI, 1.11–1.21); Begg z = 0.54, p = 0.59; Egger bias = Note: ( 0.56, p = 0.23. See Table 6.17 note i ) for studies excluded. There was one study with a cutpoint differing from 20 pack-years by more S than ± 5 years: 28 or more years (Li et al. 2005). Size of square is proportional to the weights used in the meta-analysis; error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and associated = confidence interval; = relative risk. RR CI 95% CI. S ). This 1.11−1.21) for 20 or more pack-years (Table 6.17 a no active/no passive exposure reference group (John - son et al. 2000; Kropp and Chang-Claude 2002; Gammon result was primarily due to the cohort studies, for which et al. 2004a), the overall summary estimate for 20 or more the summary estimate for fewer than 20 pack-years was = 0.872). The result for fewer pack-years was slightly attenuated (RR = 1.14; 95% CI, 1.04 (95% CI, 1.00–1.09; p h - than 20 pack-years of smoking for case-control studies = 0.829), and the RR for case-control stud 1.10–1.18; p h was substantially higher (RR = 1.20; 95% CI, 1.05–1.37) ies was reduced by 6%. The overall summary estimate was = 0.023). The but the heterogeneity was significant (p 1.14 (95% CI, 1.10–1.18; p = 0.900) with the exclusion h h summary estimate and the extent of heterogeneity for of the case-control study by Li and colleagues (2005), these case-control studies were substantially decreased which had a higher cutpoint (more than 52 pack-years) and included only postmenopausal women. The exclusion when the three studies (Johnson et al. 2000; Kropp and of this study sharply reduced the RR for the case-control Chang-Claude 2002; Gammon et al. 2004a) using a no = 0.795). active/no passive exposure reference group were excluded studies to 1.09 (95% CI, 0.96–1.24; p h (RR = 1.10; 95% CI, 0.97–1.24; p = 0.154). Overall, accu- The summary estimate for less than 20 pack-years h = 0.099), mulating 20 or more pack-years increased risk for breast of smoking was 1.09 (95% CI, 1.03–1.15; p h cancer by a significant average of 16%, while smoking which was below the summary estimate of 1.16 (95% CI, 239 Cancer

269 Surgeon General’s Report for less than 20 pack-years was associated with a smaller first two categories were combined so that all 22 studies had estimates for those younger than 20 years of age at increased risk of 9%. The estimate for 20 or more pack- smoking initiation. Sensitivity analyses stratified the stud years of smoking may be conservative, because some - ies by design and excluded studies with large differences studies indicate that risk continues to rise with more pack-years (Xue et al. 2011). in cutpoints or those that used only a no active/no passive Thirteen of the 16 studies with estimates for pack- exposure reference group. Figure 6.38 shows results from all 22 studies for years of smoking also provided risk by duration (Table those younger than 20 years of age at smoking initiation. 6.16 S ). Estimates across levels of duration and pack- The RR summary estimate was 1.11 (95% CI, 1.07–1.16; years of smoking were not necessarily consistent for the S two measures within a study; the Spearman correlation p = 0.088) (Table 6.17 ). The Begg and Egger tests were h across studies was 0.62 (p = 0.02). Nonetheless, the sum not significant (see notes to Figure 6.38; funnel plot not - shown). The estimate for the 8 cohort studies (RR = 1.09; mary estimates suggest that long duration of smoking = 0.541) was similar to that for the 95% CI, 1.06–1.13; p and higher numbers of pack-years of smoking signifi - h 14 case-control studies (RR = 1.12; 95% CI, 1.02–1.22; cantly increase risk for breast cancer by a similar amount, S = 0.029) (Table 6.17 p approximately 11–21% based on the CIs, depending on ). One cohort study (Gram et al. h study design and sensitivity analysis restrictions (Table 2005) and 5 case-control studies (Johnson et al. 2000; S ). The summary estimate from case-control stud 6.17 - Kropp and Chang-Claude 2002; Gammon et al. 2004a; Mechanic et al. 2006; Young et al. 2009) were excluded ies tended to be higher for both duration and pack-years of smoking than for cohort studies but also less stable. from the analysis because estimates were based on a no Taken together, the meta-analyses for duration, cigarettes active/no passive exposure reference group. One study smoked per day, and pack-years provide similar evidence was excluded because the age cutpoint was 16 years of age or younger (Egan et al. 2003). These exclusions did not for a dose-response relationship between smoking and breast cancer. meaningfully alter the summary estimate (RR = 1.09; 95% = 0.597). Nineteen studies (7 cohort, 12 CI, 1.06–1.13; p h case-control) estimated risk when smoking was initiated Timing of Exposure to Tobacco Smoke at 16 or fewer years of age (RR = 1.08; 95% CI, 1.00–1.15; The timing of smoking relative to critical periods of = 0.065). p h - change in the size and morphology of breast tissue—time Only 13 studies (6 cohort, 7 case-control) reported frames such as menarche, during adolescence, or before estimates of risk when smoking initiation occurred first pregnancy—may be important. Based on in vitro stud - from 16–19 years of age (RR = 1.11; 95% CI, 1.07–1.15; ies, hypothesized that smoking is more likely Russo (2002) p = 0.757). Additionally, results for the meta-analysis of h to induce neoplastic changes during these periods, when - the 19 studies that reported estimates for smoking initia the susceptibility of the breast to carcinogens is increased. - tion at 20 years of age and older showed a significant sum Breast cancer also is more likely to develop in undifferen - = 0.672) mary estimate (RR = 1.08; 95% CI, 1.05–1.12; p h tiated tissues that may be susceptible to tobacco-related (Table 6.17 S ). This estimate was only slightly lower than and other carcinogens. Results of epidemiologic studies that for those younger than 20 years of age. Thus, these substantiate that nulliparous women have a higher risk studies did not reveal a clear trend for a change in sum- than parous women of breast cancer. The lower risk for mary estimates across categories for age at initiation. parous women is attributed to having an early full-term Few studies tested for trends across age categories and pregnancy and the subsequent increased differentiation in estimates for most studies included in the meta-analyses the terminal ducts of the breast (Russo et al. 1992, 2000; were similar for those 16 years of age and younger and Russo and Russo 1995, 2008) . those 20 years of age or younger (Spearman rank-order correlation = 0.81, p <0.0001). Of note, the estimates Age at Smoking Initiation in the tails of the distribution of the RRs across studies with either significant protective or increased estimates Twenty-two studies with data for age at smoking are from studies that used a no active/no passive exposure initiation were evaluated: 8 cohort studies and 14 case- reference group. Taken together, the meta-analyses of S , see notes for Figure 6.38 control studies (Table 6.16 these studies did not provide clear evidence that initiating for exclusions). The cutpoints for age varied among these smoking during adolescence or young adulthood confers studies. Therefore, estimates were allocated into the clos - any greater risk than initiation at older ages. est of the following categories: younger than 16 years of age, 16–19 years of age, and 20 years of age and older. The 240 Chapter 6

270 —50 Years of Progress The Health Consequences of Smoking Figure 6.38 Forest plot showing association between less than 20 years of age at smoking initiation and risk for breast cancer, based on the subset of cohort studies published before 2012 and case-control studies published from 2000 to 2011 (n = 22) Note: * = cohort study; ^ = case-control study. Meta-analysis RR = 1.11 (95% CI, 1.07–1.16); Begg z = 0.59, p = 0.55; Egger bias = 0.63, p = 0.12. See Table 6.17 S ( note j ) for studies excluded. There were six studies with a cutpoint differing from 20 years of age at smoking initiation by more than ± 2 years: 15 years of age and younger (Prescott et al. 2007), 16 years of age and younger (Egan et al. 2003), and 18 years of age and younger (Gammon et al. 2004a; Mechanic et al. 2006; Olson et al. 2005; Rollison et al. 2008). Size of square is proportional to the weights used in the meta-analysis; error bars show the associated 95% CI. Solid vertical line represents RR = confidence interval; = relative risk. the null value. Diamond represents the summary estimate and associated 95% CI. CI 241 Cancer

271 Surgeon General’s Report Smoking Before or During First Full-Term 13 case-control studies (Innes and Byers 2001; Band et al. Pregnancy 2002; Kropp and Chang-Claude 2002; Fink and Lash 2003; Gammon et al. 2004a; Li et al. 2005; Lissowska et al. 2006; The effects of smoking before versus after a first full- Magnusson et al. 2007; Prescott et al. 2007; Rollison et al. term pregnancy may be confounded by effects associated 2008; Slattery et al. 2008; Young et al. 2009; DeRoo et al. with early age at smoking initiation and age at first preg - - , see notes for Figure 6.39 for exclu 2011a) (Table 6.16 S nancy (Cui et al. 2006). Few studies have examined the sions). For these 22 studies, the RR summary estimate risk of smoking during pregnancy, for which the results ). S = <0.001) (Table 6.17 was 1.10 (95% CI, 1.04–1.17; p h may differ for women who stop smoking when pregnant This summary result is higher and statistically significant than for those who continue to smoke during pregnancy. compared with that of Lawlor and colleagues (2004), pri - Lawlor and colleagues (2004) conducted a meta-analysis marily because it included 5 recent, large cohort studies of 11 studies, 2 of which were based on smoking during that reported significant estimates (Al-Delaimy et al. 2004; pregnancy (Innes and Byers 2001; Fink and Lash 2003), to Gram et al. 2005; Olson et al. 2005; Luo et al. 2011b; Xue assess the effect of smoking before a first full-term preg - et al. 2011). The RR summary estimate was 1.16 (95% CI, nancy. The analysis included estimates from their own 1.12–1.20; p = 0.746) for the 9 cohort studies and 1.05 h cohort, the British Women’s Heart and Health Study, 2 (95% CI, 0.94–1.18; p = 0.001) for the 13 case-control h earlier cohort studies (Egan et al. 2002; Reynolds et al. studies (Table 6.17 S ). After excluding 1 cohort study 2004b), and 8 case-control studies (Adami et al. 1988; (Gram et al. 2005) and 3 case-control studies (Gammon Hunter et al. 1997; Lash and Aschengrau 1999, 2002; et al. 2004a; Kropp and Chang-Claude 2002; Young et Innes and Byers 2001; Band et al. 2002; Kropp and Chang- al. 2009) that were based on estimates using only a no Claude 2002; Fink and Lash 2003). Based on 6,528 cases, active/no passive exposure reference group, the overall the RR summary estimate was 1.07 (95% CI, 0.94–1.22). summary estimate increased slightly (RR = 1.11; 95% CI, The risk was attenuated when 2 influential studies with ≤0.001) due to the increase for case-control 1.05–1.18; p h wide CIs (Lash and Aschengrau 1999; Innes and Byers ≤0.001). The studies (RR = 1.09; 95% CI, 0.96–1.23; p h 2001) were removed (RR = 1.03; 95% CI, 0.93–1.14), additional exclusion of the 3 case-control studies, which which also reduced heterogeneity. These 2 studies and an reported estimates for smoking only during pregnancy earlier one based on the NHS-I (Hunter et al. 1997) were (Innes and Byers 2001; Fink and Lash 2003; DeRoo et al. 3 of the 11 that reported statistically significant results. 2011a), further increased the RR for case-control studies DeRoo and colleagues (2011b) published a meta- to 1.13 (95% CI, 1.05–1.23), eliminating the significant analysis on a larger number of studies than the earlier - = 0.727). In addition, the overall sum heterogeneity (p h These authors review by Lawlor and colleagues (2004). mary estimate was increased to 1.16 (95% CI, 1.12–1.20; included an additional 15 reports (Morabia et al. 1996; - p = 0.830). Thus, the 3 case-control studies with risk esti h Egan et al. 2003; Al-Delaimy et al. 2004; Gammon et al. mates for smoking only during pregnancy produced het- 2004a; Gram et al. 2005; Li et al. 2005; Olson et al. 2005; Cui erogeneity and attenuated summary estimates, but those et al. 2006; Lissowska et al. 2006; Ha et al. 2007; Magnus - that used a no active/no passive exposure reference group son et al. 2007; Prescott et al. 2007; Rollison et al. 2008; had little or no effect on the summary estimates. Slattery et al. 2008; Young et al. 2009). They excluded 2 These summary estimates for smoking before or studies of smoking during first pregnancy based on linked during first pregnancy are only slightly higher than those birth and cancer registry data (Innes and Byers 2001; Fink for ever smoking, and they are quite similar to those for and Lash 2003) and 1 study (Hunter et al. 1997) that over - duration of 20 or more years and 20 or more pack-years of lapped with a subsequent report (Egan et al. 2002); these 3 smoking. Overall, the studies conducted since 2000 do not (i.e., all but Egan et al. 2002) were included in Lawlor and provide clear evidence that smoking before first pregnancy colleagues’ (2004) meta-analysis. DeRoo and colleagues’ confers a greater risk than smoking at any other time in (2011b) summary estimate was 1.11 (95% CI, 1.06–1.16). a woman’s life. Taken together, the results for earlier age This higher estimate than that of Lawlor and colleagues at smoking initiation and smoking before first pregnancy (2004) was influenced by several large cohort and case- do not support the hypothesis that smoking has greater control studies published between January 2004 and 2009. carcinogenic effects during periods in which breast tissue Twenty-two studies included in this report pro- is less differentiated and theoretically more susceptible. vided RR estimates for smoking before or during first full-term pregnancy for the meta-analysis: 9 cohort stud- Menopausal Status ies (Al-Delaimy et al. 2004; Lawlor et al. 2004; Reynolds et al. 2004b; Gram et al. 2005; Olson et al. 2005; Cui et al. Risk for breast cancer is associated with duration 2006; Ha et al. 2007; Luo et al. 2011b; Xue et al. 2011) and and level of estrogen exposure and evidence suggests that Chapter 6 242

272 —50 Years of Progress The Health Consequences of Smoking Forest plot showing association between smoking before or during first full-term pregnancy and risk Figure 6.39 for breast cancer, based on the subset of cohort studies before 2012 and case-control studies published from 2000 to 2011 (n = 22) * = cohort study; ^ = case-control study. Meta-analysis RR = 1.10 (95% CI, 1.04–1.17); Begg z = -1.16, p = 0.25; Egger bias = Note: ) for studies excluded. Estimates based on exposure before first pregnancy with the exception note m ( S -0.50, p = 0.43. See Table 6.17 of three studies, which reported an estimate based on exposure during first pregnancy (Innes and Byers 2001; Fink and Lash 2003; DeRoo et al. 2011a). Size of square is proportional to the weights used in the meta-analysis; error bars show the associated 95% CI. Solid vertical line represents the null value. Diamond represents the summary estimate and associated 95% CI. = confidence inter- CI RR val; = relative risk. the phenotypic heterogeneity of breast cancer is linked to metabolism, the risk of breast cancer due to smoking is menopausal status (Lipton 2005). Spicer and Pike (2005) similarly modified by menopause. hypothesized that because menopause is associated with a For some risk factors, such as obesity, risk estimates differ when analyses are stratified by menopausal status decreased rate of breast cell proliferation compared with - that in the premenopausal period, it modifies susceptibil (van den Brandt et al. 2000). Menopause could modify the risk of breast cancer associated with smoking by altering ity to exposures such as obesity, hormone therapy, and hormone metabolism and the sensitivity of breast tissue alcohol. It is plausible that if smoking affects hormone 243 Cancer

