Tobacco Documents Online

This Report should galvanize the health community to stress repeatedly at every opportunity the value of smoking cessation to the 50 million Americans who continue to smoke. James O. Mason, M.D., Dr.P.H. William L. Roper, M.D. Assistant Secretary for Health Director Public Health Service Centers for Disease Control ui TIMN 438387

PREFACE This Report of the Surgeon General is the 21st Report of the U.S. Public Health Service on the health consequences of smoking and the first issued during my tenure as Surgeon General. Whereas previous reports have focused on the health effects of smoking, this Report is devoted to the benefits of smoking cessation. The public health impact of smoking is enormous. As documented in the 1989 Surgeon General's Report, an estimated 390,000 Americans die each year from diseases caused by smoking. This toll includes 115,000 deaths from heart disease; 106,000 from lung cancer; 31,600 from other cancers; 57,000 from chronic obstructive pulmonary disease; 27,500 from stroke; and 52,900 from other conditions related to smoking. More than one of every six deaths in the United States are caused by smoking. For more than a decade the Public Health Service has identified cigarette smoking as the most important preventable cause of death in our society. It is clear, then, that the elimination of smoking would yield substantial benefits for public health. What are the benefits, however, for the individual smoker who quits? A large body of evidence has accumulated to address that question and derives from cohort and case-control studies, cross-sectional surveys, and clinical trials. In studies of the health effects of smoking cessation, persons classified as former smokers may include some current smokers; this misclassification is likely to cause an underestimation of the health benefits of quitting. Taken together, the evidence clearly indicates that smoking cessation has major and immediate health benefits for men and women of all ages. Overall Benefits of Smoking Cessation People who quit smoking live longer than those who continue to smoke. To what extent is a smoker's risk of premature death reduced after quitting smoking? The answer depends on several factors, including the number of years of smoking, the number of cigarettes smoked per day, and the presence or absence of disease at the time of quitting. Data from the American Cancer Society's Cancer Prevention Study II (CPS-II) were analyzed in this Report to estimate the risk of premature death in ex-smokers versus current smokers. These data show, for example, that persons who quit smoking before age 50 have one-half the risk of dying in the next 15 years compared with continuing smokers. Smoking cessation increases life expectancy because it reduces the risk of dying from specific smoking-related diseases. One such disease is lung cancer, the most common cause of cancer death in both men and women. The risk of dying from lung cancer is TIMN 438388 v

FOREWORD More than 38 million Americans have quit smoking cigarettes, and nearly half of all living adults who ever smoked have quit. Unfortunately, some 50 million Americans continue to smoke cigarettes, despite the many health education programs and anti- smoking campaigns that have been conducted during the past quarter century, despite the declining social acceptability of smoking, and despite the consequences of smoking to their health. Twenty previous reports of the Surgeon General have reviewed the health effects of smoking. Scientific data are now available on the consequences of smoking cessation for most smoking-related diseases. Previous reports have considered some of these data, but this Report is the first to provide a comprehensive and unified review of this topic. The major conclusions of this volume are: 1. Smoking cessation has major and immediate health benefits for men and women of all ages. Benefits apply to persons with and without smoking-related disease. 2. Former smokers live longer than continuing smokers. For example, persons who quit smoking before age 50 have one-half the risk of dying in the next 15 years compared with continuing smokers. 3. Smoking cessation decreases the risk of lung cancer, other cancers, heart attack, stroke, and chronic lung disease. 4. Women who stop smoking before pregnancy or during the first 3 to 4 months of pregnancy reduce their risk of having a low birthweight baby to that of women who never smoked. 5. The health benefits of smoking cessation far exceed any risks from the average 5-pound (2.3-kg) weight gain or any adverse psychological effects that may follow quitting. With the long-standing evidence that smoking is extremely harmful to health and the mounting evidence that smoking cessation confers majorr health benefits, we remain faced with the task of developing effective strategies to curtail the use of tobacco. Two broad categories of intervention are available: prevention of smoking initiation among youth and smoking cessation. Resources for tobacco control are limited, and policymakers must decide how best to allocate those resources to smoking prevention and cessation. The goal of public health is to intervene as early as possible to prevent disease, disability, and premature death. From that standpoint, prevention of smoking initiation i TIMN 438385

Public opinion polls tell us that most smokers want to quit. This Report provides smokers with new and more powerful motivation to give up this self-destructive behavior. Antonia C. Novello, M.D., M.P.H. Surgeon General xii TIMN 438395

should be a major priority. More than 3,000 teenagers become regular smokers each day in the United States. Because of the strength of nicotine addiction, some have argued that public health efforts should focus on smoking prevention rather than smoking cessation. However, this need not be an "either-or" situation. Public health practitioners have categorized interventions into primary, secondary, and tertiary prevention. Primary prevention generally refers to the elimination of risk factors for disease in asymptomatic persons. Secondary prevention is defined as the early detection and treatment of disease, and is practiced using tools such as Pap smears and blood pressure screening. Tertiary prevention consists of measures to reduce impairment, disability, and suffering in people with existing disease. Smoking cessation falls under the category of primary prevention as does the prevention of smoking initiation. Smoking cessation meets the definition of primary prevention by reducing the risk of morbidity and premature mortality in asymptomatic people. In addition, parents who quit smoking reduce or eliminate the risk of passive- smoking-related disease among their children and reduce the probability that their children will become smokers. Thus, there should be no debate about the need for smoking prevention versus cessation-both are important. Public awareness of the health effects of smoking has increased substantial ly through the years. Nevertheless, important gaps in public knowledge still exist. Some smokers may have failed to quit because of a lack of appreciation of the health hazards of smoking and the benefits of quitting. In the 1987 National Health Interview Survey of Cancer Epidemiology and Control, respondents were asked whether smoking increases the risk of various diseases (lung cancer, cancer of the mouth and throat, heart disease, emphysema, and chronic bronchitis) and whether smoking cessation reduces the risk. Thirty to forty percent of smokers either did not believe that smoking increases these risks or did not believe that cessation reduces these risks. These proportions correspond to 15 to 20 million smokers in the United States. Clearly, our efforts to educate the public on the health hazards of smoking and the benefits of quitting are not yet complete. As we continue and intensify our efforts to inform the public of these findings, we must make available smoking cessation programs and services to those who need them. Although 90 percent of former smokers quit without using smoking cessation programs, counseling, or nicotine gum, smokers who do need this assistance should have it available. We endorse the view expressed in the Preface to the 1988 Surgeon General's Report that treatment of nicotine addiction should be considered at least as favorably by third-party payors as treatment of alcoholism and illicit drug addiction. Good smoking cessation treatments can achieve abstinence rates of 20 to 40 percent at 1-year followup. Those success rates, combined with the enormous health benefits of smoking cessation, would likely make payment for some smoking cessation treatments cost- beneficial. For example, research by the Centers for Disease Control suggests that a smoking cessation program offered to all pregnant smokers could save $5 for every dollar spent by preventing low birthweiQht-associated neonatal intensive care and long-term care. ~ ii TIMN 438386

CHAPTER 1 INTRODUCTION, OVERVIEW, AND CONCLUSIONS TIMN 438416

ACKNOWLEDGMENTS This Report was prepared by the Department of Health and Human Services under the general editorship of the Office on Smoking and Health, Ronald M. Davis, M.D., Director. The Managing Editor was Susan A. Hawk, Ed.M., M.S. The scientific editors of the Report were: Jonathan M. Samet, M.D. (Senior Scientific Editor), Professor of Medicine and Chief, Pulmonary Division, Department of Medicine and the New Mexico Tumor Registry, Cancer Center, University of New Mexico, Albuquerque, New Mexico Ronald M. Davis, M.D., Director, Office on Smoking and Health, Center for Chronic Disease Prevention and Health Promotion (CCDPHP), Centers for Disease Control (CDC), Rockville, Maryland Neil E. Grunberg, Ph.D., Professor, Department of Medical Psychology, Uniformed Services University of the Health Sciences, Bethesda, Maryland Judith K. Ockene, Ph.D., Professor of Medicine, and Director, Division of Preventive and Behavibral Medicine, Department of Medicine, University of Massachusetts Medical School, Worcester, Massachusetts Diana B. Petitti, M.D., M.P.H., Associate Professor, Department of Family and Com- munity Medicine, University of California at San Francisco, School of Medicine, San Francisco, California Walter C. Willett, M.D., Dr.P.H., Professor of Epidemiology and Nutrition, Harvard School of Public Health, and The Channing Laboratory, Department of Medicine, Harvard Medical School and Brigham and Women's Hospital, Boston, Mas- sachusetts The following individuals prepared draft chapters or portions of the Report: Robert Anda, M.D., Epidemiologist, Office of Surveillance and Analysis, CCDPHP, CDC, Atlanta, Georgia John Baron, M.D., Associate Professor of Medicine, Department of Medicine, Dartmouth Medical School, Hanover, New Hampshire Tim Byers, M.D., M.P.H., Chief, Epidemiology Branch, Division of Nutrition, CCDPHP, CDC, Atlanta, Georgia Arden G. Christen, D.D.S., M.S.D., M.A., Chairman, Professor, Department of Preven- tive and Community Dentistry, Indiana University School of Dentistry, Indianapolis, Indiana Graham Colditz, Dr.P.H., Assistant Professor of Medicine, Harvard School of Public Health, and the Channing Laboratory, Department of Medicine, Harvard Medical School and_Brigham and Women's Hospital, Boston, Massachusetts TIMN 438396 X"'

22 times higher among male smokers and 12 times higher among female smokers compared with people who have never smoked.The risk of lung cancer declines steadily in people who quit smoking; after 10 years of abstinence, the risk of lung cancer is about 30 to 50 percent of the risk for continuing smokers. Smoking cessation also reduces the risk of cancers of the larynx, oral cavity, esophagus, pancreas, and urinary bladder. Coronary heart disease (CHD) is the leading cause of death in the United States. Smokers have about twice the risk of dying from CHD compared with lifetime nonsmokers. This excess risk is reduced by about half among ex-smokers after only 1 year of smoking abstinence and declines gradually thereafter. After 15 years of abstinence the risk of CHD is similar to that of persons who have never smoked. Compared with lifetime nonsmokers, smokers have about twice the risk of dying from stroke, the third leading cause of death in the United States. After quitting smoking, the risk of stroke returns to the level of people who have never smoked; in some studies this reduction in risk has occurred within 5 years, but in others as long as 15 years of abstinence were required. Cigarette smoking is the major cause of chronic obstructive pulmonary disease (COPD), the fifth leading cause of death in the United States. Smoking increases the risk of COPD by accelerating the a~e-related decline in lung function. With sustained abstinence from smoking, the rate of decline in lung function among former smokers returns to that of never smokers, thus reducing the risk of developing COPD. Influenza and pneumonia represent the sixth leading cause of death in the United States. Cigarette smoking increases the risk of respiratory infections such as influenza, pneumonia, and bronchitis, and smoking cessation reduces the risk. Cigarette smoking is a major cause of peripheral artery occlusive disease. This condition causes substantial mortality and morbidity; complications may include inter- mittent claudication. tissue ischemia and gangrene, and ultimately, loss of limb. Smoking cessation substantially reduces the risk of peripheral artery occlusive disease compared with continued smoking. The mortality rate from abdominal aortic aneurysm is two to five times higher in current smokers than in never smokers. Former smokers have half the excess risk of dying from this condition relative to current smokers. About 20 million Americans currently have, or have had, an ulcer of the stomach or duodenum. Smokers have an increased risk of developing gastric or duodenal ulcers, and this increased risk is reduced by quitting smoking. Benefits at All Ages According to a 1989'Gallup survey, the proportion of smokers who say they would like to give up smoking is lower for smokers aged 50 and older (57 percent) than for smokers aged 18-29 (68 percent) and 30-49 (67 percent). Older smokers may be less motivated to quit smoking because the highly motivated may have quit already at younger ages, leaving a relatively "hard-core" group of older smokers. But many long-term smokers may lack motivation to quit for other reasons. Some may believe they are no longer at risk of smoking-related diseases because they have already survived smoking for many years. Others may believe that any damage that may have vi TIMN 438389

TABLE OF CONTENTS Foreword ........................................................... i Preface ............................................................. v Acknowledgments .................................................. xiii List of Tables ...................................................... xxv List of Figures ..................................................... xxxi 1. Introduction, Overview, and Conclusions ............................. 1 2. Assessing Smoking Cessation and Its Health Consequences .............. 17 3. Smoking Cessation and Overall Mortality and Morbidity ................ 71 4. Smoking Cessation and Respiratory Cancers ......................... 103 5. Smoking Cessation and Nonrespiratory Cancers ...................... 143 6. Smoking Cessation and Cardiovascular Disease ...................... 187 7. Smoking Cessation and Nonmalignant Respiratory Diseases ............ 275 8. Smoking Cessation and Reproduction .............................. 367 9. Smoking, Smoking Cessation, and Other Nonmalignant Diseases ........ 425 10. Smoking Cessation and Body Weight Change ........................ 469 11. Psychological and Behavioral Consequences and Correlates of Smoking Cessation ............................................. 517 Volume Appendix. National Trends in Smoking Cessation ................ 579 Glossary .......................................................... 617 Index............................................................. 619 xxui TIMN 438406

Previous Surgeon General's reports, in particular the landmark 1964 Report of the Surgeon General's Advisory Committee on Smoking and Health and the 1982 Surgeon. General's Report on smoking and cancer, examined these associations with respect to the epidemiologic criteria for causality. These criteria include the consistency, strength, specificity, coherence, and temporal relationship of the association. Based on these criteria, previous reports have recognized a causal association between smoking and cancers of the lung, larynx, esophagus, and oral cavity; heart disease; stroke; peripheral artery occlusive disease; chronic obstructive pulmonary disease; and intrauterine growth retardation. This Surgeon General's Report is the first to conclude that the evidence is now sufficient to identify cigarette smoking as a cause of cancer of the urinary bladder; the 1982 Report concluded that cigarette smoking is a contributing factor in the development of bladder cancer. The causal nature of most of these associations was well established lona before publication of this Report. Nevertheless, it is worth noting that the findings of this Report add even more weight to the evidence that these associations are causal. The criterion of coherence requires that descriptive epidemiologic findings on disease occurrence correlate with measures of exposure to the suspected agent. Coherence would predict that the increased risk of disease associated with an exposure would diminish or disappear after cessation of exposure. As this Report shows in great detail. the risks of most smoking-related diseases decrease after cessation and with increasing duration of abstinence. Evidence or} the risk of disease after smoking cessation is especially important for the understanding of smoking-and-disease associations of unclear causality. For ex- ample, cigarette smoking is associated with cancer of the uterine cervix, but this association is potentially confounded by unidentified factors (in particular by a sexually transmitted etiologic agent). The evidence reviewed in this Report indicates that former smokers experience a lower risk of cervical cancer than current smokers, even after adjusting for the social correlates of smoking and risk of sexually acquired infections. This diminution of risk after smoking cessation supports the hypothesis that smoking is a contributing cause of cervical cancer. Conclusion The Comprehensive Smoking Education Act of 1984 (Public Law 98-474) requires the rotation of four health warnings on cigarette packages and advertisements. One of those warnings reads, "SURGEON GENERAL'S WARNING: Quitting Smoking Now Greatly Reduces Serious Risks to Your Health." The evidence reviewed in this Report confirms and expands that advice. The health benefits of quitting smoking are immediate and substantial. They far exceed any risks from the average 5-pound weight gain or any adverse psychological effects that may follow quitting. The benefits extend to men and women, to the young and the old, to those who are sick and to those who are well. Smoking cessation represents the single most important step that smokers can take to enhance the length and quality of their lives. xi TIMN 438394