273 Surgeon General’s Report provides the RR for case-control studies, with the study to tobacco carcinogens (Kendall et al. 2007). Women who by Kruk (2007) excluded because of its extreme estimates. smoke—primarily current, heavy smokers—experience menopause at an earlier age than those who do not smoke (Baron et al. 1990; Midgette and Baron 1990; Kato et al. Menopausal Status—Ever Smoking 1998; Mikkelsen et al. 2007; Sun et al. 2012) and have a Among 17 studies, 3 cohort (Hiatt and Fireman higher risk for osteoporosis even when on estrogen ther - 1986; Manjer et al. 2000b; Xue et al. 2011) and 3 case-con - apy (North American Menopause Society 2010), which trol (Band et al. 2002; Lissowska et al. 2006; Kruk 2007) may be due to altered estrogen metabolism and lower studies reported a significantly increased risk for pre- - estrogen levels (Kiel et al. 1992). These observations sup menopausal women associated with ever smoking. All but port an anti-estrogenic effect of smoking (Kendall et al. 6 studies had an RR greater than 1.10, and no significant 2007). However, smokers also tend to be leaner, drink inverse associations were reported. The summary esti- more alcohol, and have poorer diets than nonsmokers; all mate (RR) associated with premenopausal smoking for all of these factors are also associated with early menopause studies combined was 1.26 (95% CI, 1.11–1.43; p ≤0.001) h (Sampson 2002). Moreover, results from several studies ). This RR was reduced to 1.18 (95% CI, 1.08– S (Table 6.18 have not provided sufficient evidence that estradiol levels 1.29; p = 0.005) when the single outlying estimate for a h in current smokers differ from those in former or never case-control study (RR = 2.34) (Kruk 2007) was excluded smokers (Longcope and Johnston 1988; Baron et al. 1990; S ). The summary estimate for the case-control (Table 6.18 Key et al. 1991; Cassidenti et al. 1992; Kendall et al. 2007; = studies was reduced from 1.30 (95% CI, 1.04–1.62; p h Arslan et al. 2009). Even so, in a recent cross-sectional 0.001) to 1.20 (95% CI, 1.02–1.42; p = 0.075) when the h analysis of the association between endogenous hormones outlier was excluded, a value that is quite similar to the RR and several risk factors for breast cancer, the levels of for the 4 cohort studies (RR = 1.16; 95% CI, 1.08–1.24; p h all sex hormones were reported to be higher for women S ). = 0.628) (Table 6.18 who smoked 15 or more cigarettes per day than for never A total of 17 studies reported results for smoking smokers. Hormonal levels, particularly for estrogen, were by postmenopausal women. Four out of 6 cohort studies attenuated with adjustment for BMI, whereas further reported positive associations of 1.10 or greater, of which adjustment for alcohol did not result in any meaningful 2 were significant (Olson et al. 2005; Luo et al. 2011b). - change (Endogenous Hormones and Breast Cancer Col One cohort study (Xue et al. 2011), however, reported a laborative Group 2011). significant inverse association (RR = 0.91; 95% CI, 0.86– Previous reviews did not find evidence to suggest 0.96). Three of the 11 case-control studies that included that menopause modifies the risk of breast cancer from postmenopausal women reported significant positive smoking (Egan et al. 2002; Terry and Rohan 2002). The associations for this group (Johnson et al. 2000; Li et al. Collaborative Group on Hormonal Factors in Breast Can - 2005; Kruk 2007). Five studies reported an RR greater cer and colleagues (2002) reported an RR of 1.07 (stan- than 1.10, and none reported a significant inverse asso - dard error = 0.05) for premenopausal women and an RR of ciation. The summary estimate associated with postmeno - 1.12 (standard error = 0.06) for women 50 years of age and pausal women for all studies combined was 1.10 (95% CI, older who experienced natural menopause. S 1.02–1.19; p ). This RR was reduced = 0.001) (Table 6.18 h Several studies have examined menopausal status to 1.07 (95% CI, 1.00–1.14; p = 0.001) when the outlying h specifically, and several have conducted formal tests for estimate (RR = 1.76) (Kruk 2007) was removed. The sum - interaction with smoking. Table 6.18 S shows results for mary estimate for the case-control studies was reduced ever smoking from 14 studies stratified by menopausal from 1.13 (95% CI, 1.01–1.27; p = 0.001) to 1.07 (95% CI, h status and 6 studies in which the entire study sample 0.98–1.16; p = 0.147) when the outlier was removed, an h included only one menopausal group. Of the 20 studies estimate virtually identical to the estimate based on the 6 listed, 7 reported data for pack-years of smoking for both = 0.001) cohort studies (RR = 1.07; 95% CI, 0.97–1.19; p h menopausal groups and 3 reported results for postmeno - S (Table 6.18 ). pausal women only. Overall, results for ever smoking Several issues should be considered when evaluat- - were highly variable for both premenopausal and post - ing these results for ever smoking in premenopausal ver menopausal risks. Menopause can be difficult to define in sus postmenopausal women. First, the estimates reported observational studies, however, which can result in mis - by Kruk (2007) are outliers for both menopausal groups classification bias, particularly when age is the only crite - and, when these estimates are included, the summary rion for menopause. Furthermore, not all studies in Table estimates (RRs) are positively biased. The significant S 6.18 accounted for residual confounding by hormonal inverse association in postmenopausal women reported status or use of HRT. A sensitivity analysis (Table 6.18 S ) Chapter 6 244

274 —50 Years of Progress The Health Consequences of Smoking by Xue and colleagues (2011) for the NHS-I contrasts with with breast cancer in postmenopausal women. The pooled estimate for 20 or more pack-years was statistically sig- the significant positive associations reported by two other large cohort studies, Women’s Health Initiative (Luo et al. nificant in Reynolds and colleagues (2004b) (pooled RR = 1.17; 95% CI, 1.01–1.35), Olson and colleagues (2005) 2011b) and the Iowa Women’s Health Study (Olson et al. - 2005). Previous reports from NHS-I (London et al. 1989; (pooled RR = 1.17; 95% CI, 1.04–1.31), and Luo and col leagues (2011b) (pooled RR = 1.12; 95% CI, 1.03–1.21). Egan et al. 2002) have indicated a null association and no Luo and colleagues (2011b) also found a statistically sig - meaningful difference between menopausal groups, but they were based on fewer cases and less follow-up time nificant increased risk for smoking more than 50 pack- years (RR = 1.18; 95% CI, 1.02–1.22). In contrast, there than the recent report by Xue and colleagues (2011). was a trend toward lower risk with more pack-years of Among the case-control studies, the study by Johnson and colleagues (2000) also provided estimates smoking in Xue and colleagues (2011), which reached sta - for smoking by menopausal status that used a small no tistical significance for the highest level of more than 15 active/no passive exposure reference group: for premeno - pack-years (RR = 0.88; 95% CI, 0.79–0.99). In contrast, only two (Johnson et al. 2000; Li et al. 2005) of the six - pausal women, OR = 2.3 (95% CI, 1.2–4.5), and for post menopausal women, OR = 1.5 (95% CI, 1.0–2.3). These - case-control studies reported statistically significant asso ciations for the highest level of pack-years of smoking in estimates contrast strongly with their results when using postmenopausal women (RR = 1.60; 95% CI, 1.00–2.60, a no active-only reference group (Table 6.18 S ): premeno- and RR = 1.30; 95% CI, 1.00–2.60, respectively). It should pausal women (OR = 1.0; 95% CI, 0.80–1.3); postmeno- be noted that the estimates reported by Johnson and col - pausal women (OR = 1.2; 95% CI, 1.0–1.4). No other study - leagues (2000) were based on a no active/no passive expo has contrasted estimates using these two reference groups - by menopausal status. It is important to note that John sure reference group. son and colleagues (2000) restricted their analysis using Only one cohort study (Reynolds et al. 2004b) formally tested for interaction between menopause and a no active/no passive exposure reference group to the smoking across multiple measures. This study found no approximate 60% of women who reported their resi - dential exposure to passive smoke for at least 90% of significant results by the likelihood ratio test for dura - their lifetime. This makes a direct comparison of their tion of smoking (p = 0.80); cigarettes/per day (p = 0.42); pack-years of smoking (p = 0.07); and years since cessa- results difficult. tion (p = 0.76). Menopausal Status—Pack-Years of Smoking Menopausal Status—Summary Several studies have reported results for pack-years by menopausal status: 7 for premenopausal and 10 for S - indicate that consider The results in Table 6.18 ). The results across these - able heterogeneity exists among studies that report esti postmenopausal (Table 6.18 S mates for the association of smoking with breast cancer studies are variable and inconsistent. Two cohort studies by menopausal status, although none of the summary that reported results for premenopausal women (Reynolds estimates was associated with statistically significant pub - et al. 2004b; Xue et al. 2011) found significantly increased risks for the highest category of pack-years of smoking lication bias. Although the results of the meta-analysis suggest that risk is greater in premenopausal than in (≥30) (RR = 2.05; 95% CI, 1.20–3.49 and RR = 1.27; 95% CI, 1.16–1.38, respectively). Among 5 case-control studies postmenopausal women, it remains uncertain whether the association of smoking with breast cancer differs by offering estimates for premenopausal women, 2 reported menopausal status. - statistically significant positive associations for the high est level of pack-years of smoking (Band et al. 2002: RR = 1.69; 95% CI, 1.10–2.61 for ≥20 pack-years; Slattery et Hormone Receptor Status al. 2008: RR = 1.6; 95% CI, 1.1–2.4 for >15 pack-years) in ERs and progesterone receptors (PRs) mediate the non-Hispanic Whites, while 1 (Johnson et al. 2000) found effects of estrogen and progesterone on the growth, prolif - significant increased risks for fewer pack-years of expo- eration, and differentiation of breast tumors; response to sure (RR = 2.30; 95% CI, 1.10–4.70 for 11–20, and RR = hormonal treatment; recurrence; and survival. Palmer and 2.40; 95% CI, 1.20–4.70 for 1–10 pack-years). The other 2 postulated that the expression status of Rosenberg (1993) - studies (Zheng et al. 2002a; Ahern et al. 2009) were essen - ERs could modulate the anti-estrogenic effects of smok tially null for the association between breast cancer and Meek and Finch (1999) reported that smoking ing, and pack-years of smoking in premenopausal women. alters the expression of ERs. The presence (+) or absence Four cohort and six case-control studies reported (–) of ER expression in breast tumors is increasingly estimates for the association of pack-years of smoking 245 Cancer

275 Surgeon General’s Report - recognized as a potential biomarker of etiologically dis The recent case-control study by Rabstein and colleagues (2010) found a significant inverse association with ER+ (Anders et al. 2008; Bertucci et al. 2009; tinct subtypes . Consequently, some of the more Onitilo et al. 2009) breast cancer (RR = 0.79; 95% CI, 0.65–0.95), but no asso- recent studies stratify analyses on ER expression. The ciation with ER– breast cancer. The remaining studies information added by cross-classification with the status reported null results (McTiernan et al. 1986; Stanford et al. 1987; London et al. 1989; Huang et al. 2000a; Britton of PRs remains controversial. In addition to reporting the expression status of ERs and PRs, studies have begun et al. 2002; Cotterchio et al. 2003; Gammon et al. 2004a; to cross-classify cases by the status of human epidermal Lissowska et al. 2006; Trivers et al. 2009). growth factor receptor 2 (HER2) because the so-called triple negative phenotype (i.e., the combination of nega - Hormone Receptor Status—Cigarettes Smoked Per Day tive ER, PR, and HER2 status) is increasingly recognized (Bauer et al. 2007; as distinct and having a poor prognosis Only six studies have reported results on the associa - . ReisFilho and Tutt 2008; Gluz et al. 2009) tion between cigarettes smoked per day and breast cancer - Many studies have assessed the risk of breast can defined by ER status, and these are also very inconsistent cer based on the status of ER expression. In 2 early, small ). One study (London et al. 1989) reported a S (Table 6.19 hospital-based studies, Daniell (1980) and Ranocchia and significantly increased risk for ER+ breast cancer with 25 colleagues (1991) observed that the prevalence of smoking or more cigarettes smoked per day (RR = 1.38; 95% CI, was higher among breast cancer cases with ER– tumors 1.04–1.84), and another (Al-Delaimy et al. 2004) reported than in cases with ER+ tumors, but these studies were significantly increased risks for ER+ breast cancer with underpowered and the data were not rigorously analyzed. fewer cigarettes smoked per day: RR = 1.46; 95% CI, 1.14– S Table 6.19 summarizes data from 17 studies that assessed 1.87 for 5–14 cigarettes smoked per day; and RR = 1.45; whether the risk for breast cancer differs by ER expres- - 95% CI, 1.09–1.93 for 1–4 cigarettes smoked per day. Man sion status for ever smoking or by the highest category of jer and colleagues (2001) found significantly increased cigarettes smoked per day. Althuis and colleagues (2004) risks for ER– breast cancer regardless of number of - reviewed 10 of the studies shown in Table 6.19 S (McTier cigarettes smoked per day, and Morabia and colleagues nan et al. 1986; Stanford et al. 1987; Cooper et al. 1989; (1998) reported significantly increased risks for both ER+ London et al. 1989; Yoo et al. 1997; Morabia et al. 1998; and ER– breast cancer regardless of level, although the Huang et al. 2000a; Manjer et al. 2001; Britton et al. 2002; association was somewhat stronger in women with ER– Cotterchio et al. 2003) with hormone receptor-defined tumors. The remaining two studies reported essentially breast cancer and found no evidence for a differential asso - null results (Li et al. 2005; Lissowska et al. 2006). ciation between breast cancer and smoking by hormonal phenotype, but they did not provide a numerical analysis. Hormone Receptor Status—Methodologic Four of these studies (Cooper et al. 1989; London et al. Issues 1989; Yoo et al. 1997; Morabia et al. 1998) were reviewed Some issues affect the interpretation of published in the 2006 Surgeon General’s report. - results for smoking and breast cancer by hormone recep tor status. First, all but two studies (London et al. 1989; Hormone Receptor Status—Ever Smoking used case-control in Table 6.19 Al-Delaimy et al. 2004) S Findings from the 17 studies on the association of designs, which are more subject to bias than other study ever smoking with breast cancer defined by ER status are designs. Second, methods for detecting ER expression highly inconsistent (Table 6.19 S ). Four studies reported have changed over time, and some older studies were based significantly increased risks for ER+ breast cancer with on a mix of methods . Many (Ross and Hortobagyi 2005) ever smoking, with RRs ranging from 1.15–1.42 (Yoo et studies rely on incomplete or inaccurate pathology and al. 1997; Al-Delaimy et al. 2004; Li et al. 2005; Luo et medical records and ER status is generally not obtained al. 2011b). Two studies reported significantly increased on in situ tumors. The completeness of data for ER sta - risks for ER–breast cancer (Cooper et al. 1989; Manjer et S tus in the studies in Table 6.19 ranged from 40–100%. al. 2001), with RRs ranging from 1.63–2.41. One study Third, few studies have identified consistent risk factors (Morabia et al. 1998) reported significantly increased for the ER– phenotype other than race and younger age - risks for both ER+ and ER– breast cancer, with a some , and thus potential confounders for (Althuis et al. 2004) what stronger association with ER– (RR = 4.01; 95% this type of breast cancer are not yet well characterized. CI, 1.90–8.46) than ER+ (RR = 2.28; 95% CI, 1.56–3.35) Last, researchers are not sure whether ER status should tumors. This study is the only one that used a no active/ be cross-classified with PR status. The most recent studies no passive exposure reference group (Morabia et al. 1998). 246 Chapter 6

276 The Health Consequences of Smoking —50 Years of Progress of the primary behavioral risk factors, along with alcohol - have characterized breast cancer phenotypes by the com consumption, obesity, and use of oral contraceptives. In bination of ER, PR, and HER2 status or by gene expression a review by Chen and colleagues (1999) of the 16 studies phenotypes (luminal A, B, basal-like) (Kwan et al. 2009; . Kabat and colleagues (2011) recently they examined, 3 included cigarette smoking as a factor Trivers et al. 2009) published an analysis from the Women’s Health Initiative of interest (Kato et al. 1986; Horn and Thompson 1988; Bernstein et al. 1992), but there was no strong evidence on risk of the triple negative phenotype compared with risk for ER+ breast cancer in relation to smoking. RRs of a significant increased risk. These 3 studies, along with (not shown in Table 6.19 S ) were significantly increased 4 reports published in 2001 or later (Fowble et al. 2001; Trentham-Dietz et al. 2007a; Knight et al. 2009; Li et al. in women with ER+ breast cancer for former smoking S (1.14; 95% CI, 1.05–1.24), duration of 30 or more years . Overall, the find- 2009a), are summarized in Table 6.20 (1.14; 95% CI, 1.01–1.28), 40 or more pack-years of smok- ings of these 7 studies are inconclusive with regard to the risk of a second primary contralateral breast cancer in ing (1.25; 95% CI, 1.06–1.44), and younger than 20 years of age at initiation (1.16; 95% CI, 1.05–1.28). In contrast, smokers. In the largest cohort of women diagnosed with - - invasive cancer, the findings for both former and cur there were no significant associations in women with tri rent smoking were not significant (Trentham-Dietz et al. ple negative breast cancer. These results are quite similar - 2007a). In the most recently conducted study, which cov to those reported by Luo and colleagues (2011b), who also ered a 15-year follow-up period, Li and colleagues (2009a) analyzed tumors by ER/PR status only (not HER2) data - from the Women’s Health Initiative cohort. reported a significant association between cigarette smok ing and both a contralateral breast cancer diagnosis (RR Hormone Receptor Status—Summary = 2.2; 95% CI, 1.2–4.0) and risk of the first primary breast cancer diagnosis (RR = 1.8; 95% CI, 1.1–3.2). Although In summary, results from studies conducted to date Knight and colleagues (2009) evaluated a number of are inconsistent on the association of smoking with dif - smoking measures, including duration, average packs per ferent phenotypes of breast cancer defined on the basis of - day, pack-years, and age at initiation, they found little evi hormone receptor status. dence for an association between cigarette smoking and risk of a primary contralateral breast cancer. That study was focused primarily on premenopausal women, whereas Exposure to Tobacco Smoke in the study by Li and colleagues (2009a) the majority of women (81%) were postmenopausal and diagnosed with and Risk of Second Primary ER+ cancer. Taken together, the results for the association Contralateral Breast Cancer between smoking and having a contralateral breast cancer remain inconclusive. Although a recent study indicates that there was a downward trend in the incidence of contralateral breast cancer in the United States from 1975–2006 (Nichols et al. Genetic Susceptibility to Smoking 2011), a summative review published in 1999 documented prevalence estimates ranging from 2–11% (Chen et al. The 2004 Surgeon General’s report summarized 1999), and a follow-up of 305,533 breast cancer cases in eight studies on the smoking-genotype interaction: one the SEER Program database provided an estimate of 4.3% 2 on family history BRCA1 , one on (Couch et al. 2001) / for the development of a second primary contralateral NAT2 and NAT1 , three on (Brunet et al. 1998) (Ambro- breast cancer (Bernstein et al. 2003). sone et al. 1996; Hunter et al. 1997; Millikan et al. 1998) , A second primary breast cancer has most frequently (Ambrosone et al. 1999a) , and two on GSTM1 one on been defined as a new and independent tumor, although (Ambrosone et al. 1995; Ishibe et al. 1998) . The CYP1A1 studies have varied on whether carcinoma in situ has been report concluded that susceptible subgroups of women included. The risk of developing a second primary con - could not be “reliably identified” (USDHHS 2004, p. 312) . tralateral breast cancer has been evaluated in a number The Cal/EPA (2005) provided descriptive summaries of of studies (Kato et al. 1986; Horn and Thompson 1988; studies that focused on susceptible subgroups (i.e., deter - Bernstein et al. 1992; Fowble et al. 2001; Trentham-Dietz mined by family history, genotype, tumor phenotype); the et al. 2007a; Knight et al. 2009; Li et al. 2009a), primarily Canadian Expert Panel tabulated data on the interaction over the past decade, as the number of women who have between smoking and a number of genotypes and consid - survived breast cancer has steadily increased and there (Collishaw ered the evidence for to be “persuasive” NAT2 has been a growing interest in modifiable risk factors for et al. 2009, p. 47) ; and the 2009 IARC Monograph Working this disease. Cigarette smoking has been examined as one 247 Cancer

277 Surgeon General’s Report a family history of breast cancer: fewer than 30 pack-years Group concluded that results from studies of interactions (RR = 0.98; 95% CI, 0.87–1.10) and 30 or more pack-years between smoking and genes were “ambiguous, with the NAT2 ” (Secretan et al. 2009, p. 1034) possible exception of (RR = 0.97; 95% CI, 0.72–1.31). These studies provide . strong evidence that genetic factors represented by fam- Family History ily history of breast cancer modify the risk for that cancer - associated with smoking. More studies are needed to rep Having a family history of first-degree relatives with licate this interaction of smoking and family history and breast cancer is associated with a doubling to tripling of to identify underlying genetic mechanisms. risk for breast cancer (Goldgar et al. 1994; Pharoah et al. 1997; Poole et al. 1999). This risk is further increased in BRCA1/BRCA2 women with benign breast disease and a family history of An estimated 5−10% of all diagnosed breast cancer breast cancer, especially those with atypical hyperplasia is inherited, with 2–3% involving mutations in one of the (Collins et al. 2006). This finding provides strong evidence (Ashworth et BRCA2 or BRCA1 tumor suppressor genes for a genetic predisposition to breast cancer and has led al. 2010). These mutations account for nearly 40–50% to rapidly expanding efforts to identify specific genetic of familial breast cancer cases (Chen et al. 2006b; Ash - variants that increase such risk. These may be either rare variants with large effects or the joint action of common worth et al. 2010), and women with these mutations are variants (SNPs) with small effects that modify susceptibil - at high risk for developing breast cancer, especially at an ity to behavioral or environmental exposures associated early age (Chen et al. 2006b). The cumulative incidence - of breast cancer is also high for those who carry an inher with breast cancer. This section considers evidence for heritable genetic susceptibility to smoking as a risk factor BRCA ited mutation, with an estimated lifetime risk of at least 43–46% by age 70 (Chen et al. 2006b), although for breast cancer. estimates of 60−80% have been proposed (Ashworth et al. Most studies on smoking and breast cancer have - controlled for family history, but only a few have assessed 2010). These estimates have varied considerably depend the interaction of smoking and family history (Couch et al. ing on the patients selected and patterns of inheritance. As a result, there is considerable inconsistency among 2001; Suzuki et al. 2007). Couch and colleagues (2001) reported that among 132 families with three or more reports to date. incident cases of breast or ovarian cancer in sisters and Eight studies (Brunet et al. 1998; Ghadirian et al. daughters, ever smokers had an increased risk (RR = 2.4; 2004; Colilla et al. 2006; Gronwald et al. 2006; Nkondjock 95% CI, 1.2–5.1) for breast cancer compared with never et al. 2006; Breast Cancer Family Registry (BCFR) 2008; smokers. Risk for ever smokers was even higher (RR = 5.8; Ginsburg et al. 2009; Moorman et al. 2010) have examined mutations are more whether carriers of BRCA1 and BRCA2 95% CI, 1.4–23.9) in 35 families with five or more breast and/or ovarian cancers. Suzuki and colleagues (2007) also susceptible or less susceptible to cigarette smoke than are reported a significant interaction between a positive fam noncarriers. Terry and Goodman (2006) reviewed four of - these studies (Brunet et al. 1998; Ghadirian et al. 2004; ily history of cancer and smoking on risk of breast cancer Colilla et al. 2006; Gronwald et al. 2006); in the earliest (p = 0.01). In comparisons with never smokers who did one, Brunet and colleagues (1998) reported inverse associ - not have a family history, risk was over four times as high (RR = 4.33; 95% CI, 1.65–11.40) in women with a family ations between breast cancer and accumulating 4 or more (OR = 0.47; 95% CI, 0.26– history of breast cancer who reported more than 30 pack- BRCA1 pack-years in carriers of genes (OR = 0.39; 95% CI, 0.10–1.49). BRCA2 0.86) and years of smoking but only about one and one-half times as A subsequent study by the same team of investigators, high in those with a family history who never smoked (RR - = 1.44; 95% CI, 1.21–1.71). In addition, Suzuki and col based on an extended dataset of subjects from 52 centers in 11 countries, failed to replicate this finding (Ghadirian leagues (2007) found a strong dose-response relationship in smokers who had a family history of breast cancer. Risk et al. 2004). Overall, risk of breast cancer from smoking for breast cancer was nearly twice as high in women who in this study was not significantly decreased for carri - had such a family history and accumulated 30 or fewer BRCA2 ers of BRCA1 (OR = 1.09; 95% CI, 0.87–1.33) or pack-years (RR = 1.95; 95% CI, 1.36–2.81) but more than (OR = 0.97; 95% CI, 0.68–1.38), and no trend was observed with lifetime smoking (Ghadirian et al. 2004). However, four times as high in women who had a family history of using a retrospective cohort study design that included a breast cancer and accumulated more than 30 pack-years (RR = 4.33; 95% CI, 1.65–11.40) in comparisons with subset of participants from the same study population as women without a family history who did not smoke. In in Ghadirian and colleagues (2004), Colilla and colleagues (2006) reported a reduced risk of breast cancer among ever contrast, the study did not find an association between BRCA1 smokers with mutation (RR = 0.63; 95% CI, 0.47– smoking and risk for breast cancer among women without Chapter 6 248

278 —50 Years of Progress The Health Consequences of Smoking 0.87) and inverse dose-response relationships for both In a case-only analysis, Moorman and colleagues (2010) reported no significant interactions between ever smoking less than 20 pack-years (RR = 0.72; 95% CI, 0.52–1.00) status. and 20 or more pack-years (RR = 0.41; 95% CI, 0.23–0.71; and BRCA1 or BRCA2 p trend = 0.0007). The study also reported that women Lecarpentier and colleagues (2011) evaluated the BRCA1 mutation and who also had a specific association of smoking and breast cancer in the French with the A1B1 carrier cohort. Sixty-five percent BRCA1/2 National (estrogen receptor coactivator) polymorphism in the gene who accumulated 20 or more pack-years of smoking mutations and of the cohort (863 women) had BRCA1 the remainder (474) had had greatly reduced risk (OR = 0.19; 95% CI, 0.07–0.54). mutations. Among the BRCA2 BRCA1 Ginsburg and colleagues (2009) expanded the num - - carriers, risk was increased among current smok carriers from the study by Ghadirian and ber of ers who reported no alcohol consumption (RR = 2.09; BRCA1/2 colleagues (2004). The reanalysis of the expanded data set 95% CI, 0.94–4.65) but not among those who reported ever use of alcohol (HR = 0.87; 95% CI, 0.52–1.43). This revealed no significant association between risk for breast cancer and ever smoking in carriers of BRCA1 (OR = 1.09; difference between nonusers and ever users of alcohol was 95% CI, 0.95–1.24) or even greater among those with 21 or more pack-years of (OR = 0.81; 95% CI, 0.63– BRCA2 smoking (RR = 3.29; 95% CI, 1.09–9.95 vs. RR = 0.87; 95% 1.05) (Ginsburg et al. 2009). Carriers of BRCA1 who were carriers, there was no sig former smokers, however, had a significantly greater risk CI, 0.45–1.68). Among BRCA2 - nificant increase in risk of breast cancer for either current of breast cancer (OR = 1.27; 95% CI, 1.06–1.50), but no association was found among carriers of BRCA1 who were (RR = 1.39; 95% CI, 0.73–2.63) or former smokers (RR = 1.18; 95% CI, 0.60–2.33), but risk was significantly higher current smokers (OR = 0.95; 95% CI, 0.81–1.12). Total for women who reported 21 or more pack-years (RR = lifetime cigarette consumption was significantly and 2.25; 95% CI, 1.05–4.82). positively associated with breast cancer among carriers In summary, studies of effect modification of smok - who were former smokers (p trend = 0.007). of BRCA1 BRCA1 BRCA2 or ing by on breast cancer have been The study did not find a significant association with for - inconsistent. Two studies reported an inverse association BRCA2 mer smoking in carriers, but current smoking (Brunet et al. 1998; Colilla et al. 2006), four reported no had a nonsignificant inverse association (OR = 0.71; 95% association (Ghadirian et al. 2004; Gronwald et al. 2006; CI, 0.50–1.00). Smoking before the age of 18 was not Nkondjock et al. 2006; Moorman et al. 2010), and one (OR = 1.11; 95% significant in either carrier group: BRCA1 reported a significant positive association (BCFR 2008). (OR = 0.78; 95% CI, 0.59–1.04). BRCA2 CI, 0.89–1.18) or Two studies (Ginsburg et al. 2009; Lecarpentier et al. 2011) Results have been inconsistent in other studies. In a reported positive results for some measures of smoking matched case-control study in Poland, Gronwald and col - but these were inconsistent and difficult to interpret. leagues (2006) reported no association of smoking with For example, Ginsburg and colleagues (2009) reported a carriers (OR = 1.10; 95% CI, BRCA1 breast cancer for mutations who BRCA1 positive association in women with 0.8–1.5). Nkondjock and colleagues (2006) conducted a were former but not current smokers; there were no asso- - nested case-control study in a cohort of 80 French-Cana mutation. As noted BRCA2 ciations in women with the BRCA1/2 was present in 89 breast dian families in which previously, Lecarpentier and colleagues (2011) reported - cancer cases and 48 controls. The study reported no asso positive associations only in women with BRCA1 who ciation with 14 or fewer pack-years (OR = 0.86; 95% CI, reported never using alcohol; risk was only significantly 0.34–2.21) or more than 14 pack-years (OR = 0.74; 95% BRCA2 increased in carriers who reported 21 or more CI, 0.31–1.75) of smoking (p trend = 0.49). However, the pack-years of smoking. Of note, four of these reports were BCFR (2008)—a consortium of research groups in the - based on overlapping participant populations and contra Australia, Canada, and the United States—obtained differ - dictory results (Brunet et al. 1998; Ghadirian et al. 2004; ent results in their case-control study. That study reported Colilla et al. 2006; Ginsburg et al. 2009). increased risks for current smokers with BRCA1/2 muta- tions (OR = 2.08; 95% CI, 1.41–3.06) and in BRCA1 Carcinogen Metabolism BRCA2 carriers (OR = 2.33; 95% CI, 1.56–3.47) and (OR = 2.64; 95% CI, 1.78–3.90). Moreover, in carriers of Researchers have also addressed common polymor - - both mutations, risk of breast cancer increased signifi phisms with low penetrance and small additive or multi - cantly with duration of smoking (approximately 7% per plicative impacts on risk of breast cancer (Pharoah et al. pack-year; p <0.001). Overall, the cumulative incidence of - . With regard to smoking, researchers have consid 2002) BRCA1 breast cancer by 50 years of age for those with a ered common genetic variants in biologic pathways that mutation was about 60% in smokers versus 35% in non- regulate the metabolism and detoxification of tobacco- mutation it was 35% in smok- smokers, and for a BRCA2 related carcinogens (Ambrosone and Shields 1999b; Coyle ers compared with 15% in nonsmokers (BCFR 2008). 249 Cancer