Smoking cessation programs and messages should emphasize that weight gain after quitting is small on average. Not only is the average postcessation weight gain small, but the risk of large weight gain after quitting is extremely low. Less than 4 percent of those who quit smoking gain more than 20 pounds. Nevertheless, special advice and assistance should be available to the rare person who does gain considerable weight after quitting. For these individuals, the health benefits of cessation still occur, and weight control programs rather than smoking relapse should be implemented. Increases in food intake and decreases in resting energy expenditure are largely responsible for postcessation weight gain. Thus, dietary advice and exercise should be helpful in preventing or reducing postcessation weight gain. Unfortunately, minor weight control modifications to smoking cessation programs do not generally yield beneficial effects in terms of reducing weight gain or increasing cessation rates. A few studies have investigated pharmacologic approaches to postcessation weight control; preliminary results are encouraging but more research is needed. High priority should be given to the development and evaluation of effective weight control programs that can be targeted in a cost-effective manner to those at greatest need of assistance. Psychological and Behavioral Consequences of Smoking Cessation Nicotine withdrawal symptoms include anxiety, irritability, frustration, anger, dif- ficulty concentrating, increased appetite. and urges to smoke. With the possible exception of urges to smoke and increased appetite, these effects soon disappear. Nicotine withdrawal peaks in the first I to 2 days following cessation and subsides rapidly during the following weeks. With long-term abstinence, former smokers are likely to enjoy favorable psychological changes such as enhanced self-esteem and increased sense of self-control. Although most nicotine withdrawal symptoms are short-lived, they often exert a strong influence on smokers' ability to quit and maintain abstinence. Nicotine withdrawal may discourage many smokers from trying to quit and may precipitate relapse among those who have recently quit. In the 1986 Adult Use of Tobacco Survey, 39 percent of current smokers reported that irritability was a "very important" or "somewhat important" reason why they resumed smoking after a previous quit attempt. Smokers and ex-smokers should be counseled that adverse psychological effects of smoking subside rapidly over time. Smoking cessation materials and programs, nicotine replacement, exercise, stress management, and dietary counseling can help smokers cope with these symptoms until they abate, after which favorable psychologi- cal changes are likely to occur. Support for a Causal Association Between Smoking and Disease Tens of thousands of studies have documented the associations between cigarette smoking and a large number of serious diseases. It is safe to say that smoking represents the most extensively documented cause of disease ever investigated in the history of biomedical research. x TIMN 438393

Carlo C. DiClemente, Ph.D., Associate Professor, Department of Psychology, Univer- sity of Houston, Houston, Texas Douglas W. Dockery, Sc.D., Associate Professor, Department of Environmental Health, Environmental Epidemiology Program, Harvard School of Public Health, Boston, Massachusetts Gary A. Giovino, Ph.D., Acting Chief, Epidemiology Branch, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Deborah Grady, M.D.; Assistant Professor, Departments of Epidemiology and Medicine, University of California at San Francisco, School of Medicine, San Francisco, California Neil E. Grunberg, Ph.D., Professor, Department of Medical Psychology, Uniformed Services University of the Health Sciences, Bethesda, Maryland John R. Hughes, M.D., Associate Professor, Human Behavioral Pharmacology Laboratory, Departments of Psychiatry, Psychology, and Family Practice, University of Vermont, Burlington, Vermont Robert W. Jeffery, Ph.D., Professor, Division of Epidemiology, School of Public Health, University of Minnesota, Minneapolis, Minnesota LTC James W. Kikendall, M.D., Assistant Chief, Gastroenterology Section, Walter Reed Army Medical Center, Washington, D.C. Robert Klesges, Ph.D., Associate Professor, Department of Psychology, Memphis State University, Memphis, Tennessee Lynn Kozlowski, Ph.D., Head, Behavioral Tobacco Research, Socio-behavioral Re- search Department, Addiction Research Foundation, Toronto, Ontario, Canada Stephen Marcus, Ph.D., Epidemiologist, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland James L. McDonald, Jr., Ph.D., Assistant Chaitman. Professor, Department of Preven- tive and Community Dentistry, Indiana University School of Dentistry, Indianapolis, Indiana Sherry L. Mills, M.D.. M.P.H., Medical Officer, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Judith K. Ockene, Ph.D., Professor of Medicine, and Director. Division of Preventive and Behavioral Medicine, Department of Medicine, University of Massachusetts Medical School, Worcester, Massachusetts Carole Tracy Orleans, Ph.D., Director, Smoking Cessation Services, Fox Chase Cancer Center, Cheltenham, Pennsylvania Diana B. Petitti; M.D., M.P.H., Associate Professor, Department of Family and Com- munity Medicine, University of California at San Francisco, School of Medicine, San Francisco, California John P. Pierce, Ph.D., Associate Professor, Director, Population Studies and Cancer Prevention, Tobacco Control Project. University of California, San Diego Cancer Center, San Diego, California Paul R. Pomrehn, Ph.D., M.S., Associate Professor, Department of Preventive Medicine and Environmental Health. University of Iowa College of Medicine, Iowa City, Iowa James O. Prochaska, Ph.D., Professor, Director, Cancer Prevention Research Unit, Department of Psychology, University of Rhode Island, Kingston, Rhode Island xiv TIMN 438397

smoking. Smoking cessation reduces the risk of respiratory infections such as pneumonia, which are often the immediate causes of death in patients with an under- lying chronic disease. The important role of health care providers in counseling patients to quit smoking is well recognized. Health care providers should give smoking cessation advice and assistance to all patients who smoke, including those with existing illness. Benefits for the Fetus Maternal smoking is associated with several complications of pregnancy including abruptio placentae, placenta previa, bleeding during pregnancy, premature and prolonged rupture of the membranes, and preterm delivery. Maternal smoking retards fetal growth, causes an average reduction in birthweight of 200 g, and doubles the risk of having a low birthweight baby. Studies have shown a 25- to 50-percent higher rate of fetal and infant deaths among women who smoke during pregnancy compared with those who do not. Women who stop smoking before becoming pregnant have infants of the same birthweight as those born to women who have never smoked. The same benefit accrues to women who quit smoking in the first 3 to 4 months of pregnancy and who remain abstinent throughout the remainder of pregnancy. Women who quit smoking at later stages of pregnancy, up to the 30th week of gestation, have infants with higher birthweight than do women who smoke throughout pregnancy. Smoking is probably the most important modifiable cause of poorpregnancy outcome among women in the United States. Recent estimates suggest that the elimination of smoking during pregnancy could prevent about 5 percent of perinatal deaths, about 20 percent of low birthweight births, and about 8 percent of preterm deliveries in the United States. In groups with a high prevalence of smoking (e.g., women who have not completed high school), the elimination of smoking during pregnancy could prevent about 10'percent of perinatal deaths. about 35 percent of low birthweight births, and about 15 percent of preterm deliveries. The prevalence of smoking during pregnancy has declined over time but remains unacceptably high. Approximately 30 percent of U.S. women who are cigarette smokers quit after recognition of pregnancy, and others quit later in pregnancy. However, about 25 percent of pregnant women in the United States smoke throughout pregnancy. A shocking statistic is that half of pregnant women who have not completed high school smoke throughout pregnancy. Many women who do not quit smoking during pregnancy reduce their daily cigarette consumption; however, reduced con- sumption without quitting may have little or no benefit for birthweight. Of the women who quit smoking during pregnancy. 70 percent resume smoking within I year of delivery. Initiatives have been launched in the public and private sectors to reduce smoking during pregnancy. These programs should be expanded, and less educated pregnant women should be a special target of these efforts. Strategies need to be developed to address the problem of relapse after delivery. viii TIMN 438391

LIST OF TABLES Chapter 2 Table 1. Measures of false reports of not smoking from studies using nicotine and cotinine as a marker .................................... 38 Table 2. Measures of false reports from studies using CO as a marker ......... 41 Table 3. Examples of potential methodologic problems in investigating the health consequences of smoking cessation .......................... 47 Chapter 3 Table 1. Summary of longitudinal studies of overall mortality ratios relative to never smokers among male current and former smokers according to duration of abstinence (when reported) ..................... 76 Table 2. Overall mortality ratios among current and former smokers, relative to never smokers, by sex and duration of abstinence at date of enrol lment, ACS CPS-II ........................................... 78 Table 3. Estimated probability of dying in the next 16.5-year interval for quitting at various ages compared with never smoking and continuing to smoke, by amount smoked and sex ................................ 83 Table 4. Summary of overall mortality ratios in intervention studies in which smoking cessation was a component ...................... . ..... 84 Table 5. Summary of studies of medical care utilization among smokers andformersmokers .............................................. 88 Table 6. Relation of smoking cessation to various measures of general health status ...................................................90 Table 7. Age- and sex-specific mortality rates among never smokers, continuing smokers, and former smokers by amount smoked and duration of abstinence at time of enrollment for subjects in ACS CPS-II study who did not have a history of cancer, heart disease, or stroke and were not sick at enrollment ................................ 95 Table 8. Estimated probability of dying in the next 16.5-year interval (95% CI) for quitting at various ages compared with never smoking and continuing to smoke, by amount smoked and sex .................... 97 TIMN 438407 Xxv .

Chapter 4 Table 1. Histologic changes (%) in bronchial epithelium by smoking status .........................................................109 Table 2. Relative risks of lung cancer among never, former, and current smokers in selected epidemiologic studies ............................ 111 Table 3. Lung cancer mortality ratios among never, current, and. former smokers by number of years since stopped smoking (relative to never smokers), prospective studies ...................................... 112 Table 4. Relative risks of lung cancer among former smokers, by number of years since stopped smoking, and current smokers, from selected case-control studies ...................................... 115 Table 5. Relative risks of lung cancer among never, current, and former smokers, by number of years since stopping smoking and histologic type .......................................................... 119 Table 6. Relative risks of lung cancer among never, former, and current smokers by types of tobacco products smoked ........................ 120 Table 7. Standard mortality ratios of lung cancer among former smokers in ACS CPS-II (relative to never smokers) by years of smoking abstinence, daily cigarette consumption at time of cessation, and history of chronic disease ......................................... 130 Table 8. Histologic changes in laryngeal epithelium by smoking status ...... 132 Table 9. Relative risks of laryngeal cancer by smoking status . .............. 133 Chapter 5 Table 1. Studies of oral cancer and smoking cessation ................... 148 Table 2. Studies of esophageal cancer that have examined the effect of smoking cessation ............................................... 153 Table 3. Studies of cancer of the pancreas and smoking cessation .......... 156 Table 4. Studies of bladder cancer and smoking cessation ................ 160 Table 5. Bladder cancer risk according to smoking dose, duration of smoking, and smoking status ...................................... ~ 165 Table 6. Studies of cervical cancer and smoking cessation ................ 167 Table 7. Studies of breast cancer and smoking cessation .................. 170 Table 8. Studies of cancer at selected sites that have examined the effect of smoking cessation ............................................. 173 xxvi TI~~ 438408

Chapter 8 Figure 1. Perinatal, neonatal, and fetal mortality rates by birthweight in singleton white males, 1980 ....................................... 380 Chapter 11 Figure 1. Performance on a meter (i.e., visual) vigilance task. Performance on the continuous clock task, a visual vigilance task ........ 527 Figure 2. Self-reported withdrawal discomfort among abstinent smokers .... 531 Figure 3. Drinking relative to smoking status for men, 1983 NHIS ......... 558 Figure 4. Drinking relative to smoking status for women, 1983 NHIS ....... 559 Appendix Figure 1. Trends in the quit ratio, United States, 1965-87, by gender ........ 590 Figure 2. Trends in the quit ratio, United States, 1965-87, by race .......... 591 Figure 3. Flow chart of quitting history, attempts lasting longer than 1 year, NHEFS .................................................. 597 Figure 4. Estimated duration of abstinence on first 1-year or longer quit attempt, product-limit method, N=3,363 ............................. 598 Figure 5. Percentage of ever smokers who never tried to quit, by education, United States, 1974-87 .................................. 601 Figure 6. Percentage of persons smoking at 12 months prior to the survey interview who quit for at least I day during those 12 months, United States, 1978-80, 1987, by education .......................... 602 Figure 7. Percentage of ever smokers who had been abstinent for less than 1 year, United States, 1966-87, by education ..................... 603 Figure 8. PercentaQe of ever smokers who had been abstinent for 1-4 years, United States, 1966-87, by education .......................... 604 Figure 9. Percentage of ever smokers who had been abstinent for 5 years or more, United States, 1966-87, by education ........................ 605 ~jM-S 438415 xXXiii

Table 6: Summary of data from 1987 BRFSS, behaviors of former smokers aged 18 and older by duration of abstinence ................... 552 Table 7. Percent distribution of persons aged 18 and older by tobacco product and use status, according to gender and cigarette smoking status, United States, 1987 ........................................ 557 Table 8. Physician visits and medical tests within the past year among AARP members aged 50 and older, by smoking status .................. 563 Volume Appendix Table 1. Quit ratio in selected States, by age group and gender-BRFSS, 1988 ........................................................... 586 Table 2. Cigarette smoking continuum by year, percentage of ever cigarette smokers, by NHISs, United States, 1978-87, adults aged 20 and older ...................................................... 589 Table 3. Trends in quit ratio (%) (percentage of ever cigarette smokers who are former cigarette smokers), by age and by education. NHISs, United States, 1965-87, adults aged 20 and older ...................... 592 Table 4. Effect of adjusting for use of other tobacco products on quit ratio (percentage of ever cigarette smokers who are former cigarette smokers), 1987, NHIS, United States ................... ............. 594 Table 5. Selected measures of quitting activity (%), NHISs, United States, adults aged 20 and older .................................... 600 Table 6. Percentage of those intending to smoke in 5 years, by gender, AUTSs, United States, 1964-86, current smokers aged 21 and older ....... 609 Table 7. Percentage who report having ever received advice to quit from a doctor, by smoking status and gender, United States. 1964-87, adults aged 21 and older ...............................................610 xxx Tlldil~ 438412

William A. Robinson, M.D., M.P.H., Director, Office of Minority Health. Department of Health and Human Services, Washington, D.C. William L. Roper, M.D., M.P.H., Director, CDC, Atlanta, Georgia Richard B. Rothenberg, M.D., Assistant Director for Science, CCDPHP, CDC. Atlanta, Georgia Thomas C. Schelling, Ph.D., Director, Institute for the Study of Smoking Behavior and Policy, Lucius N. Littauer Professor of Political Economy, Harvard University, Cambridge, Massachusetts Marc B. Schenker, M.D., M.P.H., Associate Professor and Division Chief, Occupation- al and Environmental Medicine. University of California, Davis, Davis, California David Schottenfeld, M.D., Professor and Chairman, Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, Michigan Kathleen L. Schroeder, D.D.S., M.Sc., Assistant Professor, Section of Oral Biology, The Ohio State University College of Dentistry, Columbus, Ohio Mary J. Sexton, Ph.D., M.P.H., Professor, Department of Epidemiology and Preventive Medicine. University of Maryland School of Medicine, Baltimore, Maryland Saul Shiffman, Ph.D., Associate Professor, Department of Psychology, University of Pittsburgh, Pittsburgh, Pennsylvania Donald Shopland, Smokins. Tobacco, and Cancer Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland Amnon Sonnenberg, M.D., Associate Professor, Gastroenterology Section, Medical College of Wisconsin, Veterans Administration Medical Center, Milwaukee. Wis- consin Frank E. Speizer, M.D., Professor of Medicine, Harvard Medical School, Professor of Environmental Epidemiology, Harvard School of Public Health, Co-Director, The Channing Laboratory, Department of Medicine, BriQham and Women's Hospital, Boston, Massachusetts ~ Jesse Steinfeld, M.D., San Diego, California Steven D. Steliman, Ph.D., Assistant Commissioner, New York City Department of Health, New York, New York Ira B. Tager, M.D., M.P.H., Associate Professor of Medicine and Epidemiology and Biostatistics, University of California, San Francisco, Veterans Administration Medi- cal Center, San Francisco, San Francisco, California Kenneth Warner, Ph.D., Senior Fellow, Institute of Gerontology, University of Michigan, Ann Arbor, Michigan Jonathan S. Weiss, M.D., Assistant Professor of Dermatolosy, Section of Dermatology, Emory Clinic, Atlanta, Georgia ~ Noel S. Weiss, M.D., Dr.P.H., Professor and Chairman, Department of Epidemiology, University of Washington, Seattle, Washington y Gail R. Wilensky, Ph.D., Administrator, Health Care Financing Administration. Washington, DC ~ Deborah Winn, Ph.D., Deputy Director, Division of Health Interview Statistics, Na- tional Center for Health Statistics, CDC, Hyattsville, Maryland Philip A. Wolf, M.D., Professor of Neurology, Department of Neurology, Boston University School of Medicine, Boston, Massachusetts Ernst L. Wynder, M.D., President, American Health Foundation, New York, New York TIMN 438402 xiX