279 Surgeon General’s Report . Thus, a growing number of studies have been 2004) the categorization of rapid activity (NAT2*4, NAT2*12, designed to examine genetic polymorphisms in enzyme NAT2*13), slow activity (NAT2*5, NAT2*6, NAT2*7, systems—such as GST, cytochrome P-450, and NATs. NAT2*14), and intermediate activity (one allele associated with rapid acetylation activity and one with slow activity). Acetyltransferase Very slow activity is associated with being homozygous for Polymorphisms N- NAT2*5 (Hein 2009a). The strongest evidence to date for genetic suscep - In the mid-1990s, Ambrosone and colleagues (1996) tibility to smoking and breast cancer has been for the reported that the association between smoking and breast arylamine NATs, which are enzymes involved in both the cancer was elevated in women with NAT2 slow acetylator detoxification and activation of heterocyclic and aromatic - status, while those with a rapid acetylator status had a non amines (carcinogenic compounds found in cigarette significant decreased risk. This finding was replicated 12 smoke) (Hein 2002). The polymorphisms in the genes years later in a meta-analysis and pooled analysis reported for the NAT1 and NAT2 enzymes are very complex; as a by Ambrosone and colleagues (2008) that in total involved result, past studies have been subject to misclassifica- 4,889 premenopausal and 7,033 postmenopausal women. - tion of the metabolic phenotype, with consequent diffi Women with a history of ever smoking who were slow culty in detecting and interpreting associations. Since the acetylators were at increased risk (vs. never smokers) both first consensus nomenclature was published (Vatsis et al. overall (RR = 1.27; 95% CI, 1.16–1.40) and by menopausal 1995), the classification has become better standardized status (RR = 1.34; 95% CI, 1.17–1.53 for postmenopausal with continuing updates (University of Louisville 2013). and 1.28; 95% CI, 1.09–1.50 for premenopausal) (Table - This improvement has reduced bias in assessing the inter 6.21 S ). No increased risk was reported in women who were action between NAT phenotypes and smoking and has ever smokers and rapid acetylators (RR = 1.05; 95% CI, improved comparisons across studies and the derivation 0.95–1.17). Risk was further increased in slow acetylators of pooled estimates of effects (Deitz et al. 2004). Evidence among those with 20 or more pack-years (meta-analysis clearly indicates that polymorphisms in the NAT2 gene - RR = 1.44; 95% CI, 1.23–1.68), but not in their coun affect the efficiency of the enzyme system in detoxifying terparts who were rapid acetylators (RR = 1.04; 95% CI, carcinogenic amines and that acetylation status (rapid, 0.87–1.25); this pattern was seen for both premenopausal intermediate, slow, and very slow) is correlated with ). The associa- and postmenopausal women (Table 6.21 S - carcinogen metabolism, resulting in activation or deac tion was also present for duration of smoking 15 or more - tivation of xenobiotics (Hein et al. 2000a,b, 2002). In com years in slow acetylators regardless of menopausal status: parisons with rapid acetylator phenotypes, the slow and - premenopausal, RR = 1.35 (95% CI, 1.11–1.65); post very slow acetylator phenotypes have been reported to be menopausal, RR = 1.40 (95% CI, 1.11–1.76) versus never associated with an increased frequency of DNA adducts, smokers. Results from the pooled analysis were consistent a phenomenon that appears to be due to reduced detoxi - with the meta-analysis, with an overall RR summary esti- fication of carcinogenic amines (Pfau et al. 1998; Firozi mate of 1.49 (1.08–2.04) for women with a history of 20 or et al. 2002). Although the prevalence of slow acetylator more pack-years of smoking and the NAT2 slow acetylator status varies across populations, it has been reported to be phenotype compared with never active smokers who had as high as 50–60% in some (Wacholder et al. 2000), with the rapid acetylator phenotype. The interaction of NAT2 evidence for racial/ethnic variation in the frequencies of genotype with smoking was significant for ever smoking NAT2 genotypes (Garcia-Martin 2008). Previous studies of (p = 0.02), pack-years of smoking (p = 0.03), and duration NAT2 have reported associations with other cancers that of smoking (p = 0.007) (Ambrosone et al. 2008). -hydroxyl- N may vary due to activation or inactivation of - Before the publication from Ambrosone and col ated heterocyclic amines. Slow acetylation increases the leagues (2008), 1 summary review and 1 meta-analysis risk for bladder cancer and rapid acetylation increases the reported on the interaction of NAT2 with smoking on risk for colon cancer (Abel and DiGiovanni 2008). risk for breast cancer. Terry and Goodman’s (2006) meta- Several studies have evaluated the associations of analysis was based on 13 studies and reported an increased polymorphisms with breast cancer and NAT1 and NAT2 risk for breast cancer among postmenopausal women who many of these have examined interactions with smoking. smoked and were classified as slow acetylators (Table Only a few studies have examined NAT1 (Millikan et al. 6.21 S ). Ochs-Balcom and colleagues’ (2007) review of 12 1998; Krajinovic et al. 2001; Lee et al. 2003; van der Hel studies also found evidence that modified risk for NAT2 et al. 2003b; Zheng et al. 1999), as the majority of studies breast cancer among women who smoked. A recent meta- . Even with standardization, con- NAT2 have focused on analysis by Zhang and colleagues (2010) provided results tinuous updates have been made with the identification NAT2 for the association of with breast cancer modified by of new alleles. Currently, acetylation status is based on 250 Chapter 6

280 —50 Years of Progress The Health Consequences of Smoking smoking rather than modification by NAT2 of the associa- - genic intermediates generated in this pathway can dam age DNA gene encodes CYP1A1 (Sillanpaa et al. 2007) . The tion of smoking with risk of breast cancer. As such, the esti - a Phase I enzyme that contributes to aryl hydrocarbon mates from this meta-analysis cannot be compared with hydroxylase activity and metabolism of PAHs, which have previous findings. Zhang and colleagues (2010) extracted - been detected in both normal and cancerous breast tis data from studies to recalculate ORs for the main effects modified by pack-years of smoking, but NAT2 CYP1B1 . (Terry and Rohan 2002; Masson et al. 2005) sues of NAT2 and - in doing this, they could not take into account covariates is involved in estrogen homeostasis in normal breast tis (Rylander-Rudqvist from original analyses for the effect of smoking modified sue and is expressed in breast tumors . et al. 2003) by NAT2 . Nonetheless, a significant interaction was found. Studies have not documented an interaction of Taken together, the results of these meta-analyses suggest genotype modifies the risk for breast cancer genotypes smoking and polymorphisms in these NAT2 that the CYP in women who smoke. In addition, there is an increased on risk for breast cancer. Masson and colleagues (2005) reviewed five studies with data on the interaction of smok - risk of about 40–50% in women who have the NAT2 slow polymorphisms on risk for breast cancer CYP1A1 ing and acetylation phenotype who smoke. - (Ambrosone et al. 1995; Bailey et al. 1998; Ishibe et al. Two studies have been published since the compre 1998; Taioli et al. 1999; Basham et al. 2001), but only one hensive meta-analysis from Ambrosone and colleagues (Ambrosone et al. 1995) provided evidence for a possible (2008). In a case-control study (717 cases and 735 con - interaction, and a formal statistical test was not conducted trols) of Hispanic and non-Hispanic White women in New in that study. Furthermore, results from these studies are Mexico, Baumgartner and colleagues (2009) reported an difficult to interpret because of their small samples and interaction between a history of ever smoking and the differences in reference groups, categories of smoking, phenotype that approached significance in non- NAT2 Hispanic White women only (p for interaction = 0.06). and definition of interactions. Terry and Goodman (2006) The risk estimate (OR) for ever smokers with the very conducted a meta-analysis of four studies (Ambrosone et slow phenotype was 2.57 (95% CI, 1.49–4.41). In this al. 1995; Ishibe et al. 1998; Basham et al. 2001; Li et al. 2004), three of which (all but Li et al. 2004) were reviewed study, risk was increased similarly in former and current by Masson and colleagues (2005). The summary estimate smokers with the very slow phenotype. In Germany, Rab - stein and colleagues (2010) reported results for a case- among smokers with the wild-type genotype (OR = 1.3; control study involving 1,155 cases and 1,143 controls. 95% CI, 1.0–1.6) did not differ significantly from those The study did not find an interaction between smoking and with variant alleles (OR = 1.2; 95% CI, 0.6–2.1), suggest - the NAT2 phenotype, even when results were stratified by ing no interaction. ER phenotype. CYP1B1 Studies of the interaction between poly- morphisms and smoking on risk for breast cancer have - Finally, a report from the Breast and Prostate Can produced mixed results. Saintot and colleagues (2003) cer Cohort Consortium (Cox et al. 2011) pooled data for reported increased risk for breast cancer among former - 6,900 cases and 9,903 controls from seven separate stud smokers (OR = 1.33; 95% CI, 0.59–2.96) and current ies (CPS-II/1998, NHS-I/1989 and NHS-II/1999, EPIC CYP1B1 smokers (OR = 2.32; 95% CI, 1.00–5.38) with the 1992, Multi-Ethnic Cohort Study/1996, Prostate, Lung, LEU/LEU genotype compared with nonsmokers with VAL Colorectal and Ovarian Cancer Screening Trial/1993, and alleles. In contrast, Rylander-Rudqvist and colleagues Women’s Health Study/1993). A significant interaction (2003) reported no association between smoking and any was not found between duration or pack-years of smoking CYP1B1 genotype on risk for breast cancer. The case-con - and the NAT2 acetylation phenotype. Risk of breast cancer trol study conducted by Sillanpaa and colleagues (2007) was increased in those with more than 20 pack-years of reported unstable findings because of small samples in smoking and fast acetylation status, which was defined as some strata: for example, risk was increased significantly a combination of rapid and intermediate phenotypes (OR among smokers who consumed 1–9 cigarettes per day and = 1.24; 95% CI, 1.08–1.42), as well as in slow acetylators CYP1B1 VAL allele (OR = 2.63; 95% (a) were carriers of the (OR = 1.25; 95% CI, 1.11–1.39). Adjustment included a genotype (OR = 5.09; VAL/VAL CI, 1.07–6.46) or (b) had the number of covariates, but not the use of alcohol. This 95% CI, 1.30–19.89; p trend = 0.005), but these increased report weakens the evidence for as an effect modi- NAT2 risks were not observed in women who smoked more than fier of smoking on the risk of breast cancer. 10 cigarettes per day. Results for duration of smoking and pack-years of smoking were also contradictory. Polymorphisms Cytochrome P-450 Sillanpaa and colleagues (2007) also reported a sig - CYP1B1 CYP1A1 and are gene-encoding enzymes nificant increased risk for breast cancer in smokers with involved in the metabolism of estradiol and PAHs. Muta- VAL allele who were NAT2 slow acetylators the CYP1B1 251 Cancer

281 Surgeon General’s Report , which was not included Mitrunen and colleagues (2001a) (OR = 2.46; 95% CI, 1.11–5.45), suggesting a potential three-way interaction between smoking, CYP1B1, and in the pooled analysis by Vogl and colleagues (2004), did NAT2 . Van Emburgh and colleagues (2008b) reported a not detect any interaction between a history of smoking GSTM1, GSTM3, GSTP1, or GSTT1 genetic and either significant interaction (p = 0.02) between smoking and - 119S CYP1B1 the polymorphisms. Subsequent studies have not reported allele on risk for breast cancer in Afri can Americans but not in Whites. Taken together, these GST significant interactions between polymorphisms studies do not provide strong or consistent evidence for and smoking on risk for breast cancer (Linhares et al. modification of risk for breast cancer from smoking by 2005; Ahn et al. 2006; Olsen et al. 2008; Van Emburgh polymorphisms in genes for the CYP enzyme system. et al. 2008b; McCarty et al. 2009; Andonova et al. 2010) . Thus, with the possible exception of GSTM1, the evidence to date does not support modification of the breast can Glutathione S-transferases - cer–smoking association by polymorphisms in the GST GSTs are Phase II enzymes that metabolize and enzyme system. - detoxify endogenous and exogenous substances, includ ing tobacco smoke carcinogens—specifically PAHs (Terry Sulfotransferase 1A1 and Goodman 2006) . DNA adducts are more common in - SULT enzymes activate or inactivate PAHs and het- smokers with breast cancer who have certain polymor (van der Hel et al. phisms in genes for the GST enzymes erocyclic amines from cigarette smoke through sulfonate (ARG213HIS) conjugation. A common polymorphism 2003b) . The GST enzyme system contains eight families of - genes, and polymorphisms have been described in several in SULT1A1 results in reduced enzyme activity and effi (Terry and Goodman 2006) . Only of these families—mainly mu (M1), theta (T1), and pi (P1) ciency of this pathway (Vogl et al. 2004; Terry and Goodman 2006) GSTM1 and three studies to date have examined interactions between . are deletion ( GSTT1 this polymorphism and smoking on risk for breast can - null ) polymorphisms that result in the absence of protein expression. (Saintot et al. 2003; Lilla et al. 2005; Sillanpaa et al. cer Saintot and colleagues Terry and Goodman (2006) performed a meta- . The case-only study by 2005b) (2003) suggested interactions between the allele and HIS analysis of seven studies (Ambrosone et al. 1995; Garcia- both duration of smoking (>20 years) (OR = 1.71; 95% Closas et al. 1999; Millikan et al. 2000; Zheng et al. 2002a,b; CI, 0.97–3.03) and intensity of smoking (>5 cigarettes/ - van der Hel et al. 2003b, 2005) that investigated the poten day) (OR = 1.65; 95% CI, 0.97–2.80). In contrast, two sub - GSTM1 tial modification by of the association GSTT1 and sequent case-control studies did not find evidence of an between smoking and risk for breast cancer. Six studies (Lilla et al. interaction between SULT1A1 and smoking were population-based or nested case-control designs and . 2005; Sillanpaa et al. 2005b) one was a case-cohort study. Using categories for longest duration of smoking, the RRs from the meta-analysis were Oxidative Metabolism Genotypes 1.4 (95% CI, 1.1–1.9) for versus 1.10 (95% CI, GSTM1null 0.80–1.40) for suggesting possible effect GSTM1present, Smoking is associated with increased oxidative modification. In contrast, smoking was associated with stress (Pryor and Stone 1993), and superoxide dismutase genotype: GSTT1 breast cancer regardless of GSTT1null 2 (SOD2) is a mitochondrial enzyme that protects against (meta-RR = 1.20; 95% CI, 0.90–1.70) and GSTT1present oxidative stress. A common polymorphism in the gene for (meta-RR = 1.30; 95% CI, 1.10–1.60). - SOD2 reduces the activity of this enzyme and is report Several studies have examined the main effects of edly associated with several cancers, including breast GST polymorphisms on risk for breast cancer stratified by cancer (Millikan et al. 2004; Gaudet et al. 2005). Terry smoking status. Although these studies did not provide and Goodman (2006) reviewed four case-control studies estimates by genotype for modification of the association on the modification of risk for breast cancer by smoking between smoking and breast cancer, many included tests and SOD2 (Mitrunen et al. 2001b; Millikan et al. 2004; for interaction that can be interpreted as evidence that a - Tamimi et al. 2004; Gaudet et al. 2005); in one of the stud Vogl and colleagues polymorphism alters this association. ies, Millikan and colleagues (2004) reported a significant pooled results from seven case-control studies (2004) increased risk of breast cancer for smoking duration of (Bailey et al. 1998; Maugard et al. 1998; Nedelcheva et al. more than 20 years in women homozygous for the vari- 1998; Ambrosone et al. 1999a; Zhao et al. 2001; da Fonte ant ALA allele (OR = 1.5; 95% CI, 1.0–2.2). However, an and found no de Amorim et al. 2002; Zheng et al. 2002b) increased risk for ever smokers who were homozygous for evidence of significant interaction between smoking and allele (OR = 2.6; 95% CI, 1.1–6.3) was VAL the wild-type GSTM1, GSTT1, or GSTP1 polymorphisms. A study by reported (as calculated by Terry and Goodman [2006] for Chapter 6 252