U Sharon K. Faupel, Staff Assistant, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Leanna Fernando, Administrative Assistant, New Mexico Tumor Registry, University of New Mexico, Albuquerque, New Mexico David Fry, Editor, The Circle, Inc., McLean, Virginia Lynn Funkhauser, Word Processing Specialist, The Circle, Inc., McLean, Virginia Amy Garson, Student Intern, Office on Smoking and Health, CCDPHP, CDC, Rock- ville, Maryland Mary Graber, Secretary, University of California at San Francisco, School of Medicine, Department of Family and Community Medicine, San Francisco, California Gwen Harvey, Program Analyst, CCDPHP, CDC, Atlanta, Georgia Patricia Healy, Technical Information Specialist, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Phyllis E. Hechtman. Editorial Assistant, The Circle, Inc., McLean, Virginia Timothy K. Hensley, Technical Publications Writer-Editor, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Julian Hudson, Courier, The Circle, Inc., McLean, Virainia Beth Jacobsen, Student Intern, Office on Smoking and Health, CCDPHP. CDC, Rockville, Maryland Renee Kolbe, Program Specialist, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Matt Kreuter, Public Information Specialist, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Peggy Lytton. Editor, The Circle, Inc.. McLean, Virginia Diana Lord, Research Psychologist, Department of Medical Psychology, Uniformed Services University of the Health Sciences, Bethesda, Maryland Daniel F. McLaughlin, Editor, The Circle, Inc., McLean, Virginia Jackie L. Meador, Desktop Publishing/Word Processing Specialist, The Circle. Inc., McLean, Virginia Elaine Medoff-IVicGovern, Medical Secretary, Division of Preventive and Behavioral Medicine, Department of Medicine, University of Massachusetts Medical School, Worcester, Massachusetts Nancy A. Miltenberger, M.A., Production Editor, The Circle, Inc., McLean. Virginia Rebecca Mosher, Staff Assistant, New Mexico Tumor Registry, University of New Mexico, Albuquefque, New Mexico Millie R. Naquin. Research Assistant, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Thomas E. Novotny, M.D., Chief, Program Services Activity, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Cathie M. O'Donnell, Project Director, The Circle, Inc., McLean, Virginia Christine Pappas, Editorial Research Assistant, The Channing Laboratory, Harvard School of Public Health, Boston, Massachusetts Stacey M. Parcover, Secretary, Office on Smoking and Health, CCDPHP, CDC, Rockville, Maryland Lida Peterson, Computer Systems Manager, The Circle, Inc., McLean, Virginia TIMN 438404 xxi

LIST OF FIGURES Chapter 2 Figure 1. Cyclical model of the stages of change ......................... 23 Figure 2. Hypothetical examples of disease incidence rates for current, fotmer, and never smokers, by age .................................... 55 Chapter 3 Figure 1. Compared with never smokers, relative risk of mortality in current and former smokers aged 50-54, 60-64, and 70-74 at enrollment, by amount smoked and duration of abstinence ................ 81 Figure 2. Estimated probability of dying in the next 16.5-yr interval for quitting at ages 55-59 compared with never smoking and continuing to smoke, by sex ................................................... 98 Chapter 4 Figure 1. Risk of lung cancer by number of cigarettes smoked per day before quitting, number of years of abstinence, sex, and histologic types ... 121 Figure 2. Relative risk of lung cancer among ex-smokers compared with continuing smokers as a function of time since stopped smoking, estimated from locristic regression model, pattern adjusted for smoking duration compared with pattern unadjusted for duration ................. .123 Figure 3. Incidence of bronchial carcinoma among continuing cigarette smokers in relation to age and duration of smokine and among never smokers in relation to age, double logarithmic scale .................... 127 Chapter 6 Figure 1. Hypothetical effects of smoking cessation on risk of CHD if mechanisms' are predominantly rapidly reversible ...................... 198 Figure 2. Estimated relative risk of MI after quitting smoking among men under age 55, adjusted for age ................................. 204 Figure 3. Mortality ratios due to coronary artery diseases; rates for men who have stopped smoking are compared with those for men who never smoked and those for men still smoking in 1952 .................. 214 Figure 4. Mortality ratios for all cardiovascular diseases and CHD, by daily cigarette consumption, US Veterans Study, 1954-69 ............... 219 Figure 5. Mortality ratio for current and former cigarette smokers by years of smoking cessation, US Veterans Study, 1954-69 ................ 220 TIMN 438413 Xxxi

been caused by smoking is irreversible after decades of smoking. For similar reasons, many physicians may be less likely to counsel their older patients to quit. CPS-II data were used to estimate the effects of quitting smoking at various ages on the cumulative risk of death during a fixed interval after cessation. The results show that the benefits of cessation extend to quitting at older ages. For example, a healthy man aged 60-64 who smokes I pack of cigarettes or more per day reduces his risk of dying during the next 15 years by 10 percent if he quits smoking. These findings support the recommendations of the Surgeon General's 1988 Workshop on Health Promotion and Aging for the development and dissemination of smoking cessation messages and interventions to older persons. I am pleased that a coalition of organizations and agencies is now working toward implementation of those recommendations, including the Centers for Disease Control; the National Cancer Institute; the National Heart, Lung, and Blood Institute; the Administration on Agine; the Department of Veterans Affairs; the Office of Disease Prevention and Health Promotion; the American Association of Retired Persons; and the Fox Chase Cancer Center. The major message of this campaign will be that it is never too late to quit smoking. Two facts point to the urgent need for a strong smoking cessation campaign targeting older Americans: (1) 7 million smokers are aged 60 or older; and (2) smoking is a major risk factor for 6 of the 141eading causes of death among those aged 60 and older, and is a complicating factor for 3 others. ~ Benefits for Smokers with Existing Disease Many smokers who have already developed smoking-related disease or symptoms may be less motivated to quit because of a belief that the damage is already done. For the same reason, physicians may be less motivated to advise these patients to quit. However, the evidence reviewed in this Report shows that smoking cessation yields important health benefits to those who already suffer from smoking-related illness. Among persons with diagnosed CHD, smoking cessation markedly reduces the risk of recurrent heart attack and cardiovascular death. In many studies, this reduction in risk has been 50 percent or more. Smoking cessation is the most important intervention in the management of peripheral artery occlusive disease; for patients with this condi- tion, quitting smoking improves exercise tolerance, reduces the risk of amputation after peripheral artery surgery, and increases overall survival. Patients with gastric and duodenal ulcers who stop smoking improve their clinical course relative to smokers who continue to smoke. AlthouQh the benefits of smoking cessation amona stroke patients have not been studied, it is reasonable to assume that quitting smoking reduces the risk of recurrent stroke just as it reduces the risk of recurrence of other cardiovascular events. Even smokers who have already developed cancer may benefit from smoking cessation. A few studies have shown that persons who stopped smoking after diagnosis of cancer had a reduced risk of acquiring a second primary cancer compared with persons who continued to smoke. Although relevant data are sparse, longer survival might be expected among smokers with cancer or other serious illnesses if they stop vu TIMN 438390

Table 12. Prospective studies of COPD mortality in relation to cigarette smoking status .................................................. 343 Table 13. Standardized mortality ratios for COPD among current and former smokers broken down by years of abstinence ................... 346 Chapter 8 Table 1. Possible mechanisms for effect of smoking on pregnancy and pregnancy outcome .............................................. 372 Table 2. Summary of studies of fertility among smokers and former smokers ....................................................... 375 Table 3. Summary of studies of perinatal and neonatal mortality in smokers and nonsmokers during pregnancy .......................... 377 Table 4. Estimated relative risk of fetal plus infant mortality for maternal smoking in several birthweight groups, adjusting for maternal marital status, education, age, and parity ................................... 378 Table 5. Summary of studies of perinatal mortality in smokers throughout pregnancy, smokers who quit in the early months of pregnancy, and nonsmokers during pregnancy ........................ 379 Table 6. Summary of studies of mean birthweight, by smoking status ....... 382 Table 7. Summary of nonexperimental studies of smoking cessation after conception. mean increase (+) or decrease (-) in birthweight (g) according to timing of cessation ........................... '......... 384 Table 8. Summary of nonexperimental studies of relative risk of low birthweight for smoking cessation after conception .................... 385 Table 9. Summary of birthweight outcome in randomized trials of smoking cessation in preQnancy .................................... 388 Table 10. Smoking and smoking cessation during pregnancy, summary of results of two surveys of national probability samples ................ 391 Table 11. Patterns of smoking cessation during pregnancy among selected populations ............................................. 394 Table 12. Summary of studies that estimated relative risk of various pregnancy outcomes for smoking based on a "synthesis" of the literature, and attributable risk percent based on several estimates of the prevalence of smoking during pregnancy .......................... 395 Table 13. Summary of studies reporting relationship of cigarette smoking and age at natural menopause .............................. 397 Table 14. Summary of studies of age of natural menopause amonQ former smokers ..................................................399 Table 15. Sexual performance among male former smokers ............... 404 Table 16. Sperm quality among smokers and nonsmokers ................ 406 xxviii - TIMN 438410

Figure 6. Effect of smoking cessation on survival among men with documented coronary atherosclerosis; pooled survival among quitters (N=1,490) and continuers (N=2,675) ................................ 238 Figure 7. Mortality ratios for stroke for current smokers and ex-smokers compared with never smokers, by daily cigarette consumption, US Veterans Study, 1954-69 ......................................... 252 Figure 8. Survival free of stroke in cigarette smokers, never smokers, and former smokers, aged 60, using Cox proportional hazard regression model, among men and women ........................... 259 Chapter 7 Figure 1. Nonproportional Venn Diagram of the interrelationship among chronic bronchitis, emphysema, asthma, and airways obstruction ......... 280 Figure 2. Theoretical curves depicting varying rates of decline of FEV I ..... 281 Figure 3. Hypothesized mechanisms by which airway hyperresponsiveness may be associated with developing or established COPD without necessarily being a preexisting risk factor ............... 284 Figure 4. Symptom ratio (number of observed symptoms to number of possible symptoms) in nonmodifiers, modifiers, and quitters at each test period; symptoms are cough, sputum production, wheezing, and shortness of breath .............................................. ~287 Figure 5. Prevalence of cough and phlegm by smoking group ............. 293 Figure 6. Prevalence of dyspnea by smoking group ...................... 295 Figure 7. Sex-specific mean height-adjusted FEV I residuals versus pack-years for current and ex-smokers, and distributions of number of subjects by pack-years ........................................... 317 Figure 8. Mean values FVC and FEV 1, expressed as a percentage of predicted values, in 15 quitters and 42 smokers during 30 months after 2 smoking cessation clinics ....................................... 322 Figure 9. Mean values for the ratio of CV/VC, of CC/TLC, and slope for phase III of the single breath N2 test (N2/L), expressed as a percentage of predicted values in 15 quitters and 42 smokers during 30 months after 2 smoking cessation clinics ................................... 326 Figure 10. Percent-predicted diffusing capacity (%pDL) by pack-years of smoking, current smokers and former smokers, in a study of adults in Tucson, AZ .................................................. 329 Figure 11. Mean AFEV I values in never smokers, consistent ex-smokers, subjects who quit smoking during followup, and consistent smokers in several age groups ............................. 334 Figure 12. Effects of quitting smoking during followup among men aged 50-69 ......................................................... 336 xxxii TIMN 438414

CONTENTS Introduction ......................................................... 5 MajorConciusions .................................................... 8 Development of the Report ............................................. 8 Chapter Conclusions .................................................. 9 Chapter 2: Assessing Smoking Cessation and Its Health Consequences ........ 9 Chapter 3: Smoking Cessation and Overall Mortality and Morbidity .......... 9 Chapter 4: Smoking Cessation and Respiratory Cancers .................... 10 Chapter 5: Smoking Cessation and Nonrespiratory Cancers ................ 10 Chapter 6: Smoking Cessation and Cardiovascular Disease ................ 10 Chapter 7: Smoking Cessation and Nonmalignant Respiratory Diseases ...... 11 Chapter 8: Smoking Cessation and Reproduction ........................ 11 Chapter 9: Smoking, Smoking Cessation, and Other Nonmalignant Diseases .. 12 Chapter 10: Smoking Cessation and Body Weight Change ................. .12 Chapter 11: Psychological and Behavioral Consequences and Correlates of Smoking Cessation ............................................. 13 Volume Appendix: National Trends in Smoking Cessation ................ 13 References ......................................................... 15 ,VIMIS 438417 3

Chapter 3. Smoking Cessation and Overall Mortality and Morbidity Chapter 4. Smoking Cessation and Respiratory Cancers Chapter 5. Smoking Cessation and Nonrespiratory Cancers Chapter 6. Smoking Cessation and Cardiovascular Disease Chapter 7. Smoking Cessation and Nonmalignant Respiratory Diseases Chapter 8. Smoking Cessation and Reproduction Chapter 9. Smoking, Smoking Cessation, and Other Nonmalignant Diseases Chapter 10. Smoking Cessation and Body Weight Change Chapter 11. Psychological and Behavioral Consequences and Correlates of Smoking Cessation Volume Appendix. National Trends in Snioking Cessation A key to acronyms and terms used throughout the Report is found at the end of the volume. Other publications of the Public Health Service have reviewed determinants of smoking cessation and abstinence (US DHEW 1979; US DHHS 1980, 1988) and methods of smoking 'cessation and relapse prevention (Schwartz 1987; US DHHS 198$); hence, these topics are not cove'red in this Report. Beginning with the 1964 Report, the evidence on active smoking and disease has been reviewed for causality to evaluate the associations of smoking with disease. The explicit criteria used in this evaluation include the consistency, strength, specificity, temporal relationship, and coherence of the association (US PHS 1964; US DHHS 1989). These criteria have provided a consistent and effective framework for examin- ing the epidemiologic, clinical, and experimental data on active smoking. Although the criteria cannot be applied in the same fashion to associations of smoking cessation with changes in disease occurrence, the criteria of consistency, an appropriate temporal relationship, and coherence must be maintained with evidence on smoking cessation and health. Thus, this Report examines data for consistency among investigations of the associa- tions of cessation with disease occurrence and other outcomes, and considers the biologic plausibility of the known or presumed associations in the context of the mechanisms by which cigarette smoking is known or thought to cause disease. The appropriate time sequence of cessation with its effect is evident; cessation must always precede its presumed effect. In an observational study, this sequence may be reversed by the tendency of persons with initial symptoms of a cigarette-related disease or with frank disease to reduce cigarette consumption or to stop smoking (Chapter 2). The findings of longitudinal studies among former smokers document high mortality rates among short-term former smokers, which is consistent with reversal of the causal 6 TIMN 438419

scientists for their review and comment on the entirety of its contents. Simultaneously, the draft Report was submitted to 10 institutes and agencies within the U.S. Public Health Service for review. Comments from the senior scientific reviewers and the agencies were then used to prepare the final draft of the Report, which was then reviewed by the Office of the Assistant Secretary for Health and the Secretary, Department of Health and Human Services. CHAPTER CONCLUSIONS Chapter 2: Assessing Smoking Cessation and Its Health Consequences 1. Most former smokers have cycled several times through the process of smoking cessation and relapse before attaining long-term abstinence. Any static measure of smoking status is thus a simplification of a dynamic process. 2. In studies of the health effects of smoking cessation, persons classified as former smokers may include some current smokers. Consequently, the health benefits of smoking cessation are likely to be underestimated. 3. In contexts other than intervention trials, self-reported smoking status at the time of measurement and concurrent biochemical assessment are highly concordant. This high concordance supports self-report as a valid measure of smoking status in observational studies of the health effects of smoking cessation. Chapter 3: Smoking Cessation and Overall Mortality and Morbidity 1. Former smokers live longer than continuing smokers, and the benefits of quitting extend to those who quit at older ages. For example, persons who quit smoking before age 50 have one-half the risk of dying in the next 15 years compared with continuing smokers. 2. Smoking cessation at all ages reduces the risk of premature death. 3. Among former smokers, the decline in risk of death compared with continuing smokers begins shortly after quitting and continues for at least 10 to 15 years. After 10 to 15 years of abstinence, risk of all-cause mortality returns nearly to that of persons who never smoked. 4. Former smokers have better health status than current smokers as measured in a variety of ways, including days of illness, number of health complaints, and self-reported health status. TIMN 438422 9

4. Quitting activity, as measured by the proportion of people smoking at 12 months before a survey who quit for at least I day during those 12 months, has increased slightly over time. Between 1978 and 1987, this proportion increased from 27.8 to 31.6 percent. 5. Female smokers were more likely than male smokers to have quit smoking cigarettes for at least I day during the previous year; however, there were no gender differ- ences in the proportion abstinent for I to 4 years. Men were more likely than women to have been abstinent for 5 years or more. These findings do not take into account the use of tobacco products other than cigarettes. 6. Black smokers were more likely than white smokers to have quit for at least I day during the previous year. Blacks, however, were less likely than whites to have been abstinent for 1 year or more. 7. Younger smokers (aged 20 to 44) were more likely than older smokers to have quit for at least 1 day during the previous year. 8. Smokers with less education tend to be less likely to have quit for at least 1 day during the previous year compared with those having more education. In addition, those with lower levels of education are less likely to have been abstinent for 1 year or more. 9. In .1964, about three-fourths of all current smokers predicted that they would "definitely" or "probably" be smoking in 5 years. In 1986, fewer than half of all current smokers felt the same way. Moreover, while more than 20 percent of current smokers in 1964 predicted that they would "definitely" be smoking in 5 years, only about 7 percent of current smokers in 1986 so predicted. 10. Current smokers in 1987 were more than three times as likely as current smokers in 1964 to report having received advice from a doctor to stop smoking. 14 TIMN 438427

CHAPTER 2 ASSESSING SMOKING CESSATION AND ITS HEALTH CONSEQUENCES TIMN 438429 17

The editors also arknowled~qe the cOrltrlhuflorls of the,fulloai•in" .stuJf'nrenrhrrs curcl others who assisted in the preparation of'tlris Repcrrt: Carmen Aguirre, Secretary. Office on Smoking and Health. CCDPHP. CDC. Rockvil le. Maryland Andrea Anderson, Student Intern. Office on Smokin- and Health, CCDPHP. CDC. Rockville, Maryland Margaret Anglin. Secretary, Office on Smoking and Health. CCDPHP. CDC. Rock- viile. Maryland Cathy Arney, Graphic Artist. The Circle. Inc.. McLean, Virginia John Artis, Courier. The Circle. Inc.. McLean. Virginia Michele Asrael. Conference Coordinator, The Circle. Inc.. McLean. Vir-uinia •John L. Bagrosky, Associate Director for Program Operations. Office on Smoking anci Health, CCDPI-IP, CDC, Rockville. Maryland Sonia Balakirsky. Secretary. Office on Smoking and Health. CCDPHP. CDC. Rock- ville, Maryland Barbara Barnes. Administrative Assistant. The Circle. Inc.. McLean. Vir,,inia Carol A. Bean. Ph.D.. Actin~_ Manazin~_ Editor. Artemis Technolo`_ies. Inc.. Springfield. Virginia Marissa A. Bernstein. Editor. The Circle. Inc.. McLean. Viminia Em' Ria Briscoe, Conference Coordinator. The Circle. Inc.. McLean. Vir-6nia Karen Broder. Public Information Specialist.Oftice on Smokin-and Health. CCDPHP. CDC, Rockville. Maryland Barbara M. Brown. Editorial Assistant. Office on Smokim_ and Health. Rockville. Maryland Catherine E. Burckhardt. Public Information Specialist. Office on Smoking and Health. CCDPHP, CDC. Rockville, Maryland Lee Chapell, Courier. The Circle, Inc.. McLean. Virginia Won Choi. Research Assistant. Office on Smokin,_, and Health. CCDPHP. CDC. Rockville, Maryland ~ Trish Davidson. Student Intern. Office on Smokino and Health. CCDPHP. CDC, Rockville, Maryland Susan E. Day. Secretary. Office on Smoking and Health. CCDPHP. CDC. Rockville. Maryland ~ Karen M. Deasy, Associate Director for Policy, Office on Smoking and Health. CCDPHP, CDC, Rockville. Maryland ~ June Dow, Public Health Service Congressional Reports Coordinator. Office of Health Plannin- and Evaluation. Office of the Assistant Secretarv for Health. Washin~_ton. D.C. ~ Joanna Ebling. Word Processing Specialist. The Circle. Inc.. McLean. Virginia Pam Edwards. System Administrator. MSA. Inc.. Rockville. Maryland Rita Elliott. Technical Editor, New Mexico Tumor Re:ititrv. Universitv of New Mexico. Albuquerque. New Mexico ~ . Seth Emont, Ph.D., Epidemic Intelligence Service Officer. Office on Smoking and Health. CCDPHP, CDC. Rockv.ille. Maryland xx TIMN 438403

Less risky life of terminators Relapse %NTER HERE DCIT 3 ~ ~--~ Avoid further failure Riskier life of precontemplators FIGURE 1.-Cyciical model of the stages of change SOURCE: Prochaska et ai. (in press). On any single cessation attempt (action stage), the majority of smokers relapse and return to regular smoking. A National Heart. Luna, and Blood Institute consensus conference defined relapse as at least one puff per day for 7 days and recommended that this definition be applied uniformly (Shumaker and Grunberg 1986): however, this definition is not used in all studies. Any return to smoking that is less than the criterion for relapse is considered a "lapse" or a"slip," which may or may not cause a return to regular smoking (Brownell et al. 1986; Marlatt. Curry. Gordon. 1988). Although 75 to 80 percent of relapse occurs at 6 months and before (Hunt. Barnett, Branch 1971: Hunt and Bespalec 1973: Hughes et al. 1981: Garvey. Heinold, Rosner 1989), individuals who maintain abstinence for 6 months continue to relapse by 11 months and beyond. For example, in a review of (0 studies in which minimal or no intervention occurred (i.e., self-change studies ), relapse rates at 12 months for smokers who had previously maintained abstinence for at least 6 months ranged from 7 to 35 percent (Cohen et al. 1989). Data from the National Health and Nutrition Examination 23 TIMN 438433

2. The excess risk of CHD caused by smoking is reduced by about half after 1 year of smoking abstinence and then declines gradually. After 15 years of abstinence, the risk of CHD is similar to that of persons who have never smoked. 3. Among persons with diagnosed CHD, smoking cessation markedly reduces the risk of recurrent infarction and cardiovascular death. In many studies, this reduction in risk of recurrence or premature death has been 50 percent or more. 4. Smoking cessation substantially reduces the risk of peripheral artery occlusive disease compared with continued smoking. 5. Among patients with peripheral artery disease, smoking cessation improves exercise tolerance, reduces the risk of amputation after peripheral artery surgery, and increases overall survival. 6. Smoking cessation reduces the risk of both ischemic stroke and subarachnoid hemorrhage compared with continued smoking. After smoking cessation, the risk of stroke returns to the level of never smokers; in some studies this has occurred within 5 years, but in others as long as 15 years of abstinence were required. Chapter 7: Smoking Cessation and Nonmalignant Respiratory Diseases 1. Smoking cessation reduces rates of respiratory symptoms such as cough, sputum production, and wheezing, and respiratory infections such as bronchitis and pneumonia, compared with continued smoking.. 2. For persons without overt chronic obstructive pulmonary disease (COPD), smoking cessation improves pulmonary function about 5 percent within a few months after cessation. 3. Cigarette smoking accelerates the age-related decline in lung function that occurs among never smokers. With sustained abstinence from smoking, the rate of decline in pulmonary function among former smokers returns to that of never smokers. 4. With sustained abstinence, the COPD mortality rates among former smokers decline in comparison with continuing smokers. Chapter 8: Smoking Cessation and Reproduction 1. Women who stop smoking before becoming pregnant have infants of the same birthweight as those born to never smokers. 2. Pregnant smokers.who stop smoking at any time up to the 30th week of gestation have infants with higher birthweight than do women who smoke throughout pregnancy. Quitting in the first 3' to 4 months of pregnancy and abstaining 'I'IAIN 438424 11

throughout the remainder of pregnancy protect the fetus from the adverse effects of smoking on birthweight. 3. Evidence from two intervention trials suggests that reducing daily cigarette con- sumption without quitting has little or no benefit for birthweight. 4. Recent estimates of the prevalence of smoking during pregnancy, combined with an estimate of the relative risk of low birthweight outcome in smokers, suggest that 17 to 26 percent of low birthweight births could be prevented by eliminating smoking during pregnancy; in groups with a high prevalence of smoking (e.g., women with less than a high school education), 29 to 42 percent of low birthweight births might be prevented by elimination of cigarette smoking during pregnancy. 5. Approximately 30 percent of women who are cigarette smokers quit after recogni- tion of pregnancy, with greater proportions quitting among married women and especially among women with higher levels of educational attainment. 6. Smoking causes women to have natural menopause I to 2 years early. Former smokers have an age at natuial menopause similar to that of never smokers. Chapter 9: Smoking, Smoking Cessation, and Other Nonmalignant Diseases 1. Smokers have an increased risk of development of both duodenal and gastric ulcer, and this increased risk is reduced by smoking cessation. 2. Ulcer disease is more severe among smokers than amons nonsmokers. Smokers are less likely to experience healing of duodenal ulcers and are more likely to have recurrences of both duodenal and gastric ulcers within specified timeframes. Most ulcer medications fail to alter these tendencies. 3. Smokers with gastric or duodenal ulcers who stop smoking improve their clinical course relative to smokers who continue to smoke. 4. The evidence that smoking increases the risk of osteoporotic fractures or decreases bone mass is inconclusive, with many conflicting findings. Data on smoking cessation are extremely limited at present. 5. There is evidence that smoking is associated with prominent facial skin wrinkling in whites, particularly in the periorbital ("crow's foot") and perioral areas of the face. The effect of cessation on skin wrinkling is unstudied. Chapter 10: Smoking Cessation and Body Weight Change 1. Average weight gain after smoking cessation is only about 5 pounds (2.3 kg). This weight gain poses a minimal health risk. 12 TIMN 438425

Benefits for Infants and Children As a pediatrician, I am particularly concerned about the effects of parental smoking on infants and children. Evidence reviewed in the 1986 Surgeon General's Report, The Health Consequences of Involuntary Smoking, indicates that the children of parents who smoke, compared with the children of nonsmoking parents, have an increased frequency of respiratory infections such as pneumonia and bronchitis. Many studies have found a dose-response relationship between respiratory illness in children and their level of tobacco smoke exposure. Several studies have shown that children exposed to tobacco smoke in the home are more likely to develop acute otitis media and persistent middle ear effusions. Middle ear disease imposes a substantial burden on the health care system. Otitis media is the most frequent diagnosis made by physicians who care for children. The myringotomy- and-tube procedure, used to treat otitis media in more than I million American children each year, is the most common minor surgical operation performed under general anesthesia. The impact of smoking cessation during or after pregnancy on these associations has no[been studied. However, the dose-response relationship between parental smoking and frequency of childhood respiratory infections suggests that smoking cessation during pregnancy and abstinence after delivery would eliminate most or all of the excess risk by eliminating most or all of the exposure. If parents are unwilling to quit smoking for their own sake, I would urge them to quit for the sake of their children. Passive-smoking-induced infections in infants and young children can cause serious and even fatal illness. Moreover, children whose parents smoke are much more likely to become smokers themselves. . Smoking Cessation and Weight Gain The fear of postcessation weiQht gain may discourage many smokers from trying to quit. The fear or occurrence of weight gain may precipitate relapse among many of those who already have quit. In the 1986 Adult Use of Tobacco Survey, current smokers who had tried to quit were asked to judge the importance of several possible reasons for their return to smoking. Twenty-seven percent reported that "actual weight gain" was a "very important" or "somewhat important" reason why they resumed smoking; 22 percent said that "the possibility of gaining weight" was an important reason for their relapse. Forty-seven percent of current smokers and 48 percent of former smokers agreed with the statement that "smoking helps control weight." Fifteen studies involving a total of 20,000 persons were reviewed in this Report to determine the likelihood of gaining weight and the averaae weiQht gain after quitting. Although four-fifths of smokers who quit gained weight after cessation, the average weight gain was only 5 pounds (2.3 kg). The average weight gain among subjects who continued to smoke was 1 pound. Thus, smoking cessation produces a 4-pound greater weight gain than that associated with continued smoking. This weight gain poses a minimal health risk. Moreover, evidence suggests that this small weight gain is accompanied by favorable changes in lipid profiles and in body fat distribution. ix TIMN 438392

Lester Breslow, M.D., M.P.H., Professorof Public Health and Director, Health Services Research, Division of Cancer Control, Jonsson Comprehensive Cancer Center, University of California, Los Angeles, Los Angeles, California Samuel Broder, M.D., Director, National Cancer Institute, National Institutes of Health, Bethesda, Maryland David Bums, M.D., Associate Professor, Pulmonary Division, Division of Pulmonary Medicine and Critical Care, University of California at San Diego Medical Center, San Diego, California Benjamin Burrows, M.D., Director, Division of Respiratory Sciences, University of Arizona Health Sciences Center, University of Arizona School of Medicine, Tucson, Arizona Jane Cauley, Dr.P.H., Assistant Professor of Epidemiology, Department of Epidemiol- ogy, University of Pittsburgh, Pittsburgh, Pennsylvania Gregory N. Connolly, D.M.D.,. M.P.H., Director, Office on Nonsmoking and Health, Massachusetts Department of Public Health, Boston, Massachusetts Thomas M. Cooper, D.D.S., Professor, University of Kentucky Medical Center, Col- lege of Dentistry, Lexington, Kentucky Stephen Corbin, D.D.S., M.P.H., Policy Analyst, Disease Prevention, Center for Preventive Services (CPS). CDC. Bethesda. Maryland K. Michael Cummings, Ph.D., M.P.H., Cancer Control and Epidemiology. Roswell Park Cancer Institute, Buffalo, New York Joseph W. Cullen, Ph.D., Director. AMC Cancer Research Center, Denver, Colorado Sir Richard Doll, ICRF Cancer Studies Unit, Oxford, United Kingdom Virginia Ernster, Ph.D., Professor of Epidemiology, Department of Epidemiology and ' International Health, University of California, San Francisco, San Francisco, California Jonathan E. Fielding, M.D., M.P.H., Vice President and Health Director, Johnson and Johnson Health Management, Inc., Santa Monica, California Gary D. Friedman, M.D., M.S., Division of Research, Kaiser Permanente Medical Care Program, Northern California Reaion, Oakland, California William Foege, M.D., Executive Director, The Carter Center of Emory University, Atlanta, Georgia Lawrence J. Furman, D.D.S., M.P.H.. Chief, Dental Disease Prevention Activity, CPS, CDC, Atlanta, Georgia Lawrence Garfinkel, Vice President for Epidemiology and Statistics, Director of Cancer Prevention, American Cancer Society, Inc., New York, New York - Barbara A. Gilchrest, M.D., Professor and Chairman, Department of Dermatology, Boston University Medical Center, Boston, Massachusetts Frederick K. Goodwin, M.D., Administrator, Alcohol, Drug Abuse, and Mental Health Administration, Rockville, Maryland Robert O. Greer, Jr., D.D.S., Sc.D., Professor and Chairman, Division of Oral Pathology and Oncology, Department of Diagnostic and Biological Sciences, School of Den- tistry, University of Colorado Health Sciences Center, Boulder, Colorado Ellen Gritz, Ph.D., Director, Division of Cancer Control, Jonsson Comprehensive Cancer Center, University of California, Los Angeles, Los Angeles, California xvi TIMN 438399