282 —50 Years of Progress The Health Consequences of Smoking the study by Gaudet and colleagues [2005]). Results from Since 2000, several studies have evaluated SNPs in gene (Colilla et al. the other two studies were null. The overall meta-RR esti - the nuclear receptor coactivator AIB1 mate for the four studies was 1.5 (95% CI, 1.1–2.1). Only 2006), the IGHMBP2 gene (Shen et al. 2006), the A-T gene genes (Yang NOS3 two other case-control studies have been published since MPO and (Swift and Lukin 2008), the this review (Slanger et al. 2006; Kostrykina et al. 2009); gene (de Assis et al. 2002) for et al. 2007), and the mEH neither found significant interactions between SOD2 and - interaction with smoking on risk of breast cancer. How ever, the results have been either null or indicated only smoking or main effects of SOD2 or smoking on risk for breast cancer. weak associations. None of these studies have been rep - licated to date. Additionally, three studies evaluated the association between smoking and DNA Repair Genes mutational status p53 as a measure of apoptosis (Conway et al. 2002; Furberg Terry and Goodman (2006) reviewed seven stud- et al. 2002; Gaudet et al. 2008). A recently published anal - ies with data on modification of risk for breast cancer by ysis of more extensive data from the Long Island Breast - smoking and DNA repair genotypes, including polymor Cancer Study Project suggested that cigarette smoking , and . Five studies, which phisms in XRCC1 MGMT , XPD and passive smoking were more strongly associated with included two or three different polymorphisms in XRCC1 p53-negative cancer (Mordukhovich et al. 2010), which 399 194 280 ( ARG TRP, AND ARG GLN, ARG HIS ) and widely contrasts with results reported by Conway and colleagues different smoking exposures (ever smoking, duration >20 (2002), Van Emburgh and colleagues (2008a), and an ear - years, >5 pack-years of smoking), produced inconsistent lier analysis of the Long Island study (Gaudet et al. 2008). results (Duell et al. 2001; Metsola et al. 2005; Patel et al. 2005; Shen et al. 2005a; Pachkowski et al. 2006). The Genetic Susceptibility—Summary meta-analytic summary estimate for smoking exposure 194 ARG/ The epidemiologic studies conducted to date have was significant only for women homozygous for not established clear or consistent evidence for modifica- ARG . Two studies of the XPD LYS751GLN polymorphism tion of the association between smoking and breast cancer reported nonsignificant increased risks for smokers with genotype (as calculated by Terry and Good GLN/GLN by genes that influence susceptibility to tobacco-related - the man [2006] for the studies by Terry and colleagues [2004] carcinogens. The published reports support only genetic as a potential effect modifier of the asso NAT2 variation in and Metsola and colleagues [2005]). A study by Shen and - ciation of breast cancer with smoking, although this find - - colleagues (2005b) reported increased risk in heavy smok MGMT LEU84PHE and ILE143VAL polymor - ing has been weakened by the recent report of Cox and ers with colleagues (2011). Unfortunately, a variety of limitations phisms. In the NHS-I cohort, Han and colleagues (2003) have affected these studies. First, many have been too small to provide adequate statistical power for detecting found no evidence for effect modification of smoking by 194 399 GLN, TRP, C26602T, ARG interactions between smoking and low-frequency geno - any of four SNPs ( ARG 632 and XRCC1. ) in GLN Subsequently, Han and col- GLN Terry and Goodman (2006) types. reported that statistical leagues (2004) reported no interaction between smoking power was less than 80% for detecting a risk estimate of at genes, and and SNPs in the XRCC2, XRCC3, and LIG IV least 2.0 for breast cancer for the majority (68%) of stud - Han and colleagues (2006) did not report such an inter - ies in their review. In addition, the definitions of smoking gene. Shore and colleagues (2008) MGMT action in the - exposure have varied widely across studies, making it dif reported an interaction between smoking and a SNP in ficult to combine estimates in meta-analyses. Most stud - XPC the gene that approached significance (p = 0.08) in ies have tested only a limited number of selected SNPs in the NYU Women’s Health Study. Mechanic and colleagues specific groups of candidate genes, targeting mainly those (2006) found that the combination of smoking and four that influence carcinogen metabolism, oxidative stress, or more SNPs in several nucleotide excision repair genes or DNA repair. Not all of these studies have established significantly (XPD, XPC, RAD23B, XPG, XPF, and ERCC6) the functionality of SNPs. Only a few studies have ana - modified the risk for breast cancer in African American, lyzed interactions of smoking with haplotype combina - but not White, women. Similarly, Metsola and colleagues tions of SNPs within or across genes. Investigators will (2005) found strong evidence for modification of the asso - likely continue to examine this important area of research ciation between smoking and the combination of two by combining genomewide association studies with gene or more SNPs in XRCC1 and XPD on the risk for breast expression assays to identify functional gene variants that cancer. Future studies should emphasize interactions . modify susceptibility to smoking (Chung et al. 2010) between smoking and combinations of SNPs within and across genes (Neumann et al. 2005). 253 Cancer

283 Surgeon General’s Report The use of a no active/no passive exposure referent appears Summary and Review of Active to have a small impact on most summary estimates, but Cigarette Smoking this can be difficult to interpret because it results in a very small reference group and a loss of statistical power. - The 2004 Surgeon General’s report on active ciga Future studies need to determine whether statistical rette smoking concluded that there was (a) no consistent adjustment for exposure to passive smoking is adequate. evidence for an association between active smoking and This may require stronger techniques and methods of breast cancer, and that (b) subgroups of women could not measuring exposure to secondhand smoke. be reliably identified that were at increased risk of breast cancer due to smoking. Since the previous report, 12 cohort and 30 case-control studies have been published Major Summary Points for Active on the association of smoking with breast cancer. Several large cohort studies now provide consistent evidence for a Smoking significant, although weak, positive association. While the findings from the case-control studies are more variable, Based on 22 cohort reports published prior to 2012 1. when considered together the results are in keeping with and 27 case-control reports published from 2000– - those from the cohort studies. The meta-analyses con - 2011, evidence suggests that a history of ever smok - firm a weak but statistically significant, positive associa ing is associated with an increase in the RR for breast tion of smoking with risk of breast cancer. The estimates cancer by an average of 10%; long duration of smok- for active smoking tend to be higher when based on data ing (20 or more years), greater number of cigarettes - from case-control studies than on data from cohort stud smoked per day (20 or more), and more pack-years of ies; but there is greater heterogeneity among estimates smoking (20 or more) significantly increase risk for from case-control studies. Sensitivity analyses reveal that breast cancer by 13–16%, depending on study design this heterogeneity is largely related to issues in the design and the exclusion of studies with design or analysis or analysis of certain studies. When these studies are issues. removed, the summary estimates from the case-control studies converge to agreement with those from the cohort Studies have not clearly determined whether either 2. - studies. The sensitivity analyses also suggest that the posi early age at smoking initiation or smoking before first - tive association of smoking with breast cancer is statisti pregnancy is associated with increased risk for breast cally robust. cancer over and above the risk due to ever smoking. Ever smoking is associated with a significant increase in RR of about 10% (Table 6.17 S ). The magnitude Studies have not clearly determined whether the use 3. - of the association appears to be slightly stronger for cur of a restricted no active/no passive exposure reference rent smoking (12%) than for former smoking (9%). It is - group or adjustment for exposure to passive smok increased by 16% for duration of 20 or more years, 13% ing meaningfully alters or clarifies the association for smoking 20 or more cigarettes per day, and 16% for between smoking and risk for breast cancer. accumulating 20 or more pack-years. There is no clear evidence that earlier age at initiation of smoking (8%) or The extent to which the use of alcohol confounds the 4. smoking before first pregnancy (10%) is associated with a association between smoking and risk for breast can - greater risk of breast cancer than is ever smoking. There is cer remains uncertain and should be considered in evidence, based on the most conservative combined study relation to the duration, dose, and timing of smoking. design estimates, that among ever smokers, premeno - pausal women have a slightly higher increase in risk than 5. There is emerging evidence to suggest that the risk postmenopausal women, 17% versus 7%, respectively - of breast cancer from smoking may be greater in pre ). It remains to be established whether smok S (Table 6.18 - menopausal than postmenopausal women, 17% ver - ing is more strongly associated with a particular tumor sus 7%, or a relative difference of 9%. phenotype. There is no consistent evidence to date that subpopulations of women with genetic susceptibility to There is insufficient evidence to conclude that the 6. tobacco-related carcinogens (even , given the most NAT2 risk of breast cancer from smoking differs between recent report by Cox and colleagues [2011]), can be reli - - women diagnosed with ER+ tumors and those diag ably identified as being at increased risk for breast cancer . nosed with ER– tumors. Chapter 6 254

284 —50 Years of Progress The Health Consequences of Smoking With the possible exception of the polymorphism in risks for breast cancer. These and several subsequent 7. the NAT2 carcinogen metabolism pathway, subgroups studies had limitations, however, such as mixing inci - of women who are at increased risk of breast cancer dent and prevalent cases with breast cancer deaths; using because of the interaction between smoking and gen- - proxy reports; having limited data for duration, dose, loca otype cannot be identified reliably. tion, and timing of exposure; and adjusting inadequately for relevant confounders. Palmer and Rosenberg (1993) cited only the reports from Hirayama (1984), Sandler and colleagues (1986), and Wells (1991); the latter was a re- Exposure to Secondhand Smoke analysis of the data from the studies by Hirayama (1984) and Risk for Breast Cancer and Sandler and colleagues (1985a). They concluded that “so little research” had been conducted that it was “not Compared with directly inhaled tobacco smoke or possible to reach any conclusions” (Palmer and Rosenberg mainstream smoke, the evidence indicates that undi- 1993, p. 152). - luted sidestream smoke, the major contributor to sec Several meta-analyses and monographs about pas- ondhand smoke (passive smoke, involuntary smoking, sive smoking and breast cancer have been published or environmental tobacco smoke [ETS]), contains higher released, some not long before or after the 2006 Surgeon - levels of several substances considered to be carcino General’s report (Khuder and Simon 2000; Khuder et al. genic, cocarcinogenic, or toxic—including benzene, 2001; Morabia 2002a; Cal/EPA 2005; Johnson 2005; Lee formaldehyde, catechol, and nitrosamines (IARC 2004; N- and Hamling 2006; Nagata et al. 2006; Pirie et al. 2008; USDHHS 2010). Measuring exposure to secondhand Collishaw et al. 2009; Secretan et al. 2009). The authors smoke for assessment of cancer risk poses challenges, of these studies have drawn markedly different interpre- however, because an ideal comprehensive assessment tations and conclusions, despite considerable overlap should address duration of exposure, dosage (exposure among some of these reports in the studies reviewed and time, number of people who smoke in the immediate evaluated through meta-analysis. environment, number of cigarettes smoked by smokers, Khuder and Simon (2000) published one of the first ventilation), location of exposure (home, workplace), time systematic reviews of passive smoking and risk for breast period of exposure (childhood, adulthood), and method of cancer. That review examined 11 reports (3 cohort and 8 assessing exposure (self-report, biologic specimen). Other case-control) that were published between 1984 and 2000 relevant issues include the pervasiveness of secondhand (Hirayama 1984; Sandler et al. 1986 [based on Sandler smoke in the environment, particularly in the past in the - et al. 1985a]; Smith et al. 1994; Morabia et al. 1996; John United States and some other Western countries, changes son et al. 1998, 2000; Jee et al. 1999; Lash and Aschengrau in intensity over time, measurement error, and informa - 1999; Liu et al. 2000; Marcus et al. 2000; Wartenberg et al. tion bias that may dilute estimates of association (Kawachi 2000). Two of the three cohort studies examined breast and Colditz 1996). Methodologic issues in investigating cancer mortality (Hirayama 1984; Wartenberg et al. 2000), secondhand smoke and disease risk were addressed in the and one was reported as an abstract (Johnson et al. 1998). 2006 report of the Surgeon General. Despite strong evi- Results were summarized using the random-effects model. - dence from cotinine levels of declining exposure to sec The summary estimate of the RR for ever being exposed to ondhand smoke in the United States, there is no level of secondhand smoke was 1.41 (95% CI, 1.14–1.75). Based exposure considered to be risk free (USDHHS 2006), and on their results, Khuder and Simon (2000) suggested high levels of exposure persist for some groups (Chen et a “possible weak association between passive smoking al. 2010a). and breast cancer” (p. 1117) and that more studies were Exposure to secondhand smoke has been investi- needed. Morabia (2002a) also reviewed the associations gated as a risk factor for breast cancer over nearly three between passive smoking, as well as active smoking, decades. Sandler and colleagues (1985a) first evaluated and breast cancer. This review considered most of the the association between passive smoking exposure and same studies assessed by Khuder and Simon (2000) but breast cancer in the mid-1980s in a small hospital-based did not calculate a summary estimate. Instead, Morabia case-control study in North Carolina. In the early 1990s, (2002a) noted that ORs were greater than 1.5 in 5 of the Wells (1991) analyzed data from Hirayama’s large Japanese 11 case-control studies he reviewed and emphasized the cohort study (Hirayama 1984, 1990), which was initiated importance of separating passive from active exposures in in 1965. Both studies found nonsignificantly increased future studies. 255 Cancer