Nancy J. Haley, Ph.D.. Associate Chief, Division of Nutrition and Endocrinology, American Health Foundation, Valhalla, New York Sharon M. Hall, Ph.D., Professor of Medical Psychology, Department of Psychiatry, University of California, San _ Francisco, San Francisco Veterans Administration Medical Center. San Francisco, California Robert Harmon, M.D., Administrator, Health Resources and Services Administration, Rockville, Maryland Jeffrey E. Harris, M.D., Ph.D., Associate Professor, Department of Economics, Mas- sachusetts Institute of Technology, Cambridge, Massachusetts, Clinical Associate, Medical Services, Massachusetts General Hospital, Boston, Massachusetts Norman O. Harris, D.D.S., M.S., University of Texas Health Science Center, San Antonio, Texas Jack Henningfield, Ph.D., Chief, Clinical Pharmacology Branch, National Institute on Drug Abuse Addiction Research Center, National Institutes of Health, Baltimore, Maryland Robert A. Hiatt, M.D., Ph.D., Senior Epidemiologist, Division of Research, Kaiser Permanente Medical Care Program. Oakland California Millicent Higgins, M.D., Associate Director, Epidemiology and Biometry ProQram, Division of Epidemiology and Clinical Applications, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland Carol Hogue. Ph.D., M.P.H., Director, Division of Reproductive Health, CCDPHP, CDC, Atlanta. Georgia John Holbrook. M.D., Professor of Internal Medicine, Department of Internal Medicine, University of Utah School of Medicine, Salt Lake City, Utah Richard Hunt, M.D., Division of Gastroenterology, McMaster University Medical Center, Hamilton, Ontario, Canada Dwight Janerich, D.D.S., M.P.H., Professor of Epidemiology, Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut William Kannel, M.D., Professor of Medicine, Department of Preventive Medicine, Boston University School of Medicine, Boston, Massachusetts LTC James W. Kikendall, M.D., Assistant Chief, Gastroenterology Section, Walter Reed Army Medical Center, Washington, D.C. Dushanka V. Kleinman, D.D.S., M.Sc.D., Section Chief, National Institute on Dental Research. National Institutes of Health, Bethesda. Mary. land C. Everett Koop, M.D., Sc.D., U.S. Surgeon General, 1981-89, Bethesda, Maryland Jeffrey P. Koplan, M.D., M.P.H., Director, CCDPHP, CDC, Atlanta, Georgia Lewis H. Kuller, M.D., Dr.P.H., Professor and Chairperson, Department of Epidemiol- ogy, University of Pittsburgh Graduate School of Public Health, Pittsburgh, Pennsyl- vania Charles L. LeMaistre, M.D., President, The University of Texas M.D. Anderson Cancer Center, Houston, Texas Claude Lenfant, M.D., Director, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland Richard J. Levine, M.D., M.S., M.P.H., Chief Epidemiologist, Chemical Industry Institute of Toxicology, Research Triangle Park, North Carolina TIMN 438400 x""

Table 17. Estimated relative risk of azoospermia or oligospermia among smokers versus nonsmokers or never smokers ......................... 408 Table 18. Sperm quality among former smokers ........................ 409 Chapter 9 Table 1. Percentage of healed duodenal ulcers among smoking and nonsmoking patients .............................................. 434 Table 2. Results of statistical analysis of pooled data from Table 1......... 437 Table 3. Recurrences of duodenal ulcer in smokers and nonsmokers in clinical trials ................................................... 438 Table 4. Recurrences of gastric ulcer in smokers and nonsmokers in clinical trials ................................................... 442 Table 5. Summary of studies of smoking and bone mass .................. 445 Table 6. Summary of case-control studies of smoking and fractures ........ 450 Table 7. Summary of cohort studies of smoking and fractures ............. 454 Chapter 10 Table 1. Summary of prospective studies on smoking and body weight ...... 474 Table 2. Details of prospective studies in which change in weight relative to continuing smokers was reported .......................... 477 Table 3. Mortality ratios for all ages combined in relation to the death rate of those 90-109% of average weight ............................ 493 Table 4. Mortality ratios for all ages combined according to smoking status in relation to those 90-109% of average age ..................... 494 Chapter 11 Table 1. Diagnostic categorization and criteria for nicotine withdrawal-Nicotine-induced organic mental disorder ................. 522 Table 2. Prospective studies of quitting-related changes in mood, anxiety, stress reactivity, perceived stress, self-image, and psychological well-being ......................................... 536 Table 3. Summary of data from 1985 NHIS, behaviors of never, former, and current smokers aged 20 and older .............................. 548 Table 4. Summary of data from 1987 NHIS, behaviors of never, former, and current smokers aged 18 and older .............................. 549 Table 5. Summary of data from 1987 BRFSS, behaviors of former smokers and current smokers aged 18 and older ....................... 550 xxix TIMN 438411

Edward Lichtenstein, Ph.D., Research Scientist, Oregon Research Institute, Eugene, Oregon Jay H. Lubin, Ph.D., National Cancer Institute, National Institutes of Health, Rockville, Maryland Alfred C. Marcus, Ph.D., Director, Community Research and Applications, AMC Cancer Research Center, Denver, Colorado Denis M. McCarthy, M.D., M.Sc., Chief, Division of Gastroenterology, University of New Mexico, Department of Medicine, Veterans Administration Medical Center, Albuquerque, New Mexico J. Michael McGinnis, M.D., Deputy Assistant Secretary for Health, Disease Prevention and Health Promotion. Department of Health Human Services, Washington, D.C. Sonja M. McKinlay, Ph.D., M.Sc., M.A.. B.A., ASA, APHA. AER, SCt, Biometrics Society, Institute of Mathematical Statistics, International Menopause Society, American Association for the Advancement of Science, President, New England Research Institute, Inc., Watertown, Massachusetts Robert E. Mecklenberg, D.D.S., M.P.H.. Potomac, Maryland L. Joseph Melton, III, M.D., Head, Section of Clinical Epidemiology, Department of Health Sciences Research. Mayo Clinic and Foundation, Rochester, Minnesota Anthony Miller, B.A., M.B.B., M.R.C.P., M.F.C.M., F.R.C.P.C., Professor. Depart- ment of Preventive Medicine and Biostatistics, University of Toronto. Toronto, Ontario, Canada Gregory Morosco. Ph.D., M.P.H.. Chief, Health Education Branch and Coordinator. SmokinQ Education Program, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland Richard L. Naeye. M.D.. Professor and Chairman, Department of Pathology. Pennsyl- vania State University School of Medicine, Hershey, Pennsylvania ~ Thomas A. Pearson, M.D., M.P.H., Ph.D., Director, Mary Imogene Bassett Research Institute, Cooperstown, New York, Professorof Public Health in Medicine, Columbia University, New York, New York Terry Pechacek, Ph.D., Acting Chief, Smoking, Tobacco, and Cancer Branch. National Cancer Institute. National Institutes of Health, Bethesda. Maryland Michael G. Perri. Ph.D., Professor and Deputy Chairman, Psychology Department. Fairleigh Dickinson University, Teaneck, New Jersey Richard Peto, FRS. ICRF Cancer Studies Unit, Oxford, United Kingdom John M. Pinney, Executive Director, Institute for the Study of Smoking Behavior and Policy, John F. Kennedy School of Government, Harvard University, Cambridse, Massachusetts v William F. Raub, Ph.D., Acting Director, National Institutes of Health, Bethesda, Maryland Patrick L. Remington, M.D.. Bureau of Community Health Prevention, Wisconsin Division of Health, Madison, Wisconsin Everett R. Rhoades, M.D., Director, Indian Health Service, Rockville, Maryland Julius Richmond. M.D., John D. MacArthur Professor of Health Policy, Emeritus, Division of Health Policy, Research, and Education, Harvard University, Boston, Massachusetts xV"' TIMN 438401

Chapter 6 Table 1. Case-control studies of CHD risk among former smokers ......... 201 Table 2. Cohort studies of CHD risk among former smokers .............. 206 Table 3. Estimated probability of dying from ischemic heart disease in the next 16.5-year interval (95% CI) for quitting at various ages compared with never smoking and continuing to smoke, by amount smoked and sex ................................................. 216 Table 4. Intervention trials of smoking cessation and CHD risk ............ 225 Table 5. Studies of the effect of smoking cessation on persons with diagnosed CHD ................................................. 231 Table 6. Studies of smoking cessation and risk of death due to aortic aneurysm ...................................................... 242 Table 7. Case-control studies of smoking cessation and risk of stroke ....... 247 Table 8. Prospective cohort studies of smoking cessation and risk of stroke .......................................................253 Chapter 7 Table 1. Percentages of subjects in cross-sectional studies with respiratory symptoms, by cigarette smoking status and gender ............ 289 Table 2. Percentages of subjects in cross-sectional surveys with respiratory symptoms by smoking and occupational exposure status ....... 297 Table 3. Change (%) in presence of respiratory symptoms, longitudinal studies, by cigar•ette smoking status ................................. 300 Table 4. Percentage of subjects with respiratory symptoms by smoking status, 1961 and 1971, in a cohort of middle-aged, rural Finns ............ 305 Table 5. Age-standardized mortality ratios for influenza and pneumonia for current and former smokers compared with never smokers ............ 309 Table 6. Association between cigarette smoking status and FEV I levels in selected cross-sectional studies of adult populations .................. 311 Table 7. Spirometric studies of participants in smoking cessation programs ...................................................... 320 Table 8. Studies of closing volume (CV/VC%), closing capacity (CC/I'LC%), and slope of alveolar plateau (SBN2/L) among participants in smoking cessation programs ........................... 324 Table 9. Population-based longitudinal studies of annual decline in pulmonary function .............................................. 330 Table 10. Decline of FEV I (mL/yr) in subjects in the Copenhagen City Heart Study ....................................................337 Table 11. Mortality attributable to COPD, United States, 1986 ............. 342 ,~IM~Y 438409 xxvii .~ 1.d

Barbara Rimer, Dr.P.H., Director, Behavioral Research, Fox Chase Cancer Center, Philadelphia, Pennsylvania Mary Ann Salmon, Ph.D., Research Specialist, School of Social Work, C.A.R.E.S., University of North Carolina, Chapel Hill, North Carolina Jonathan M. Samet, M.D. (Senior Scientific Editor), Professor of Medicine and Chief, Pulmonary Division, Department of Medicine and the New Mexico Tumor Registry, Cancer Center, University of New Mexico, Albuquerque, New Mexico David Savitz, Ph.D., Associate Professor, Department of Epidemiology, School of Public Health, University of North Carolina, Chapel Hill, North Carolina Charles B. Sherman, M.D., Director, Pulmonary Division, Miriam Hospital, Providence, Rhode Island Meir Stampfer, M.D., Dr.P.H., Associate Professor of Epidemioloay, Harvard School of Public Health, and The Channing Laboratory, Department of Medicine, Harvard Medical School and Brigham and Women's Hospital, Boston, Massachusetts Wayne F. Velicier, Ph.D., Professor, Co-Director, Cancer Prevention Research Unit, Department of Psychology, University of Rhode Island, Kingston, Rhode Island Thomas Vogt, Ph.D., Principle InvestiQator, Center for Health Research. Portland, Oregon ~ Scott T. Weiss, M.D., Associate Professor, Harvard School of Public Health, and The Channing Laboratory, Department of Medicine, Harvard Medical School and Brigham and Women's Hospital, Boston, Massachusetts Anna H. Wu-Williams, Ph.D., Associate Professor, Department of Preventive Medicine, University of Southern California, Los Angeles, California -The editors acknowledge with gratitude the following distinguished scientists. physicians, and others who lent their support in the development of this Report by coordinating manuscript preparation, contributing critical rel'iews, or assistin; in other ways. In particular, the editors e_1Press appreciation to the American Cancer Soc•iety for making available data from its Cancer Prevention Study 11. David B. Abrams, Ph.D., Director, Division of Behavioral Medicine. The Miriam Hospital, Associate Professor, Psychiatry and Human Behavior, Brown University Program in Medicine, Providence, Rhode Island Duane Alexander, M.D., Director, National Institute of Child Health and Human Development, National Institutes of Health. Bethesda. Maryland bavid Bates, M.D., FRCP. FRCPC, FACP, FRSC, Professor Emeritus of Medicine, Department of Health Care, University of British Columbia, Vancouver, British Columbia James S. Benson, Acting Commissioner, Food and Drug Administration, Rockville, Maryland Trudy S. Berkowitz, Ph.D., Associate Professor, Department of Obstetrics, Gynecol- ogy, and Reproductive Science, Mount Sinai School of Medicine, New York, New York Ruth Bonita, M.P.H., Ph.D., Masonic Senior Research Fellow, Geriatric Unit, Univer- sity of Auckland, Auckland 9, New Zealand xv TIMN 438398

Survey I(NHANES-I) Epidemiologic Followup Study demonstrate that even after 1 year of prolonged abstinence, relapse continues to occur in about one-third of former smokers. Relapse continues to occur at a much lower rate after 2 years (Volume Appendix). In the Multiple Risk Factor Intervention Trial (MRFIT). a multifactor intensive intervention study, Ockene and colleagues (1982) found that among smokers who had stopped with the aid of intensive intervention, relapse continued to occur throughout the 6 years of followup. However, relapse was at a much higher rate in the first year than in years two through six. Kirscht and colleagues (1987) reported that 9.5 percent of adults who had been abstinent for 24 to 119 monthsreported smoking again in a followup survey. Even after 120 months, 2.3 percent of former smokers reported smoking again. Research would be simplified if the probability of remaining a former smoker were 100 percent after a prolonged period of abstinence. If this were the case, then there would be no concern about future misclassification of these confirmed former smokers. However, the continuous nature of the relapse process and the curves that represent this process indicate that the probability of maintained cessation will never be 100 percent. The available data (Garvey, Heinold, Rosner 1989; Ockene et al. 1982; Cohen et al. 1989; Volume Appendix) suggest that for most research purposes. 24 months of continuous abstinence can be used as a practical criterion for categorizing individuals as confirmed former smokers. However, use of this timeframe is often not feasible or applicable in many research studies, and as a general guideline for interpreting out- comes-the longer the duration of continuous abstinence, the ;reater the probability that individuals will remain former smokers. ~ Cessation is a cyclical, not linear, process; smokers can enter or leave the process at any point (Prochaska and DiClemente 1983; Prochaska et al., in press) (Figure 1). Research on self-change approaches to smoking cessation suQgeSts that the average smoker cycles three to four times through the stages before attaining lonQ-term continuous abstinence and becoming a confirmed former smoker (Prochaska and DiClemente 1984, 1986; Marlatt, Curry, Gordon 1988; Schachter 1982). In a review of self-change studies, Cohen and colleagues (1989) found that only 4.3 percent of the participants in the reviewed studies shifted immediately from current smokers to former smokers without experiencing any lapses or relapses. Most smokers who relapse return to a point where they think about stopping again, that is, the contemplation stage. A smaller proportion lose their motivation to change and regress back to the pre- contemplation stage (Prochaska and DiClemente 1984). In summary, because of the dynamic nature of change in smoking behavior, any categorization of smoking status at a single point in time becomes a simplification. A group of former smokers will include individuals who have stopped recently or who have been abstinent for varying lengths of time; some will maintain abstinence, and some will relapse. Knowledge of the dynamics of smoking cessation and its usual time course can help investigators minimize misclassification by choosing the most ap- propriate methods for assessing smoking behavior and the appropriate sampling pro- cedures (e.g., number of measurements made and time between repeated measures of smoking status). 24 TIlVIN 438434

Behavioral Measures Self-Report: Questionnaires and Interviews For health research purposes, smoking status is usually assessed by using self- administered questionnaires or interviews. However, other behavioral methods, sur- rogate assessments, and nonbehavioral methods such as biochemical assessments are also used as sources of smoking data. These other sources will be reviewed in subsequent sections. (See also reviews by Pechacek, Fox et al. 1984 and Marsh et al. 1988.) Questionnaires and interviews may include information concerning smoking at the time of the assessment or conceming a complete orpartial retrospective lifetime history. Assessment can be made once or serially over time, thus providing more valid data regarding cessation and possible relapse. Information gathered from an interview or questionnaire about smoking categorizes respondents as never, current, or former smokers. Two standard items used in the National Health Interview Survey (Volume Appendix) to classify smoking status are "Have you smoked at least 100 cigarettes in your entire life?" and "Do you smoke cigarettes now?" Someone responding "yes" to the first question and "no" to the second would be classified as a former smoker. Such a broad definition for former smokers combines persons who experimented with smoking enough to have smoked 100 cigarettes with individuals who may have smoked during their entire adult life and quit in the week prior to being interviewed. The commonly used item, "Have you smoked at least 100 cigarettes in your entire life?" has an advantage of countina as never smokers those individuals who experi= mented with 1, 2, or quite a few cigarettes. Only those who have 'smoked at least 5 packs of cigarettes in their lifetime are counted as ever smokers. The arbitrariness of this definition reflects the lack of accepted and standardized definitions for.ever smokers and never smokers: A definition of never smokers that requires only minimal or no use of tobacco may result in many individuals with extremely low exposure to cigarettes being classified as former smokers, which in general would not be bioloQi- ~ ~ ~ cally appropriate. Another commonly used type of item, as in the Medical Research Council (MRC) National Survey of Health and Development (Britten 1988), for defining ever smokers is "Have you ever smoked as much as I cigarette a day for as long as~l year?" This item is used by the American Thoracic Society, Division of Lung Disease in its Adult Respiratory questionnaire; however, two other choices are added- "or 20 packs of cigarettes" or"12 ounces of tobacco" (Ferris 1978). A comparable questions is "Have you ever smoked at least 5 cigarettes per week, almost every week for at least 1 year?" (Petitti, Friedman, Kahn 198 1). These items that are used to classify ever smokers are based on a combination of the amount of cizarettes smoked (e.g.. 365) and the duration of smoking (e.g., at least 6 or 12 months). The particular question used to differentiate between ever smokers and never smokers can directly affect categorization of individuals. For example, Petitti, Friedman, and Kahn (1981) found that with a more specifically defined question such as "Have you ever smoked at least 5 cigarettes per week almost every week for at least 1 year?" which TIMN 438435 25

smoked was twice as common for 20 years earlier (17 vs. 9 percent) and six times more common for 32 years previously (37 vs. 6 percent). Persson and Norell (1989) found that in a random sample of 9,394 individuals in Sweden, retrospective information obtained 6 years later resulted in a strong tendency to overestimate previous cigarette consumption among individuals who had increased their smoking (69 percent over- estimated) and to underestimate among individuals who had decreased their smoking (49 percent underestimated). Subjects with unchanged cigarette consumption showed the highest levels of agreement (89 percent) between original and retrospective infor- mation. Rather than reconstructing full smoking cessation histories that are subject to biased reporting, many retrospective studies rely on more limited categorization such as never, former, and current smokers. Retrospective studies enable researchers to assess long periods of smoking abstinence without the need to observe the subjects over a long period of time, as would be necessary in prospective studies. Case-control studies, for example, can compare cases with smoking-related diseases with controls with histories of being abstinent for 10 to 20 years; in a prospective study, it may be impractical or impossible to study health consequences of cessation with more than 10 to 20 years of abstinence (Chapter 2, Part II). Prospective studies have the potential for more reliable and valid measures of smoking status over time, especially when using a series of assessments, than do retrospective studies. In intervention trials, for example, all subjects enter the trial as current smokers. Followina intensive intervention, subjects are identified as continuing smokers or former smokers (abstinent). By assessing subjects at specified intervals such as every 4 or 6 months over a series of years, especially when paired with biochemical verification (Chapter 2, see section on Biochemical Markers), researchers can reduce the measurement bias and be more confident in the reliability and validity of measures classifying continuing and former smokers and specifying length of abstinence for former smokers. In MRFIT (Ockene et al. 1990) for example, a series of 4-month followups over 6 years enabled researchers to classify participants into three categories: persistent quitters (continuous abstainers since the initial intervention), intermittent quitters (abstinent for periods of time since the initial intervention), and continuous smokers (not abstinent during any of the followup periods). Such precision in measurement is generally not possible or necessary in epidemiologic studies. Prospective studies may use a single assessment to categorize current, former, and never smokers. These studies then prospectively examine the categories to detect differential rates of morbidity and mortality. As discussed above, the assumption that individuals will not change their smoking status maybe a flaw with such single assessments. Improving Self-Report Measures Ideally, assessments of smoking status need to include standardized questions to determine smoking status, that is never, current, and former smokers. For example, to be categorized as a never smoker, the necessary response would be "no" to a standard question such as, "Have you ever smoked at least I cigarette per day for at least 1 year?" 28 TIMN 438438

Chapter 4: Smoking Cessation and Respiratory Cancers 1. Smoking cessation reduces the risk of lung cancer compared with continued smok- ing. For example, after 10 years of abstinence, the risk of lung cancer is about 30 to 50 percent of the risk in continuing smokers; with further abstinence, the risk continues to decline. 2. The reduced risk of lung cancer among former smokers is observed in males and females, in smokers of filter and nonfilter cigarettes, and for all histologic types of lung cancer. 3. Smoking cessation lowers the risk of laryngeal cancer compared with continued smoking. 4. Smoking cessation reduces the severity and extent of premalignant histologic changes in the epithelium of the larynx and lung. Chapter 5: Smoking Cessation and Nonrespiratory Cancers 1. Smoking cessation halves the risks for cancers of the oral cavity and esophagus, compared with continued smoking, as soon as 5 years after cessation, with further reduction over a longer period of abstinence. 2. Smoking cessation reduces the risk of pancreatic cancer, compared with continued smoking, although this reduction in risk may only be measurable after 10 years of abstinence. 3. Smoking is a cause of bladder cancer; cessation reduces risk by about 50 percent after only a few years, in comparison with continued smoking. 4. The risk of cervical cancer is substantially lower among former smokers in com- parison with continuing smokers, even in the first few years after cessation. This finding supports the hypothesis that cigarette smoking is a contributing cause of cervical cancer. y ~ 5. Neither smoking nor smoking cessation are associated with the risk of cancer of the breast. Chapter 6: Smoking Cessation and Cardiovascular Disease 1. Compared with continued smoking, smoking cessation substantially reduces risk of coronary heart disease (CHD) among men and women of all ages. 10 TIMN 438423

Renate J. Phillips. Graphic Artist. Desktop Publishing Designer, The Circle. Inc.. McLean. Virginia Margaret E. Pickerel, Public Information and Publications Specialist. Office on Smok- ing and Health. CCDPHP. CDC, Rockville. Maryland Elizabeth Precup. Student Intern. Office on Smoking and Health. CCDPHP. CDC. Rockville, Maryland Cary R. Prince, Editor, The Circle. Inc.. McLean. Virginia Dick Ray, Director of Computer Services, The Circle. Inc.. McLean, Virginia Nancy J. Rhodes, Editor, The Circle, Inc.. McLean, Virginia Rose Mary Romano. Chief. Public Information Branch. Office on Smokin- and Health. CCDPHP, CDC. Rockville. Maryland Lisa Phelps, Computer Systems Analyst. The Circle, Inc.. McLean. Virginia Sel Semler. Secretary. Office on Smoking and Health, CCDPHP. CDC. Rockville, Maryland James Sliwa. Student Intern. Office on Smoking and Health. CCDPHP. CDC, Rock- ville, Maryland Mattie Smith. Secretary. CCDPHP, CDC. Rockville. Maryland Linda R. Spiegelman. Administrative Officer. Office on Smoking and Health, CCDPHP, CDC. Rockville, Maryland Traion C. Stallings. Project Secretary. The Circle. Inc.. McLean. Virginia Sophia Stewart. Student Intern. Office on Smoking and Health. CCDPHP, CDC, Rockville, Maryland Daniel R. Tisch. Director of Publications. The Circle. Inc.. McLean. Virainia Anne Trontell, M.D., Epidemic Intelligence Service Officer. Office on Smoking and Health. CCDPHP. CDC. Rockville. Maryland Karen Tyler. Conference Coordinator. The Circle. Inc.. McLean. Virginia Godfrey R. Vaz. M.D.. Student Intern. Office on Smoking and Health. CCDPHP. CDC. Rockville, Maryland Susan Von Braunsberg. Information Specialist. The Circle. Inc.. McLean. Virginia Elyse Watson. Administrative Assistant. New Mexico Tumor Registry. University of New Mexico, Albuquerque. New Mexico Michael F. White. Associate Director for Prouram Development: CCDPHP. CDC. Rockville. Maryland Charles Wi~Qins.~M.S.P.H.. Epidemiologist. New MexicoTumor Reg-istry. University of New Mexico. Albuquerque. New Mexico Louise G. Wiseman. Technical Information Specialist. Office on Smoking and Health. CCDPHP. CDC. Rockville. Maryland Rebecca B. Wolf. Program Analyst, Office ot'Pro;,,ram Planninz, and Evaluation. CDC. Atlanta. Georgia S. Tanner Wray. Technical Intonnation Specialist. Office on Smoking and Health. CCDPHP. CDC. Rockville. Maryland xx<i 'TIMN 438405

MAJOR CONCLUSIONS More than 38 million Americans have quit smoking, and almost half of all living adults in the United States who ever smoked have quit (Volume Appendix). Neverthe- less, more than 50 million Americans continue to smoke. This Report reviews in detail the health consequences of smoking cessation for those who have quit and for those who will quit in the future. The following major volume conclusions summarize the health consequences of smoking cessation for those who quit smoking in comparison with those who continue to smoke: 1. Smoking cessation has major and immediate health benefits for men and women of all ages. Benefits apply to persons with and without smoking- related disease. 2. Former smokers live longer than continuing smokers. For example, persons who quit smoking before age 50 have one-half the risk of dying in the next 15 years compared with continuing smokers. 3. Smoking cessation decreases the risk of lung cancer, other cancers, heart attack, stroke, and chronic lung disease. 4. Women who stop smoking before pregnancy or during the first 3 to 4 months of pregnancy reduce their risk of having a low birthweight baby to that of women who never smoked. 5. The health benefits of smoking cessation far exceed any risks from the average 5-pound (2.3-kg) weight gain or any adverse psychological effects that may follow quitting. DEVELOPMENT OF THE REPORT This Report was developed by the Office on Smoking and Health (OSH), Center for Chronic Disease Prevention and Health Promotion. Centers for Disease Control, Public Health Service of the U.S. Department of Health and Human Services, as part of the Department's responsibility under Public Law 91-222 to report new and current information on smoking and health to the U.S. Congress. The scientific content of this Report was produced through the efforts of more than 120 scientists in the fields of medicine, psychology, the biologic and social sciences, and public health. Manuscripts for the Report, constituting drafts of chapters or sections of chapters, were prepared by 26 scientists selected for their expertise in specific content areas. An editorial team, including the Director of OSH, a medical psychologist with the Uniformed Services University of the Health Sciences, and four non-Federal experts, edited and consolidated the individual manuscripts into chapters. These draft chapters were subjected to an intensive outside peer review, with each chapter reviewed by an average of five individuals knowledgeable about the chapter's subject matter. Incorporating the reviewers' comments, the editors revised the chapters and assembled a draft of the complete Report. The draft Report was then submitted to 25 distinguished 8 TIMN 438421

CONTENTS Introduction ........................................................ 21 Part I. Assessing the Dynamic Process of Smoking Cessation ................ 22 The Process of Smoking Behavior Change .............................. 22 Behavioral Measures ............................................... 25 Self-Report: Questionnaires and Interviews ............................ 25 Temporal and Frequency Issues .................................. 27 Improving Self-Report Measures ................................. 28 Alternative Behavioral Measures ................................... 29 Surrogate Assessments ............................................ 30 Nonbehaq ioral Measures ............................................ 31 Physiologic Measures ............................................ 32 Biochemical Markers ............................................. 33 Terminology ..................................................33 Carbon Monoxide ............................................. 34 Thiocyanate .................................................. 35 Cotinine ..................................................... 36 Bogus Pipeline .................................................. 37 Contextual Issues Affecting Biochemical Assessment ................... 37 Part II. Assessing the Consequences of Smoking Cessation .................. 46 Study Designs Used To Assess the Consequences of Cessation .............. 46 Overvie.w of Study Design ......................................... 46 Ecologic Studies ................................................. 47 Cross-Sectional Studies ............................................ 47 Cohort Studies .................................................. 48 Case-Control Studies ............................................. 49 Intervention Trials ............................................... 50 Methodologic Issues ................................................ 51 Introduction .................................................... 51 Statistical Considerations .......................................... 52 Bias ........................................................... 52 Analytic Issues in Observation Studies ............................... 55 Summary .......................................................... 57 Conclusions ........................................................ 58 References ................................. :....................... 59 TIMN 438430 19

Biochemical Markers Cigarette smoke is a complex mixture of chemicals, some of which are present in the tobacco leaf and some of which result from chemical reactions during either the curing process or smoking (US DHEW 1979; US DHHS 1986, 1989). Three chemical constituents of tobacco smoke, carbon monoxide (CO), hydrogen cyanide (HCN), and nicotine, pass through cigarette filters and are present in inhaled tobacco smoke in concentrations high enough to be absorbed and detected in persons who smoke. These chemicals are measurable as intact compounds or as metabolic products. Exposure to CO can be assessed in the blood as carboxyhemoglobin (COHb) or as CO in expired alveolar air. Methods are available for measuring cotinine, the primary metabolite of nicotine, and SCN-, a metabolite of HCN, in urine, blood, and saliva. Other measures, such as skin-surface sampling for nicotine (Nanji and Lawrence 1988) are not as well established. Extensive reviews of the literature on the use of biochemical markers as measures of smoking status are provided by Benowitz (1983), Haley and colleagues (1986), Lee (1988), Pechacek, Fox, and colleagues (1984), -and Windsor and Orleans (1986). Cummings and Richard (1988) supplied a review of optimal cutoffs for the biochemical measures discussed here. This Section is not intended to provide an indepth review of the variability and biochemical rationale for these measures and will only provide an overview of the use of biochemical assessments for smoking status. Terminology Sensitivity and specificity, characteristics of a test such as a biochemical assessment, are measures of validity, the extent to which the test measures truth (Fletcher, Fletcher, Wagner 1987). Typically, sensitivity and specificity are determined by comparing the test results against a reference or "gold" standard. For smoking, self-reported status . has most often been used as the standard for assessing biochemical markers. The sensitivity of a biochemical test for smoking exposure is the proportion of true smokers who are classified as smokers by the biochemical test. The specificity of a biochemical test for smoking exposure is the proportion of true nonsmokers who are classified as nonsmokers by the biochemical test. A test of 100-percent sensitivity and 100-percent specificity would perfectly discriminate true smokers from true nonsmokers. However, this degree Qf validity is not reached by any presently available biochemical marker. In addition, the standard to which biochemical measures are compared, typically self-reported smoking status, may be of limited validity, and thereby cause apparent sensitivity and specificity to be reduced. When continuous measures are used to test for smoking status, a cutpoint must be chosen such that those individuals whose test value exceeds the cutpoint are classified as smokers and those with values below the cutpoint are classified as nonsmokers (Cummings and Richard 1988). The level at which the cutpoint is set determines the sensitivity and specificity of the test. Lowering the curpoint improves the sensitivity at the expense of specificity. Raising it will improve specificity at the expense of sensitivity (Cole and Morrison 1980; Browner, Newman, Cummings 1988). Selecting TIMN 438443 33