285 Surgeon General’s Report the workplace and on levels of at-home smoke exposure by The 2004 IARC monograph reviewed results from 5 cohort and 10 case-control studies and concluded that the number of cigarettes smoked per day (≤2, 3–9, 10–19, ≥20) “collective evidence on breast cancer risk associated with (Liu et al. 2000). In contrast to the estimated quadrupling of risk in the Cal/EPA report, the pooled risk estimate for involuntary exposure of never smokers to tobacco smoke is inconsistent” (p. 1410). The monograph emphasized adult home exposure was 1.47 (95% CI, 0.74–2.95) (Liu et al. 2000); this estimate was used in the meta-analysis in results from the NHS-I (Egan et al. 2002) and the CPS-II the 2006 Surgeon General’s report. Additionally, the 2006 (Wartenberg et al. 2000), noting that these large cohort studies “provided no support for a causal relation between Surgeon General’s report included the study by Bonner - involuntary exposure to tobacco smoke and breast can and colleagues (2005) that was published after the period of inclusion for studies in the Cal/EPA report had passed. cer in never smokers,” that the “lack of a positive dose- response also argue[d] against a causal interpretation of This study reported a significant inverse association for these findings,” and that “the lack of an association of exposure at the workplace (calculated pooled OR = 0.79; breast cancer with active smoking weighs heavily against 95% CI, 0.65–0.96) but no significant effect for exposure at home (calculated pooled OR = 1.16; 95% CI, 0.96–1.41). the possibility that involuntary smoking increases the risk In a meta-analysis by Johnson (2005) of the asso- for breast cancer, as no data are available to establish that ciation between passive and active smoking and breast different mechanisms of carcinogenic action operate at - the different dose levels of active and of involuntary smok cancer, the analysis for passive smoking was based on - 19 studies (7 cohort and 12 case-control) that met spe ing” (IARC 2004, p. 1410). cific quality criteria for study design and exposure mea In contrast, a report from 2005 about secondhand - smoke as a toxic air contaminant (Cal/EPA 2005), which surement (Hirayama 1984; Sandler et al. 1985a; Smith et al. 1994; Morabia et al. 1996; Millikan et al. 1998; Jee was also summarized by Miller and colleagues (2007), included an extensive section about breast cancer in which et al. 1999; Lash and Aschengrau 1999, 2002; Zhao et al. - it noted that “the weight of evidence (including toxicol 1999; Delfino et al. 2000; Johnson et al. 2000; Wartenberg et al. 2000; Nishino et al. 2001; Egan et al. 2002; Kropp ogy of ETS [environmental tobacco smoke] constituents, and Chang-Claude 2002; Shrubsole et al. 2004; Gammon epidemiological studies, and breast biology) is consistent et al. 2004a; Reynolds et al. 2004b; Hanaoka et al. 2005). with a causal association between ETS exposure and breast cancer in younger, primarily premenopausal women” These studies were mostly the same as those included in (Cal/EPA 2005, p. ES8). The pooled RR estimate was 1.68 the 2005 Cal/EPA report and the 2006 Surgeon General’s - report. The summary pooled risk estimate for all 19 stud - (95% CI, 1.31–2.15), based on a meta-analysis of 14 stud ies reporting risk for breast cancer among never-smoking ies using the broadest definition of passive smoking was premenopausal women who reported exposure to passive 1.27 (95% CI, 1.11–1.45; test for heterogeneity p <0.001). smoking. However, the overall test for heterogeneity was The broadest definition of passive smoke exposure in most - significant (p = 0.001), suggesting substantial inconsis studies included the following: exposure from any source, including husband’s smoking history; years smoked by tency across studies. When the analysis was restricted to 5 studies (Smith et al. 1994; Morabia et al. 1996; Zhao et al. spouse; lifetime residential childhood exposure; work - place exposure; and parental exposure. As in the Cal/EPA 1999; Johnson et al. 2000; Kropp and Chang-Claude 2002) report, 5 case-control studies strongly influenced the with what was considered “better exposure assessment” (Cal/EPA 2005, p. ES-3), the pooled RR estimate was 2.20 summary of pooled risk estimate (Smith et al. 1994; Mora- bia et al. 1996; Zhao et al. 1999; Johnson et al. 2000; Kropp (95% CI, 1.69–2.87), and a test for heterogeneity was not and Chang-Claude 2002), because they were considered significant (p = 0.354). to have the most complete assessments of exposure. The The Cal/EPA report differed from the 2006 Surgeon summary pooled risk estimate (RR) for these 5 studies was General’s report with respect to two studies. The Cal/EPA 1.90 (95% CI, 1.53–2.37). In contrast, the summary RR excluded the study by Liu and colleagues (2000) because was 1.16 (95% CI, 0.95–1.42) for the remaining 7 case- the panel found that the results were difficult to interpret control studies (those considered to have less complete as the study was clinic based and small (n = 186 cases) and assessments of exposure). The summary estimate for the - reported results based on a passive smoking index (num 7 cohort studies was 1.06 (95% CI, 0.97–1.16). Johnson ber of smokers times smoke exposure levels, defined as (2005) also calculated summary estimates for risk of light, medium, or very heavy). The estimate of breast can- cer risk for adult home exposure based on this index was breast cancer among premenopausal women by using data RR = 4.07 (95% CI, 2.21–7.50) (Liu et al. 2000). However, from 14 of the 19 studies. The overall summary estimate the 2006 Surgeon General’s report included estimates was higher for premenopausal women (RR = 1.68; 95% CI, from Liu based on number of smokers exposed to smoke in 1.33–2.12; p = 0.002 for heterogeneity) than for all women 256 Chapter 6

286 —50 Years of Progress The Health Consequences of Smoking to secondhand smoke. The analysis included all 21 stud and was highest for the 5 studies (as a group) considered - to have the most complete assessment of exposure (RR = ies from the 2006 Surgeon General’s report and 4 other 2.19; 95% CI, 1.68–2.84). Johnson (2005) did not calculate studies—2 case-control studies (Lissowska et al. 2006; summary estimates by timing, source, duration, or dose of Roddam et al. 2007), 1 cohort study on mortality (Sagiv exposure to passive smoking. The author concluded that et al. 2007), and results from the Million Women Study, a cohort study in the United Kingdom (Pirie et al. 2008). “studies with thorough passive smoking exposure assess - Overall, data reported for the cohort studies indicated no ment implicate passive and active smoking as risk factors for premenopausal breast cancer” but that more cohort association with breast cancer (RR = 0.99; 95% CI, 0.93– studies with thorough exposure assessments were needed 1.05), but data reported for the case-control studies noted (Johnson 2005, p. 619). a significant association (OR = 1.21; 95% CI, 1.11–1.32; p Lee and Hamling (2006) conducted a systematic <0.0002). When based on data for the cohort studies, h results reported by Pirie and colleagues (2008) for expo - review and meta-analysis of 22 studies (13 case-control, sure to passive smoking as a child and as an adult were 8 prospective cohort, and 1 nested case-control) involv- ing nonsmoking women that were published through identical (RR = 1.00; 95% CI, 0.94–1.07). Analyses were June 2005. RR estimates that adjusted for the greatest not stratified on menopausal status or source or location number of confounding variables for exposure to second - of exposure, as they were in the 2006 Surgeon General’s hand smoke at home, at the workplace, during childhood, report. Conclusions were strongly influenced by results during adulthood, or during lifetime were used when from the cohort studies: “In aggregate little or no adverse available. Results of the meta-analysis included several effect on the risk of breast cancer” was evident, and the subgroup variables from the studies—including meno - results based on the case-control studies “appear[ed], in aggregate, to be misleading” (Pirie et al. 2008, p. 1,077). pausal status (n = 11), the woman’s age or the age of The 2009 Canadian Expert Panel on Tobacco Smoke husband (n = 4), and genotype (n = 5). Results were also and Breast Cancer Risk—based primarily on its updated stratified by location, source, or timing of exposure: home review of four studies published in 2005 or later (Bon- (n = 19), workplace (n = 5), childhood (n = 9), spouse ner et al. 2005; Lissowska et al. 2006; Roddam et al. 2007; (n = 8), and lifetime (n = 6). A sensitivity analysis removed Pirie et al. 2008), previous reports by the Cal/EPA, and studies that adjusted for fewer than nine covariates but resulted in little inflation of the RR—from 1.23 (95% CI, the 2006 Surgeon General’s report—concluded that “the - 1.03–1.45) to 1.28 (95% CI, 1.07–1.53). Overall, this meta- relationship between secondhand smoke and breast can cer in younger, primarily premenopausal women is con - analysis was similar to the one reported in the 2006 Sur - sistent with causality” but determined that evidence was geon General’s report, although it excluded the study by insufficient for a conclusion on risk of postmenopausal Zhao and colleagues (1999) and did not include the study breast cancer (Collishaw et al. 2009, p. 57). In its special by Bonner and colleagues (2005), which was reported after report from November 2009 that included an assessment its publication. The review by Lee and Hamling (2006) of exposure to secondhand smoke, IARC concluded that also included two abstracts (Rookus et al. 2000; Woo et al. “evidence for female breast cancer remains inconclusive” - 2000) and a cohort study reported on by Gram and col (Secretan et al. 2009, p. 1,033). leagues (2005). The results were similar to those reported in the 2006 Surgeon General’s report: a nonsignificant summary estimate based on 9 cohort studies (RR = 1.02; 95% CI, 0.93–1.10), a significant summary estimate based Conclusions from Previous Surgeon on 13 case-control studies (RR = 1.28; 95% CI, 1.07–1.53), General’s Reports and a significant increased risk for breast cancer among premenopausal women based on 10 studies (RR = 1.54; The 1986 Surgeon General’s report was the first 95% CI, 1.16–2.05), but with significant heterogeneity to offer a conclusion on passive smoking and cancer, (p <0.01). Additionally, risk estimates for small studies but given available evidence it addressed only lung can - (<500 cases) were higher (RR = 1.27; 95% CI, 1.03–1.57) cer (USDHHS 1986). This report also concluded that the and showed significant heterogeneity compared with large effects of passive exposure were likely not greater than studies (≥500 cases) (RR = 1.01; 95% CI, 0.93–1.09). Lee those effects seen for smokers, echoing a similar conclu- and Hamling (2006, p. 1,068) noted that “one cannot con- sion of IARC Monograph 38 of WHO (IARC 1986). fidently conclude, based on the evidence available, that The 2006 Surgeon General’s report concluded that ETS exposure increases risk in nonsmokers.” the evidence on exposure to secondhand smoke was “sug- Pirie and colleagues (2008) conducted a meta-anal- gestive but not sufficient to infer a causal relationship” ysis of 8 cohort and 17 case-control studies on exposure 257 Cancer

287 Surgeon General’s Report with risk for breast cancer (p. 480), based on a review of versus cohort studies. The majority of case-control studies have reported positive associations, with summary esti- 7 prospective cohort studies (Hirayama 1984, reanalyzed by Wells [1991]; Jee et al. 1999; Wartenberg et al. 2000; mates (RRs) ranging from 1.2–1.9 depending on the stud - ies included. Results from cohort studies have mostly been Nishino et al. 2001; Egan et al. 2002; Reynolds et al. 2004b; Hanaoka et al. 2005) and 15 case-control studies (Sandler null. Compared with cohort studies, case-control studies - et al. 1985a; Smith et al. 1994; Morabia et al. 1996; Mil often include more extensive and rigorous assessments of exposure—including detailed information for tim- likan et al. 1998; Lash and Aschengrau 1999, 2002; Zhao ing (childhood, adulthood), location (home, workplace), et al. 1999; Delfino et al. 2000; Johnson et al. 2000; Liu source (parent, spouse, other), duration, and dose—but et al. 2000; Marcus et al. 2000; Kropp and Chang-Claude 2002; Shrubsole et al. 2004; Gammon et al. 2004a; Bonner these studies are more susceptible to information bias et al. 2005). In the 2006 report, pooled risk estimates were and generally considered less reliable. In addition, most of derived for all women and stratified by menopausal status the case-control studies published before 2006 were small (<100 cases) or moderate (<500 cases) in size and had and categories related to timing (childhood, adulthood), imprecise estimates. The likelihood of extreme estimates source (spouse), and location (home, workplace) of expo- is increased in small studies and leads to significant het- sure. The overall risk estimate (RR = 1.20; 95% CI, 1.08– 1.35) was based on the most comprehensive measure of erogeneity across studies. In any case, all of the previous exposure to secondhand smoke. Data from cohort studies reviews have concluded that more and larger studies are needed, particularly those with cohort designs, with more indicated no association (RR = 1.02; 95% CI, 0.92–1.13) with breast cancer, but the summary estimate from case- detailed and extensive assessments of exposure. control data showed a significant association (OR = 1.40; 95% CI, 1.17–1.67). The association was particularly strong Cohort Studies for premenopausal women (OR = 1.64; 95% CI, 1.25–2.14), - The 2006 Surgeon General’s report covered 21 stud based on estimates from 2 cohort studies (Reynolds et al. ies, identified through 2005, on the health consequences 2004b; Hanaoka et al. 2005) and 9 case-control studies of involuntary exposure to tobacco smoke. From 2006– (Sandler et al. 1985a; Smith et al. 1994; Morabia et al. 1996; 2011, 7 cohort studies have evaluated exposure to passive Millikan et al. 1998; Delfino et al. 2000; Johnson et al. smoking (Table 6.22 S ). As part of the Norwegian-Swedish - 2000; Gammon et al. 2004a; Shrubsole et al. 2004; Bon cohort, Gram and colleagues (2005) followed 102,098 ner et al. 2005). The review did not find an association for women, 30–50 years of age, for an average of 8−9 years postmenopausal women (OR = 1.00; 95% CI, 0.88–1.12) (1991/1992–2000) and ascertained 1,240 incident cases based on the same 2 cohort studies (Reynolds et al. 2004b; of breast cancer among current or former smokers and Hanaoka et al. 2005) and 7 of the 9 case-control studies never smokers. Exposure to passive smoking at home was (Sandler et al. 1985a; Millikan et al. 1998; Delfino et al. assessed from self-reports of living with a smoker, either 2000; Johnson et al. 2000; Gammon et al. 2004a; Shrub- currently or during childhood. In a multivariate model sole et al. 2004; Bonner et al. 2005). The review identi - based on 1,130 cases with complete data, the RR for breast fied several issues related to these results—including the cancer among women who never smoked but reported significant heterogeneity among studies, especially for the exposure to passive smoking (n = 24,030) was 1.21 (95% - case-control studies; the potential for selection and infor CI, 0.98–1.50) in a comparison with never smokers who mation biases; the lack of consistency between findings reported no exposure to passive smoking (n = 12,743). The - for active cigarette smoking and those for exposure to sec study adjusted for multiple covariates—including age, ondhand smoke; and biologic plausibility. menopausal status, parity, age at birth of first child, use of In summary, several reviews and meta-analyses have hormones, BMI, and use of alcohol. been conducted to date—including reports by IARC, the In the Million Women Study, Pirie and colleagues Cal/EPA, the Canadian Expert Panel, Surgeon General’s (2008) ascertained 2,344 incident cases in a cohort of reports, and several groups of investigators (Khuder and 210,647 women, 50–64 years of age, who never smoked, Simon 2000; Johnson 2005; Lee and Hamling 2006; Pirie had complete data for passive smoking exposure, and were et al. 2008). These reports have reached different conclu - followed for an average of 3.5 years. Exposure to passive sions about the presence and magnitude of association smoking was based on self-reports of living with a par - between passive exposure to smoke and breast cancer ent who smoked at the time the participant was born and despite considerable overlap in the studies reviewed and when she was 10 years of age, and of currently living with analyzed. Some of the difference in interpretation is a partner who smoked. Only 17% of women reported not related to the relative weight given by the authors of the being exposed to passive smoking during childhood or reviews and meta-analyses to results from case-control 258 Chapter 6