2. Approximately 80 percent of smokers who quit gain weight after cessation, but only about 3.5 percent of those who quit smoking gain more than 20 pounds. 3. Increases in food intake and decreases in resting energy expenditure are largely responsible for postcessation weight gain. Chapter 11: Psychological and Behavioral Consequences and Correlates of Smoking Cessation 1. Short-term consequences of smoking cessation include anxiety, irritability, frustra- tion, anger, difficulty concentrating, increased appetite, and urges to smoke. With the possible exception of urges to smoke and increased appetite, these effects soon disappear. 2. Smokers who abstain from smoking show short-term impairment of performance on a variety of simple attention tasks, which improves with nicotine administration. Memory, learning, and the performance of more complex tasks have not been clearly shown to be impaired. Whether the self-reported improvement in attention tasks upon nicotine administration is due entirely to relief of withdrawal effects or is also due in part to enhancement of performance above the norm is unclear. 3. In comparison with current smokers, former smokers have a greater perceived ability to achieve and maintain smoking abstinence (self-efficacy) and a greater perceived control over personal circumstances (locus of contrdl). 4. Former smokers, compared with current smokers, practice more health-promoting and disease-preventing behaviors. Volume Appendix: National Trends in Smoking Cessation 1. By 1987, more than 38 million Americans had quit smoking cigarettes. nearly half of all living adults who ever smoked. y 2. The percentage of ever cigarette smokers who are former cigarette smokers (quit ratio) has increased from 29.6 percent in 1965 to 44.8 percent in 1987 at an average rate of 0.68 percentage points per year. The quit ratio has increased among men and women, among blacks and whites, and among all age and education subgroups. Between 1966 and 1987, the rate of increase in the quit ratio among college graduates was twice the rate among high school dropouts. ~ 3. About one-third of all former cigarette smokers who have maintained abstinence for at least 1 year may eventually relapse. As the duration of abstinence increases, relapse becomes less likely. TIMN 438426 13

sequence of cessation followed by reduced disease occurrence; that is, disease has caused a change in exposure (Rogot and Murray 1980). Cigarette smoke in its gaseous and particulate phases contains thousands of agents, many of which can damage tissues and cause disease (US DHEW 1979; US DHHS 1986, 1989). The pathogenetic mechanisms by which cigarette smoking causes disease are diverse, ranging from longer term processes, such as carcinogenesis, to shorter term processes, such as interference with tissue oxygenation by carbon monoxide. Thus, the biologic context in which the evidence on cessation is considered must be disease- specific; a unified biologic framework for evaluating the evidence on cessation cannot be offered. For example, cigarette smoking causes emphysema, an irreversible destruction of the gas-exchanging structure of the lung, and permanent or only partially reversible damage to the airways of the lung. Little improvement of lung function after cessation would be anticipated for a long-term smoker with disabling chronic obstructive pulmonary disease (COPD) and extensive irreversible damage to the lung. However, cessation would benefit a smoker who has less extensive damage by slowing the rate of lung function decline and thereby reducing the likelihood of clinically significant impair- ment. By contrast with COPD, smoking cessation following myocardial infarction has both relatively immediate and longer term benefits. The immediately decreased risk of death in those who stop smoking in comparison with those who continue to smoke may reflect a decrease of blood coagulability, improved tissue oxygenation, and less predisposition to cardiac arrhythmias after cessation. The findings of studies on the health consequences of smoking cessation also provide evidence relevant to determining the causality of associations of active smoking with disease. A decline in disease incidence after cessation needs to be considered as a positive indication of such a causal association. However, the pattern of changing risk after cessation must be interpreted in the context of the mechanism of disease causation by active smoking. In interpreting individual studies on the consequences of smoking cessation, difficult methodologic and conceptual issues must be considered. Chapter 2 addresses these issues in depth. Because smoking cessation is a dynamic process. often involving multiple relapses to active smoking, accurate characterization of the former smoker is difficult and best accomplished by longitudinal observation. Misclassification of cigarette smoking status may lead to biased estimates of the consequences of smoking cessation. In observational studies and trials some subjects may report that they are former smokers, even though they continue to smoke: the resulting misclassification tends to result in underestimation of the benefits of cessation. Unraveling the conse- quences of smoking cessation from the effects of other factors determining the occur- rence of disease poses a substantial analytical challenge. In reviewing individual reports on the consequences of smoking cessation, the approaches to these potential methodologic issues were assessed (Chapter 2). TIMN 438420 7

accurate measure that underestimates the amount of cigarettes smoked (Haley and Hoffmann 1985; Marsh et al. 1988; Warner 1978) because subjects often underreport levels of cigarette consumption or misrepresent themselves as former smokers (Luepker et al. 1989; Murray and Perry 1987; Windsor and Orleans 1986; Russell 1982; Stookey et al. 1987). Underreporting also has been linked to "digit bias," that is, subjects tend to report in terms of multiples of ten and underestimate actual consumption (Pechacek, Fox et al. 1984; Vogt 1977; US DHHS 1989). Between 1974 and 1985, estimates of U.S. cigarette consumption based on self-report accounted for only about 70 percent of consumption estimates based on cigarettes taxed and sold (Hatziandreu et al. 1989). This ratio has remained relatively stable. Most of this discrepancy is likely to be due to underreporting or a°`rounding down" to the nearest multiple of a half-pack of daily cigarette consumption (Kozlowski 1986), although misreporting of smoking.status may play a role as well. Validation of self-reports with measures such as biochemical assessments represents a possible means of decreasing misclassification due to misreporting (Luepker et al. 1989; Windsor and Orleans 1986). However, some researchers note that biochemical validation techniques present different problems that also cause misclassification, thus favoring the use of self-report (Assaf et al. 1989: Crossen. Dougher, Belew 1984; Hansen, Malotte, Fielding 1985; Hatziandreu et al. 1989; Kornitzer et al. 1983; Petitti. Friedman, Kahn 1981). As noted above, sensitivity and specificity of the biochemical measures are not perfect. In addition. the procurement of biochemical measures from a large majority of self-reported quitters is not as feasible in large-scale intervention trials or observational studies as it is in smokins studies of a smaller scale and a more clinical nature. Subjects in the population samples do not have the same commitment to studies that volunteers have to clinical studies, and the former are more likely to leave the study area, which makes validation difficult (Ockene et al. 1989). Validation also requires more personal contact than is. eenerally employed in observational or large- scale field studies, and the additional contact may not be acceptable to the subjects or feasible in the context of the study. The section below on physiologic measures discusses methods other than behavioral measures that have been used to assess cigarette smoke exposure. These measures are then contrasted with self-report, and the varying needs for biochemical measurement among different populations are considered. Physiologic Measures Smoking behavior has been assessed by measuring physiologic changes that result from smoking (Pechacek, Fox et al. 1984). Smoking and smoke exposure are reflected in a variety of acute and chronic physiologic measures primarily because of the strong pharmacologic effects of nicotine. These effects include chanQes in heart rate, blood pressure, hand tremor, and skin temperature. Each of these measures has a wide variability under normal conditions and is affected by many factors other than smoking, thus limiting usefulness as a measure of smoking (Pechacek, Fox et al. 1984). 32 7'gMly 438442

a cutpoint depends on the relative importance of mislabeling an actual smoker as a nonsmoker with a very insensitive but specific test versus mislabeling an actual nonsmoker as a smoker with a very sensitive but nonspecific test. This tradeoff between sensitivity and specificity is discussed in more detail elsewhere (Fletcher, Fletcher, Wagner 1987). An important contextual issue concerns the validity with which the biochemical measure classifies individuals. When the test is applied to a population of smokers and nonsmokers, the proportion of the persons who test positive, that is, above the specified cutpoint, who are actually smokers becomes an important concern. This issue, distinct from the question of what proportion of smokers are above the cutpoint, is the crucial measure of how much misclassification occurs. This proportion, the positive predictive value of a test, depends not only on specificity and sensitivity but also on the prevalence of the condition in the population being tested (smoking in this example). The less prevalent smoking is in the screened population the lower the positive predictive value of a test (Browner, Newman, Cummings 1988). The relative misclassification rates for smokers and nonsmokers, determined in part by the estimated prevalence of smoking in the population to which the cutpoints are applied, are particularly important in studies which use biochemical tests to verify self-reported smoking cessation (Cummings and Richard 1988; Ruth and Neaton, in press). For example, the pressure to quit smoking that is present in formal smoking cessation programs may result in a high proportion of continuing smokers who report not smoking. The use of cotinine validation in such circumstances (high prevalence of false reporting) results in a high positive predictive value, as opposed to the lower positive predictive value when the same test is applied to self-reported former smokers identified in a population-based survey (low prevalence of false reporting). In biochemical validation studies, such as those reported in a subsequent section of this Chapter, after optimal cutpoints are set using self-report in one population as the gold standard, the biochemical marker then becomes the gold standard against which self-reported smoking status is measured in another population. ~ Carbon Monoxide High concentrations of CO are present in cigarette smoke (US DHEW 1979; US DHHS 1986, 1989). Absorbed rapidly into the bloodstream during smoke inhalation, CO has a half-life of 4 to 5 hours in sedentary adults (Stewart 1975). Direct measure- ments of CO can be taken from exhaled alveolar air or estimated by measuring the percentage of hemoglobin combined with CO (COHb) (Stewart 1975). Sensitivity of exhaled CO for classifying active smoking is generally in the range of 80 to 85 percent but can be affected by diurnal variability as well as other factors (Benowitz 1983). Given the short half-life of CO, levels are influenced by time of day and time elapsed since last cigarette. Measurements taken late in the day, standardized from time since last cigarette, are likely to give the best estimates of CO levels (Frederiksen and Martin 1979; Horan, Hackett, Linberg 1978; Hughes, Frederiksen, Frazier 1976). Using self-report of recency of smoking can increase sensitivity (Bauman, Koch, Bryan 1982). Sensitivity is poor for light smokers (Fortmann et al. 34 TIMN 438444

Whenever possible, questions should be used that allow continuous rather than dichotomous scales for response. A question such as "Do you smoke regularly?" results in a dichotomous response scale. This scale provides much less information than does a continuous scale, such as the question, "On the average, how many cigarettes do you smoke per day?" which can range from 0 to 20, 40, 60, or more. Multiple questions such as, "Have you smoked even a puff of a cigarette in the past 7 days?"; "How many cigarettes do you typically smoke each day?"; and "How many cigarettes do you typically smoke each week?" can be used to refine a category such as current smokers. Inclusion of other indices, such as biochemical markers of smoking (e.g., saliva cotinine levels), can also be used to describe smoking status. In a followup study, measures of smoking status optimally should be repeated over multiple occasions, especially for dynamic categories like current smokers and former smokers, which are open to change over time. Repeated measures over a series of occasions provide further reliability and validity for assessments and also provide greater statistical power for detecting differences between groups. Nevertheless, studies with only a single or a few assessments of smoking behavior have been extremely informative. ~ Alternative Behavioral Measures As a measure of smoking, self-report by questionnaires and interviews is the most common, the least expensive, the easiest to use. and the most feasible in epidemioloQic studies-(Frederiksen. Martin, Webster 1979: Pechacek. Fox et al. 1984). However, other behavioral measures have also been used in clinical studies. Because these measures are generally not used in large-scale epidemioloQic studies, they will be presented only briefly in this Chapter. Self-monitoring by the smoker, a measure of smoking commonly used in intervention studies, involves recording by paper, pencil, and mechanical counters each ciQarette as it is smoked. The monitoring itself may be•a reactive measure and alter the behavior, depending on the nature of the monitored behavior and motivation (Abrams and Wilson 1979; Frederiksen, Martin, Webster 1979; Lipinski et al. 1973: McFall 1978; Orleans and Shipley 1982). It is an intrusive measure that is normally restricted to smali-studies of high intensity. Other behavioral measures, such as direct observation, collecting and counting cigarette butts (McFall 1978), and measuring their length (Auaer, Wright, Simpson 1972), are even more costly and intrusive and less appropriate ~for epidemiologic and large intervention studies. Alternative types of behavioral reports for validation of smoking status include verification by an informant (Shipley 1981), by self-report measures using multiple questions about smoking behavior or status as part of the same interview or question- naire (see above), and by sampling on multiple occasions. Examples of the latter usually involve lonQ periods of time and often result in multiple sources of dis- crepancy. (See Lee 1988 for summary.) 29 TIMN 438439

1984; Vogt 1982), and specificity can be reduced by exposure to CO present in the environment as a result of industrial and automobile pollution, environmental tobacco smoke, indoor combustion sources, and use of products such as marijuana (Biglan et al. 1985; Frederiksen and Martin 1979; Stewart 1975). In spite of this, only 2 to 5 percent of nonsmokers in general populations will exceed 1 percent COHb (Janzon et al. 1981; Kahn et al. 1974). Using COHb levels from a national probability sample, the Radford and Drizd (1982) reported the 95th percentile for COHb to be 1.77 percent in nonsmokers, aged 12 to 74. If a 2-percent cutpoint is applied to this sample, 3.6 percent of nonsmokers would be incorrectly classified as smokers. Thiocyanate High concentrations of HCN, a toxic gas, are present in cigarette smoke. However, HCN is very active chemically and is rapidly detoxified by the liver into SCN- (Langer and Greer 1977; Boxer and Rickards 1952). Because SCN-accumulates in body fluids, such as saliva, urine, and blood, it is used as a biochemical measure of exposure to tobacco smoke. The biologic half-life of SCIY- has been found to vary quite a bit (Bliss and O'Connell 1984) although the length of time usually noted is between 10 and 14 days (Langer and Greer 1977; Vesey 1981). Salivary SC1V- can be measured most reliably in parotid gland secretions (Shannon, Suddick, Dowd 1974); however, parotid gland secretions show some seasonal and diurnal variability (Shannon, Suddick, Dowd 1974). When serum and saliva samples are compared, the levels are 15 to 20 times higher in saliva than serum (Langer and Greer 1977; Pechacek et al. 1979; Vesey 198 1). However, saliva levels are more variable (Pechacek et al. 1979). The increment of SCN- in light smokers is low, and there is much overlap of SCIV_ levels in light smokers compared with nonsmokers (Fortmann et al. 1984; Neaton et al. 1981; Vesey et al. 1981). However, detection of light smoking in adults using SCN- levels is better than in adolescents (Windsor et al. 1985). This is likely to be related to the fact that adolescents are often in the process of learning how to smoke and inhale, and they may not have an established pattern of smoking (Pechacek, Murray et al. 1984). For example, among younger adolescents only one-third or less could be identified on a single assessment (Hunter, Webber, Berenson 1980; Luepker et al. 1989: Pechacek. Murray et al. 1984). Specificity represents a more severe problem than sensitivity. A large number of food products are sources of either cyanogenic glycosides (e:g., almonds, bamboo shoots, sugar cane) or naturally occurring SCN- (e.g.. cauliflower, broccoli, beer) and can produce levels of SCN- in saliva equivalent to the average levels of smokers (Langer and Greer 1977; Neaton et al. 1981; Pechacek et al. 1979: Swan et al. 1985). The relatively low specificity and sensitivity of SCi%T_ testing compared with cotinine and CO make SCIN_ a less useful outcome measure for smokina cessation studies (Gillies et al. 1982; Fortmann et al. 1984) unless adjustments are made using carefully collected dietary and environmental exposure data. A prime advantage of using SC1V- for biochemical validation of smoking abstinence is its long half-life compared with other biochemical measures (Fortmann et al. 1984; Steinman 1985; Murray et al. 1987; TIMN 438445 35