288 The Health Consequences of Smoking —50 Years of Progress adulthood, leaving a relatively small reference group with (RR = 1.17, 95% CI, 0.96–1.41) in a fully adjusted analysis. There were trends toward increasing risk with duration no active/no passive exposure for the analyses. The overall RR was 0.99 (95% CI, 0.93–1.05) for any passive exposure. and intensity of exposure that reached statistical signifi- cance only in the highest category of this combined vari- After adjusting for relevant covariates, including use of 4 in postmenopausal women (RR able (>42 intensity-years) alcohol, the study found no increased risk of breast can - In this study, the unexposed = 1.25; 95% CI, 1.01–1.56). cer from exposure during childhood (RR = 0.96; 95% CI, 0.88–1.05) or adulthood (RR = 1.02; 95% CI, 0.89–1.16). reference group constituted only 14% of the women in the cohort. The measure of exposure intensity was highly Lin and colleagues (2008) reported findings from the Japan Collaborative Cohort Study for Evalustion of qualitative (self-report of “a little smoky,” “fairly smoky,” Cancer Risk based on 208 incident breast cancer cases and “very smoky”). Xue and colleagues (2011) reported updated analy - in 34,401 women, 40–79 years of age, who were followed ses for the NHS-I on active and passive smoking and risk an average of 11–13 years. The study assessed exposure of breast cancer. Their data included 2,890 incident breast to passive smoking based on self-reports—including the estimated frequency of exposure (either sometimes cancer cases among 36,017 nonsmoking women followed or almost every day)—as adults at home and in pub- from 1982–2006. No significant associations were found for any of the following categories of passive exposure: lic places, and during childhood. There were 196 cases both parents (RR = 0.90; 95% CI, 0.79–1.03), regular at among 32,023 never-smoking women, but the numbers in various analyses ranged from 140–178. After adjust- work (RR = 0.87; 95% CI, 0.78–0.98), regular at home (RR ing for relevant covariates, including use of alcohol, RRs = 1.02; 95% CI, 0.90–1.14), and living with a smoker for 40 or more years (RR = 0.99; 95% CI, 0.74–1.32). Indices for exposure during adulthood at home and in public that combined information on place (home or work) and places almost every day were less than 1.0 (RR = 0.71; duration (<20 vs. >20 years) of exposure were not signifi 95% CI, 0.48–1.05 and RR = 0.84; 95% CI, 0.51–1.40, - respectively). The RR for exposure during childhood was cantly associated with risk. All estimates were adjusted for - slightly higher (RR = 1.24; 95% CI, 0.84–1.85) but still age and multiple relevant covariates but were not strati not significant. fied by menopausal status. Reynolds and colleagues (2009) reported on passive Also as shown in Table 6.22 S , Luo and colleagues smoking and risk of breast cancer using data from the (2011b) reported results for passive smoking and incident breast cancer from the Women’s Health Initiative. There WAVE-II survey (1997) of the California Teachers Study. were a total of 1,692 incident cases among 41,022 post - This analysis was based on 1,754 women with incident invasive breast cancer among a cohort of 57,523 women menopausal women, who had never smoked, followed over who were lifetime nonsmokers and followed over 10 years. an average of 10.3 years. There were no significant associ - - This report updates one published in 2004 that was based ations between passive exposure during childhood, adult hood at home or at work, or any combination thereof, and on data from the WAVE-I survey (1995) for 1,174 cases risk of breast cancer. The only significant association was among 77,708 lifetime nonsmokers followed over 4 years (Reynolds et al. 2004b). The WAVE-II survey included for the highest combined category of exposure duration (childhood ≥10 years plus adult at home ≥20 years plus more extensive questions on frequency, duration, source, and intensity, and there was a large loss to follow-up from adult at work ≥10 years: RR = 1.32; 95% CI, 1.04–1.67), but the trend across the duration categories for increased WAVE-I to WAVE-II. The RR for breast cancer with ever- lifetime exposure in the WAVE-II survey was 1.10 (95% risk with greater exposure was not significant (p = 0.10). This is one of the only studies to examine exposure to pas- CI, 0.94–1.30), adjusting for age, race, and other relevant sive smoking in relation to breast cancer by ER/PR status, covariates (Reynolds et al. 2009). The RRs were 1.06 (95% but no significant associations were found. All estimates CI, 0.94–1.19) and 1.04 (95% CI, 0.91–1.19) for any child - were adjusted for age at enrollment and multiple relevant hood (<20 years of age) and any adulthood (≥20 years of age) exposures, respectively; and 1.04 (95% CI, 0.92–1.16) covariates. Finally, Chuang and colleagues (2011) reported the and 1.02 (95% CI, 0.93–1.13) for any home and any work exposures, respectively. Exposure before first pregnancy RR for childhood exposure from parental smoking (RR = 0.98; 95% CI, 0.91–1.06) based on data from 6 of the 23 was also associated with a nonsignificant increased risk 4 To predict risk of breast cancer for two age groups (<20 years of age and ≥20 years of age), Reynolds and colleagues (2009) combined two metrics (years of exposure and intensity) into a common metric (intensity-years) that included both intensity (smokiness) and duration (years) of exposure. 259 Cancer

289 Surgeon General’s Report centers participating in the EPIC; these centers were in Third, these cohort studies differ markedly in rates France, Italy, The Netherlands, Sweden, Denmark, and of breast cancer incidence and exposure to passive smok - Norway. There were 3,187 breast cancer cases among ing. In the Japanese cohort study (Lin et al. 2008), which included both in situ and invasive cases, participants 92,956 premenopausal and postmenopausal women, 25–70 years of age, who reported themselves to be never had a very low incidence of breast cancer (approximately 58/100,000) compared with the other cohorts (Norwegian- smokers at recruitment (1992–1998); the mean age at Swedish, approximately 114/100,000; Million Women, recruitment was 50 years. Follow-up was over an average approximately 315/100,000; and Women’s Health Initia- of 9–10 years. Significant associations were not found for the two frequency categories of exposure in childhood: few tive, approximately 428/100,000). While the difference across these studies for incidence of breast cancer partly times during a week (RR = 0.98; 95% CI, 0.88–1.10) and daily (RR = 1.06; 95% CI, 0.95–1.19). All estimates were reflects the age composition of the respective cohorts, geographic and ethnic/racial differences must be consid- adjusted for age at menarche, ever use of oral contracep - tives, parity, menopausal status, education, alcohol use, ered also. - BMI, physical activity, vegetable intake, fruit intake, non Fourth, methods for exposure assessment varied alcoholic energy intake, and adulthood passive smoking. from study to study. For example, the reported prevalence Several issues should be considered when compar - of lifetime (childhood and adulthood) exposure to second - ing and combining the results of these seven studies. hand smoke varied markedly, from approximately 24% in the Norwegian-Swedish cohort to greater than 90% in First, the categories of exposure were generally broad, particularly in the Norwegian-Swedish cohort (Gram the Women’s Health Initiative cohort study. As noted in et al. 2005). Second, with the exception of the studies by the 2006 Surgeon General’s report, these cohort studies Pirie and colleagues (2008) and Reynolds and colleagues lacked updated data about exposure to passive smoking, - (2009), analyses were not stratified by menopausal status, which can result in some misclassification, especially dur ing long-term followup periods of marked secular change use of alcohol, or breast cancer phenotype, although most in smoking habits. Xue and colleagues (2011) acknowl - studies adjusted for these potential confounders. The edged this limitation in the NHS and pointed out that the Norwegian-Swedish Cohort (Gram et al. 2005) consisted result would be to attenuate estimates toward the null mostly of premenopausal women at baseline and the Women’s Health Initiative cohort (Luo et al. 2011b) was value because any exposure misclassification may be safely assumed to be nondifferential in a cohort study design. comprised entirely of postmenopausal women; whereas the Million Women Study (Pirie et al. 2008), Japan Col - The most recent reports (Reynolds et al. 2009; Luo et al. 2011b; Xue et al. 2011) used novel indices of exposure laborative Cohort Study for Evaluation of Cancer Risk that combined available information for duration, place, (Lin et al. 2008), California Teachers Study (Reynolds et timing, and intensity. The analyses of Reynolds and col - al. 2009), EPIC (Chuang et al. 2011), and NHS-I (Xue et - leagues (2009) and Luo and colleagues (2011b) suggest al. 2011) cohorts included both premenopausal and post increased risk at only the very highest levels of these indi- menopausal women. This is important because a previous ces, while the results of Xue and colleagues are essentially cohort study by Hanaoka and colleagues (2005) (Table null. The analysis of Pirie and colleagues (2008) is unique - ) reported markedly different risks for premeno S 6.14 in restricting the data to women who reported living pausal (RR = 2.6; 95% CI, 1.3–5.2) and postmenopausal with a partner. This could be important because women women (RR = 0.7; 95% CI, 0.4–1.0). This difference in risk by menopausal status was also found in the meta-analysis who live alone cannot be passively exposed routinely in the home, a major venue of adult passive exposure. Theo - of cohort and case-control studies included in the 2006 retically, the restriction imposed by Pirie and colleagues Surgeon General’s report (USDHHS 2006). Pirie and col- (2008) could produce bias because women not living with leagues (2008) stratified estimates by menopausal status a partner are likely to differ with respect to multiple risk but included few premenopausal women (n = 60), and factors for breast cancer, especially those related to repro - thus the resulting estimate, although significant, was both ductive history. inverse and imprecise (RR = 0.54; 95% CI, 0.30–0.99). In contrast, the analysis by Reynolds and colleagues (2009) suggests that risk may be increased in postmenopausal Case-Control Studies rather than premenopausal women. Xue and colleagues The 2006 Surgeon General’s report evaluated 14 (2011), who also stratified by menopausal status, did not case-control studies on the association between passive provide results that could be used for comparison. Thus, - smoking and risk for breast cancer. Since then, 10 differ considerable inconsistency remains with regard to the ent case-control studies have been conducted, resulting in effects of passive smoking exposure by menopausal status. 11 published reports (Table 6.23 S ). Two reports (Metsola 260 Chapter 6

290 —50 Years of Progress The Health Consequences of Smoking et al. 2005; Sillanpaa et al. 2005a) were based on the same 50 years of age. Analyses were stratified by menopausal study population; the latter report included adjustment status and Hispanic/non-Hispanic White ethnicity. ORs for potential confounders. were adjusted for age, study site, BMI, use of aspirin or NSAIDs, parity, use of alcohol, physical activity, and recent use of estrogen. The study found a significant increased North American Studies risk only in premenopausal Hispanic women report- Three large case-control studies were conducted in ing more than 10 hours of exposure to passive smoking North America (Mechanic et al. 2006; Slattery et al. 2008; per week during the reference period compared with a Young et al. 2009). In a combined sample of the Ontario no active/no passive reference group (OR = 2.3; 95% CI, Women’s Health Study and the Ontario Women’s Diet and 1.2–4.5). However, there was an inverse association, albeit Health Study (2,751 nonsmoking cases and 3,097 non- nonsignificant, between fewer hours of exposure to pas - smoking controls), Young and colleagues (2009) reported sive smoking in this subgroup and risk. In this same sub- results on the association between exposure to passive group, a significant interaction with a SNP in the gene IL6 smoking and risk for breast cancer. Exposure to passive also was detected (see “Secondhand Smoke Exposure and smoking was self-reported and defined as exposure less Genotype”). The estimates for postmenopausal women than 2 hours per day during childhood and exposure of were essentially null, and those for non-Hispanic White at least 2 hours per day for workplace and nonworkplace premenopausal women were increased by about 20%. The environments (adult exposure) during the 2 years before overall lifetime summary estimate (OR) calculated for this the study interview. The study reported an overall OR of report was 1.06 (95% CI, 0.88–1.28). 0.97 (95% CI, 0.88–1.08) for exposure to passive smoking Taken together, these large case-control studies do compared with a no active/no passive exposure reference not provide evidence that exposure to secondhand smoke group. This estimate was adjusted only for age because is a risk factor for breast cancer. However, the assessment the change to the risk estimate was less than 10% when of exposure to passive smoking was relatively crude in the other potential confounders were included. Strati - two studies that did not stratify results for potential effect fied analyses by timing of exposure (childhood vs. adult - modifiers—timing of exposure or menopausal status. hood), menopausal status, or other relevant variables were Three additional case-control studies conducted in North not provided. America collected more extensive exposure data, but the In the Carolina Breast Cancer Study, which included results are difficult to interpret because of small samples both African American and White women, Mechanic and (Alberg et al. 2004; Rollison et al. 2008; Ahern et al. 2009). colleagues (2006) evaluated the association between In a case-control study in Massachusetts (242 nonsmoking exposure to passive smoking and risk for breast cancer cases, 195 nonsmoking controls), Ahern and colleagues among 1,211 nonsmoking cases and 1,087 nonsmoking (2009) collected information about exposure to passive controls. Passive smoking was broadly defined as living smoking according to stage of life (childhood, adulthood), with a smoker after 18 years of age. After adjusting for parental source during childhood (father, mother), and age, age at menarche, age at first full-term pregnancy, par - location (home, workplace). Overall, the results were null; ity, family history, and use of alcohol, the study found an only two significantly increased risks were reported: one increased risk for breast cancer among African American for exposure during childhood from a mother who smoked women (OR = 1.40; 95% CI, 1.00–1.90) but not among (OR = 1.9; 95% CI, 1.1–3.3), and the other for postmeno - White women (OR = 1.00; 95% CI, 0.80–1.20) compared pausal women exposed during childhood (OR = 1.8; 95% with a no active/no passive exposure reference group. CI, 1.0–3.3). In a small case-control study in Delaware Results were not stratified by menopausal status. For (124 nonsmoking cases, 116 nonsmoking controls), Rol- African Americans, risk for breast cancer associated with lison and colleagues (2008) collected extensive data on exposure to passive smoking appeared to increase with the exposure to passive smoking at home during childhood number of at-risk genotypes, which consisted of SNPs in and adulthood and at the workplace in adulthood. Data DNA repair genes. included estimates of the number of smokers in the house- In the 4-Corners Breast Cancer Study, Slattery and hold, number of hours of exposure per day, and intensity of colleagues (2008) examined the association between expo- exposure (packs of cigarettes smoked per day). Compared sure to passive smoking and risk for breast cancer among with a no active/no passive exposure reference group, the 1,347 nonsmoking cases and 1,442 nonsmoking controls. study did not find any significant increased ORs across Data on exposure to passive smoking was self-reported and any exposure category, but statistical power was limited captured as the number of exposure hours per week, both by the small sample. In another small case-control study in and out of the house, during a reference period of 1 year (115 cases and 115 controls matched for age, race, and before cancer diagnosis or study interview and 15, 30, and 261 Cancer

291 Surgeon General’s Report 730 nonsmoking controls). For this study, Kruk defined menopausal status), Alberg and colleagues (2004) assessed - the association between passive smoking, defined as liv exposure to passive smoking as living with a spouse who ing with a spouse who smoked, and risk for breast cancer. smoked and defined dose as number of cigarettes smoked Data were available for only 62 nonsmoking cases and 66 per day. In contrast to Lissowska and colleagues (2007), Kruk (2007) reported significant ORs for premenopausal nonsmoking controls. The OR for breast cancer was 1.2 women (2.86; 95% CI, 1.65–4.97) and postmenopausal (95% CI, 0.59–2.4). The study observed a nonsignificant women (2.57; 95% CI, 1.73–3.80). These estimates, how interaction between exposure to passive smoking and the - NAT2 phenotype. ever, were adjusted only for age among premenopausal - women and age and breastfeeding among postmeno - European Studies pausal women, and smokers were mixed with nonsmok ers in the reference group. Among case-control studies, Five reports based on four case-control studies this study provides some of the highest ORs for active and in Europe have been published since the 2006 Surgeon passive smoking. General’s report. Two of these studies were conducted in Roddam and colleagues (2007) conducted a study in Poland (Lissowska et al. 2006; Kruk 2007), one in Finland England of women, 36–45 years of age, who were mostly (Metsola et al. 2005; Sillanpaa et al. 2005a), and one in premenopausal. Exposure to passive smoking at home was England (Roddam et al. 2007). defined as living at least 1 year with a partner who smoked, - The largest European study was conducted by Lis and dose was defined as the number of years of exposure sowska and colleagues (2006) and had 1,034 nonsmoking and estimated number of cigarettes smoked per day. After cases and 1,162 nonsmoking controls. Passive smoking adjusting for relevant covariates, exposure to secondhand was self-reported and defined as adult exposure at home smoke was not significantly associated with risk for breast or in the workplace for at least 1 hour per day for at least cancer (OR = 0.89; 95% CI, 0.64–1.25) among 297 non - - 1 year. In a comparison with a no active/no passive expo smoking cases and 310 nonsmoking controls when no sure reference group, this study did not find significant passive/no active exposure was the reference group. Esti- associations between risk for breast cancer and exposure - mates were stratified on menopausal status, but the num both to passive smoking at home, at the workplace, at ber of perimenopausal/postmenopausal women (n = 23) - home and the workplace, or for either the home or work was too small to provide a meaningful result. place. After adjusting for relevant covariates, the OR was Metsola and colleagues (2005) and Sillanpaa and 1.10 (95% CI, 0.84–1.45) for either the home or work - colleagues (2005a) published results on the same case- place. The initial analyses did not stratify risk by stage - control study in Finland. Both focused on the modifica of life (childhood, adulthood), age group, or menopausal tion of active smoking by selected SNPs in DNA repair - status. A subsequent reanalysis, however, which was pub genes, but both reports provided only a cur and NAT2 - lished as a response to a letter to the editor by Johnson sory description of how exposure to passive smoking was (2007), reported results that were stratified by age group defined in terms of years at home and the workplace. The and menopausal status (Lissowska et al. 2007). Premeno - two reports provided ORs for the association between ) exhibited increasing ORs for pausal women (Table 6.23 S exposure to passive smoking and risk for breast cancer breast cancer by hours of exposure to secondhand smoke (153 nonsmoking cases, 169 nonsmoking controls), but 5 : less than 100, 1.36 (95% CI, 0.67–2.73); per day-years only the estimate from Sillanpaa and colleagues (2005a) 101–200, 1.52 (95% CI, 0.73–3.13); and more than 200, was adjusted for multiple covariates; this estimate was not 2.02 (95% CI, 0.94–4.36) (p trend = 0.08). The indicator of significant (OR = 0.85; 95% CI, 0.62–1.16). Stratification hours per day-years was calculated as the product of hours - on the NAT2 phenotype suggested that risk for breast can of exposure per day and duration of exposure. Of note, the cer was increased in women with the slow phenotype who study did not find similar trends for either of the two age were passively exposed to tobacco smoke (OR = 1.22; 95% groups (younger than 45 years of age and 45–55 years of CI, 0.75–1.98). age) that included all premenopausal women. Kruk (2007) reported results from an independent case-control study in Poland (445 nonsmoking cases, 5 Day-years: the sum of hours per day exposed to secondhand smoke multiplied by the number of years of all episodes of secondhand smoke exposure, whether at home, at work, or during leisure time. Chapter 6 262

292 The Health Consequences of Smoking —50 Years of Progress Asian Studies S provides a listing of the 39 reports for 34 Table 6.24 studies, of which 9 overlap with results on the same study Findings from case-control studies carried out in - population. Of these, 7 are included in the meta-analy Asia on secondhand smoke have not been published since ses because they are either the most recent or complete 2005. However, the 2006 Surgeon General’s report did not reports from their study. In the case of 1 cohort study include the hospital-based, cross-sectional study by Hirose (California Teachers Study) and 1 case-control study (Car - and colleagues (1995) that was conducted in Japan. Using olina Breast Cancer Study), the best exposure estimates a large administrative database that had data for cigarette for specific categories were selected for inclusion in the smoking and exposure to secondhand smoke, the study meta-analyses: California Teachers Study (Reynolds et al. identified 1,052 breast cancer cases with survey data and 2004b, 2009) and Carolina Breast Cancer Study (Millikan 23,163 controls without a cancer diagnosis. The analysis et al. 1998; Marcus et al. 2000; Mechanic et al. 2006). A for passive smoking was limited to women who reported total of 34 separate reports were included in the broadest being nonsmokers (560 cases and 11,276 controls). The category of exposure for the meta-analyses: Most compre- prevalence of smoking in the control group (14%) was . RR and OR estimates were pooled across expo - hensive similar to that in the general population of women in sure levels to fit into one of the meta-analysis categories Japan (13%). Passive smoking among women who were when necessary. nonsmokers was defined on the basis of whether the hus- band smoked and the number of cigarettes he smoked per Measures of Exposure to Secondhand Smoke day (either 0–19 or ≥20). Among premenopausal women, risk for breast cancer increased as the number of cigarettes - This meta-analysis used eight categories of mea smoked per day by the husband rose: 0–19 (RR = 0.81; sures of exposure to secondhand smoke. These categories 95% CI, 0.57–1.15) and 20 or more (RR = 1.30; 95% CI, are not mutually exclusive, and assignments are presented 1.02–1.65). There was no similar dose-response relation - S in Table 6.24 . ship in postmenopausal women: 0–19 (RR = 1.55; 95% CI, 1.10–2.17) and 20 or more (RR = 1.28; 95% CI, 0.92–1.77). 1. - This category was based on expo Spouse/partner: The study had several limitations: it was clinic based and sure during adulthood from a spouse or partner who may have included prevalent as well as incident cases, data was a smoker. were missing on passive smoking for 38% of nonsmoking women, and risk estimates were adjusted only for age and 2. This category was based on exposure Adult—home: year of first visit to a clinic. during adulthood from any smoker in the home. The category Adult— is a subset of Spouse/partner home because the location of exposure was assumed to be in the home. Meta-Analysis of Breast Cancer Risk Associated with Measures of 3. Adult—workplace: Based on exposure during adult- Secondhand Smoke hood from smokers at the workplace, an estimate - from this category could be used for any adult. How A total of 19 new published reports (7 cohort, 12 ever, most studies with a measure for exposure at case-control) were reviewed together with the 20 reports the workplace had a measure for exposure at home (5 cohort, 15 case-control) that were previously abstracted that took precedence. and analyzed for the 2006 Surgeon General’s report. Three of these update previous reports from the same studies Childhood: This category was based on exposure 4. and one overlaps with a current report (Table 6.24 ). RR S during childhood to any smoker in the home. and OR estimates were based on either single estimates or Among the 15 studies that provided a childhood were pooled across exposure strata and classified similarly estimate, the age definition of childhood varied. - to the eight categories reported in the 2006 Surgeon Gen Sixteen, 18, or 21 years of age defined the end of eral’s report. The same statistical procedures used in the (Smith et al. 1994; childhood exposure in 7 studies meta-analyses for active cigarette smoking were used for Marcus et al. 2000; Gammon et al. 2004a; Bonner the analyses of exposure to secondhand smoke. Sensitivity et al. 2005; Rollison et al. 2008; Chuang et al. 2011; analyses considered study design, sample size, and magni- he remaining studies did not Luo et al. 2011b), and t tude of exposure effect. (Johnson et al. 2000; define a specific cutoff for age 263 Cancer

293 Surgeon General’s Report home (Lash and Aschengrau 1999, 2002; Zhao et al. Liu et al. 2000; Kropp and Chang-Claude 2002; Lin et al. 2008; Pirie et al. 2008; Ahern et al. 2009; Reyn- 1999; Bonner et al. 2005; Slattery et al. 2008) ; 4 . - were based on any exposure from a spouse or a par olds et al. 2009; Xue et al. 2011) (Gammon et al. 2004a; Gram ent during the lifetime 5. Adulthood and childhood (or lifelong): This cat- ; 1 et al. 2005; Pirie et al. 2008; Chuang et al. 2011) egory was based on lifelong exposure during child - was based on having lived with a smoker or been exposed to a smoker outside of the home hood and adulthood from any individual in any (Hanaoka ; 5 were based on having lived with a et al. 2005) setting. Only seven studies defined exposure in this smoker or been exposed at the workplace (Smith manner (Smith et al. 1994; Johnson et al. 2000; et al. 1994; Morabia et al. 1996; Johnson et al. 2000; Kropp and Chang-Claude 2002; Reynolds et al. ; 2004b; Pirie et al. 2008; Ahern et al. 2009; Luo et al. Kropp and Chang-Claude 2002; Rollison et al. 2008) and 4 were based on any exposure during childhood . 2011b) or adulthood without information about location or source of exposure This category was based on the Adult—any source: 6. (Reynolds et al. 2009; Young . The broadest measure et al. 2009; Xue et al. 2011) broadest, most inclusive measure available for expo - Ever in lifetime for was selected in those studies that sure during adulthood from any source in the fol - reported more than one category of exposure during lowing priority: a general estimate for all sources of childhood and adulthood. The home was the most exposure if available, a comprehensive home expo- frequently defined location for exposure; outside the sure, spouse/partner exposure, and workplace expo - sure. Twenty-six non-overlapping reports included home and/or at the workplace were identified less frequently. Studies varied widely in specificity and measures that were coded for this category based rigor of the definition of lifetime exposure. on a number of descriptive measures, including a general report for overall and nonspecific exposure This category was based on 8. Most comprehensive: to passive smoke as an adult (Johnson et al. 2000; the broadest, most inclusive estimate of exposure Kropp and Chang-Claude 2002; Ahern et al. 2009); available from each study. In the meta-analysis, this - exposure specifically noted as from a spouse or part Adult—any source or Ever in was always either life- ner (Sandler et al. 1985a; Hirose et al. 1995; Mora- - , with preference for the latter when both esti time bia et al. 1996; Jee et al. 1999; Nishino et al. 2001; mates were reported. A careful evaluation was made Alberg et al. 2004; Gammon et al. 2004a; Kruk 2007; of the independent contributions of each category to Roddam et al. 2007; Pirie et al. 2008); cohabitants the summary estimate for the Most comprehensive (Smith et al. 1994; Delfino et al. 2000; in general Comparison of Adult—Any Source with Ever in (see Liu et al. 2000; Mechanic et al. 2006; Lin et al. 2008; Lifetime for Most Comprehensive ). Reynolds et al. 2009; Xue et al. 2011); cowork- ers (Bonner et al. 2005; Hanaoka et al. 2005); or a This meta-analysis applied some changes to the combination of cohabitants and coworkers (Shrub- studies reviewed in the 2006 Surgeon General’s report, sole et al. 2004; Sillanpaa et al. 2005a; Lissowska (Hirayama including the exclusion of two mortality studies . et al. 2006; Luo et al. 2011b) , the inclusion of a study 1984; Wartenberg et al. 2000) (Hirose Based on a report of exposure to conducted in China and published prior to 2005 7. Ever in lifetime: passive smoke during either childhood or adulthood - , and changes to several estimates for five stud et al. 1995) in studies that assessed exposure across the lifetime, (Smith et al. 1994; Millikan et al. 1998; Jee et al. 1999; ies hese changes this category can include, for example, an estimate Nishino et al. 2001; Gammon et al. 2004a). T S are detailed in the notes for Table 6.24 based on exposure during adulthood if exposure dur - . Risk estimates were abstracted for each study, classified into the eight ing childhood was also assessed and included in the categories described previously, and tabulated together Adulthood and childhood risk estimate. The category Twenty nonoverlap- Ever in lifetime. with information on adjusted covariates, including repro - is a subset of ping reports had measures that were coded for this ductive risk factors, alcohol use, BMI, family history, and category based on definitions that ranged from very menopausal status. The most fully adjusted estimates general to specific. One study estimate was based on were selected when available, and a random effects model was used to pool estimates across strata (e.g., race/ethnic - (Ahern exposure during childhood and adulthood ity, menopausal status, or dose levels) when necessary. et al. 2009) ; 5 were based on lifetime exposure in the 264 Chapter 6

294 —50 Years of Progress The Health Consequences of Smoking Adjustment for Selected Covariates studies by Kruk (2007) and Zhao and colleagues (1999) appeared to include smokers along with nonsmokers in The majority of studies that evaluated exposure to the analysis of exposure to passive smoke. Furthermore, passive smoke adjusted for covariates, most often referenc - the number of cases and controls reported in the tables in ing those that were related to reproduction or estrogen, both of these studies could not be reconciled with totals but also family history, use of alcohol, and BMI. Of the 34 provided in the text or in other tables. Excluding these two separate studies, only 4 did not adjust for any covariate or studies (Table 6.25 ) attenuated the overall risk estimate S (Sandler et al. 1985a; Jee et al. 1999; adjusted for age only = 0.001; n = 32) and the (RR = 1.08; 95% CI, 1.01–1.14; p h Alberg et al. 2004; Metsola et al. 2005) . risk estimate for the case-control studies (RR = 1.14; 95% = 0.003; n = 22). CI, 1.04–1.26; p h Most Comprehensive Measures of The extreme estimate from Smith and colleagues Passive Smoking (1994) was based on a very small subset of cases and con - Among the 34 studies included in the meta-analysis - trols (n = 193) that represented only 27% of the nonsmok of passive smoking and risk for breast cancer, only 7 did not ers (n = 703) in the full study. Other studies were also report estimates for measures of active smoking (Jee et al. based on a small number of cases. For example, estimates 1999; Liu et al. 2000; Nishino et al. 2001; Bonner et al. reported by Morabia and colleagues (1996) were based on 2005; Pirie et al. 2008; Reynolds et al. 2009; Chuang et al. only 126 cases (620 controls), and the results from the 2011). Eight of the 34 studies were based on Asian popula - cohort study by Lin and colleagues (2008) were based on tions (Hirose et al. 1995; Jee et al. 1999; Zhao et al. 1999; only 140 incident cases. However, although small studies Liu et al. 2000; Nishino et al. 2001; Shrubsole et al. 2004; are statistically more likely to produce extreme estimates, Hanaoka et al. 2005; Lin et al. 2008), and 8 studies included these studies adjusted for appropriate covariates and did data on the interaction between genotype and smoking not have other limitations to their respective designs. for risk for breast cancer (Delfino et al. 2000; Alberg et al. Limitations in study design were detected in three 2004; Gammon et al. 2004a; Metsola et al. 2005; Sillanpaa other studies that did not provide extreme estimates. Two et al. 2005a; Lissowska et al. 2006; Mechanic et al. 2006; studies included an unknown percentage of deceased Slattery et al. 2008). Figure 6.40 presents the 34 studies (10 persons for whom information was collected from prox- cohort and 24 case-control) that were based on the Most ies and did not adjust for menopausal status (Lash and comprehensive category, which was derived from either Aschengrau 1999, 2002), and one study included both (n = 20) Ever in lifetime (n = 14) or Adult—any source the incident and prevalent cases based on medical records measures. Meta-analysis provided an overall summary RR and did not adjust for covariates relevant to breast cancer - of 1.14 (95% CI, 1.06–1.23), but with significant heteroge (including menopausal status) other than age (Jee et al.