References ROGOT, E., MURRAY,I.L. Smoking and causes of death among U.S. veterans: 16 years of observation. Public Health Reports 95(3):213-222, May-June 1980. SCHWARTZ,I.L. Review and Evaluation of Smoking Cessation Methods: United States and Canada.1978-1985. U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, NIH Publication No. 87-2940, April 1987. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking for Women. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health, 1980. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Involuntary Smoking. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control. DHHS Publication No. (CDC) 87-8398, 1986. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking: Nicotirie Addiction. A Report of the Surgeon General.1988. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control, Center for Health Promotion and Education, Office on Smoking and Health. DHHS Publication No. (CDC) 88-8406. 1988. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. Reducing the Health Conse- quences of Smoking: 25 Years of Progress. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service. Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. DHHS Publication No. (CDC) 89-8411, 1989. U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE. Smoking and Health. A Report of the Surgeon General. U.S. Department of Health, Education, and Welfare. Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health. DHEW Publication No. (PHS) 79-50066, 1979. U.S. PUBLIC HEALTH SERVICE. Smoking and Health. Report of the Advisoi-v Committee to the Surgeon General of the Public Health Service. U.S. Department of Health, Education, and Welfare, Public Health Service, Center for Disease Control. PHS Publication No. 1103, 1964. WEST, S., MUNOZ, B., EMMETT, E.A., TAYLOR, H.R. Cigarette smoking and risk of nuclear cataracts. Archives of Opthalmology 107(8):1 166-1169, August 1989. TIMN 438428 15

INTRODUCTION Smoking cessation is a dynamic process that begins with a decision to stop smoking and ends with abstinence from cigarettes maintained over a long period of time. Typically, initiation of regular cigarette smoking occurs at a young age, usually during the teenage years (US DHHS 1989); cessation may be contemplated and initiated at any age. The spectrum of factors motivating cessation is diverse; some smokers quit before being adversely affected by cigarette smoking whereas others quit as a result of developing smoking-related disease. Most attempts to quit are temporarily successful, and most smokers attempting to quit return several times to regular smoking before achieving long-term abstinence. For the purpose of health research, smoking status (i.e., never, former, or current smoker) can be evaluated by using an interview or questionnaire to query subjects about their smoking behavior. However, self-reports may not fully characterize the process of cessation in individual smokers, particularly if information is collected retrospec- tively or cross-sectionally. Moreover, persons who are smoking may falsely report themselves as former or never smokers. Biochemical markers, such as cotinine and thiocyanate (SCN-) levels in body fluids, provide complementary measures of tobacco product use. However, reliance solely on biochemical markers of smoking also may lead to some misclassification. For example, intake.of some foods can result in high SCi~ levels unrelated to smoking behavior. Individuals who accurately report being quitters may fail to participate in the validation process and therefore may be misclassified as .continuing smokers if nonparticipants in biochemical testing are assumed to be smok- ing. Because proper classification of smoking behavior is critical for conducting research on the health consequences of smoking cessation and for evaluating the results of such research, it is important to consider how smoking status is assessed. The health consequences of smoking cessation have been studied using conventional approaches of epidemiologic and clinical research: ecologic study, cross-sectional study or survey. case-control study, cohort study, and intervention trial. Each design has well-described advantages for studying causes of disease and preventive factors among human populations (Kleinbaum, Kupper, Morgenstern 1982). In addition, each design type is subject to the three types of bias potentially affecting any epidemiologic study: selection bias, information bias, and confounding bias (Rothman 1986) (Chapter 2, Part II). Misclassification resulting from information bias is of particular concern in studies of smoking cessation; misclassification is addressed in detail in this Chapter. These conventional research designs have been used successfully to characterize the adverse effects of active cigarette smoking and to amass the scientific information on smoking cessation reviewed in this Report. For example, the evidence on smoking cessation and mortality derives from cohort studies (Chapter 3); evidence on cancer comes largely from case-control and cohort studies (Chapters 4 and 5); and information on respiratory morbidity and mortality is based primarily on cross-sectional and cohort studies (Chapter 7). This Chapter establishes a methodologic framework for interpreting the evidence on smoking cessation obtained from observation studies and intervention trials. Part I TIMN 438431 21

describes the process of smoking cessation and the methods used to assess smoking behavior. Part II reviews research methods used to study smoking cessation as well as the potential limitations of data obtained from observational studies and intervention trials including biases that may affect the results. PART I. ASSESSING THE DYNAMIC PROCESS OF SMOKING CESSATION This Section describes the dynamic nature of smoking behavior, the various measures of smoking status applied in observational and intervention studies, and the effect of these measures on classification of smoking status. The Process of Smoking Behavior Change Smoking behavior in U.S. populations has been changing, and three-fourths of all smokers have attempted to quit (Volume Appendix). The proportion of adult former smokers in the population is now about the same as the proportion of current smokers. These population changes have provided opportunity to describe the consequences and, thereby, the benefits of cessation. Progressing from smoking to former smoking is a complex, dynamic process and not a one-time event. Retrospective, cross-sectional, and longitudinal studies of how people quit smoking on their own have demonstrated that smokers move through a series of stages in their cessation efforts (DiClemente and Prochaska 1982; Lichtenstein and Brown 1980; Prochaska and DiClemente 1983; Prochaska et al. 1985; Rosen and Shipley 1983). These stages have been labeled motivation and commitment, initial change, and maintenance by Brownell and coworkers (1986); contemplating change, decidingt change, short-term change, and long-term change by Horn ( I976); motivation and commitment, cessation and possible relapse, and maintenance by Marlatt and Gordon (1985); precontemplation, contemplation, action, and maintenance and/or relapse by Prochaska and DiClemente (1983); and initial decision, initial control, and maintenance by Rosen and Shipley (1983). The stage model of Prochaska and DiClemente (1983; Prochaska et al., in press) has generated the most research and is described in more detail below (Figure 1). Pre- contemplation is a period in which smokers are not thinking about quitting smoking, or at least not about quitting within the next 6 months. The basis for the 6-month timeframe is the assumption that 6 months into the future is as far as most people plan a specific behavior change. Contemplation is the period in which smokers seriously consider quitting smoking within the next 6 months. Action is the period that begins when actual cessation occurs and continues for 6 months after stopping smoking. Maintenance is defined as the period beginning 6 months after cessation occurrence. In all of the proposed stage models, differentiation is made between short-term (generally up to 6 months) and long-term (generally 6 months and longer) change or between initial cessation and maintenance of cessation. Maintenance continues until relapse to regular smoking, or until a return to regular smoking is of minimalor no concern and "termination" of the behavior occurs for the confirmed ex-smoker. 22 TIMN 438432

INTRODUCTION The 1964 Report of the Surgeon General's Advisory Committee on Smoking and Health (US PHS 1964) concluded that cigarette smoking is a cause of lung cancer and laryngeal cancer in men, a probable cause of lung cancer in women, and the most important cause of chronic bronchitis. Other diseases, including emphysema and cardiovascular disease, also were found to be associated with cigarette smoking, although the evidence available at that time was not viewed as sufficient to establish the associations as causal. Even in 1964, however, the evidence for adverse health consequences of cigarette smoking was sufficient for the Committee to conclude that "cigarette smoking is a health hazard of sufficient importance in the United States to warrant appropriate remedial action" (US PHS 1964, p. 33). Subsequent reports of the Surgeon General on smoking and health expanded and strengthened the conclusions of the 1964 Report on active smoking •and documented the benefits of smoking cessation. (See US DHHS 1989 for review.) For some diseases, such as cardiovascular disease, newer evidence warranted a determination that associations with cigarette smoking were causal. Further associations of cigarette smoking with disease were identified, and involuntary (passive) smoking was found to be a cause of disease in nonsmokers (US DHHS 1986). Although cisarette smokine has been investigated intensively since the 1950s, new associations of smoking with adverse effects continue to be identified. For example, in a recent study smoking was associated with cataracts (West et al. 1989). Evidence substantiates cigarette smoking as a cause of disease in smokers and, through invbluntary smoking, in never smokers as well. This evidence has motivated the implementation of diverse and far-reaching programs for smoking prevention and 'cessatioin. The proportion of U.S. adults who smoke decreased substantially since the 1964 Report. In 1965, 29.6 percent of persons who had ever smoked had quit; by 1987, this percentage had increased to 44.8, representing more than 38 million adults. As the numbers of formerly smoking adults increased in the United States and other countries (US DHHS 1989), epidemiologic and clinical studies provided increasingly extensive information on the health benefits of smoking cessation. Thus, the 1964 Report noted that former smokers had lower overall mortality rates and lower lung cancer risk than current smokers, but the cited evidence was limited. Scientific data are now available on the consequences of cessation for most smoking-related diseases. Major benefits have been shown for overall mortality and for many specific diseases. Although past reports have considered much of the evidence, these data have not received a com- prehensive and unified review. This Report systematically reviews the findings on the health benefits and consequences of cessation. This Report includes a Foreword by the Assistant Secretary for Health and the Director of the Centers for Disease Control, a Preface by the Surgeon General of the U.S. Public Health Service, and the following chapters: Chapter 1. Introduction, Overview, and Conclusions Chapter 2. Assessing Smoking Cessation and Its Health Consequences 5 TIMN 438418

Pechacek, Fox et al. 1984), which is of particular interest in population surveys where longer term abstinence is of concern. Cotinine Cotinine, a metabolic byproduct of nicotine, is distributed throughout extracellular fluid and is excreted through the kidneys and salivary glands (Benowitz 1983). About 15 to 20 percent is eliminated in the urine unchanged, and the rest is metabolized (Benowitz 1983). The half-life estimates of cotinine are variable and range from 15 to 40 hours (Carey and Abrams 1988: Knight et al. 1985; Greenberg et al. 1984; Haley and Hoffmann 1985; Haley et al. 1987; Sepkovic. Haley, Hoffmann 1986). The differences in estimated half-life for cotinine reflect not only individual differences in metabolism but also differences between smokers and nonsmokers (Haley, Sepkovic, Hoffmann 1989; Sepkovic, Haley, Hoffmann 1986; Haley et al. 1987). Cotinine levels vary with the diurnal cycle and are best assessed late in the day (Benowitz 1983). Methods are available for measuring cotinine in saliva, urine, and blood. Urinary levels have been suggested to be too variable (Pechacek, Fox et al. 1984), and plasma or serum levels appear to be the most stable (Benowitz 1983). However, sampling saliva because of ease of procurement and accuracy in classifying smokers and nonsmokers has been recommended as a useful, noninvasive method that can be applied to large-scale intervention trials (Abrams et al. 1987). Because nicotine is unique to tobacco, cotinine is a highly valid marker for almost any tobacco use (Haley, Axelrad, Tilton 1983; Russell et al.,1981; Wald et al. 1984; Zeidenberg et al. 1977). Although nicotine has been assessed in some studies, it is recommended that cotinine be used because it has a more enduring and stable blood level (Langone. Gjika, Van Vunakis 1973). Detecting regular smokers by analysis of cotinine in blood, urine, or saliva is almost certain, and even light smokers and intermittent smokers are easily detected (Benowitz 1983: Haley, Axelrad, Tilton 1983; a et al. 1977; Carey and Abrams 1988; Williams Paxton and Bernacca 1979; Zeidenber( et al. 1979). In one investieation, 95 percent of adolescent ever smokers were detected by cotinine (Williams et al. 1979). Specificity is also hiah; regular smokers typically have blood cotinine levels of 200 to 400 n,-/mL, light smokers have 40 to 50 ng/mL, and nonsmokers are typically below 10 ng/mL. When nonsmokers are assessed, they rarely have any detectable cotinine (Benowitz 1983; Haley, Axelrad. Tilton 1983; Sepkovic and Haley 1985; Zeidenberg et al. 1977). In comparative studies of different biochemical measures of smoking, cotinine has emerged as the measure of choice (Abrams et al. 1987; Haley, Axelrad, Tilton 1983; Jarvis et al. 1984, 1987; Knight et al. 1985; Pojer et al. 1984) because of its superior sensitivity and specificity. However, it is more expensive and more analytically complex than the other biochemical measures. The value of biochemical measures is limited to short-term abstinence and cannot be used to document continuous abstinence in long-term studies. CO, with a half-life of 4 to 5 hours, can validate self-reports of not having smoked in the past 24 to 48 hours (Benowitz 1983). Cotinine, with a half-life of 15 to 40 hours, would have limited application for validation beyond a few days. SCIV-, with a half-life of 10 to 14 days, 36 TIMN 438446

interpreted cautiously. For example, in the reports of the Whitehall Civil Servants Study (Rose and Hamilton 1978; Rose et al. 1982), the criterion used to define abstinence is not indicated. The only information provided is that the smokers reported that "they were then smoking no cigarettes at all" (Rose and Hamilton 1978). Regardless of the criteria used to define abstinence, the methodology for assessing smoking status, including questionnaire items, needs to be carefully described by investigators. Optimally these items should enhance the process of obtaining informa- tion regarding the duration of abstinence, making it possible to fully determine the relationship of smoking cessation to health and disease outcomes. When reviewing studies of the health effects f smoking, the definition of the former smoker must be carefully assessed, and the effect of the definition on the findings must be carefully examined. Temporal and Frequency Issues Studies vary according to whether smoking is assessed retrospectively or prospec- tively and whether a single assessment or a series of assessments is used. The category of never smokers can be assessed retrospectively, usually relying on a single assess- ment. Requiring subjects to reconstruct more detailed smoking histories can be very demanding. Nevertheless, simply classifying individuals as former smokers or current smo s reveals very little about the amount of smoking exposure experienced. More pert. .:,nt questions regarding exposure include "How long have you been abstinent from cigarettes?"; "At what age did you start smokinQ9"; "How many cigarettes did you smoke during different periods of your life?"; "How many times did you stop smoking?"; and "How long did you remain abstinent during each of these occasions?" A series of repeated assessments can result in inconsistencies such as some in- dividuals reporting smoking at one assessment and later reporting that they never smoked. In a followup study in England, for example, Britten (1988) found 1,296 participants aged 36 who claimed that they had never smoked. Of these. 242 (18.7 percent) previously had reported smoking less than 1 cigarette per day, and 102 (7.9 percent) previously had reported smoking at least 1 cigarette per day for at least 1 year. Of the 102 who reported previously that they had been regular smokers, 93 percent reported that the last time they had smoked was at least 10 years prior to the survey. If the Britten study had used only one retrospective assessment of the subjects at age 36, 32.5 percent of the 1.296 subjects would have been classified as never smokers and 32.6 percent as former smokers. Assuming that reports at a young age were more accurate because memory bias was less likely to occur, the serial assessment indicates that a more accurate categorization would be 29.1 percent for never smokers and 36.5 percent for former smokers. Britten (1988) estimated that misclassification of this magnitude, when applied to a study by Friedman and colleagues (1979), would result in only a 5-percent increase from 2.41 to 2.53 in relative risks of death for former smokers compared with never smokers. Krall and colleagues (1989) found that of 87 middle-aged adults, 87 percent accurate- ly recalled their smoking status of 20 years earlier, but only 71 percent accurately recalled the amount that they had smoked. Furthermore, underestimation of the amount 27 TIMN 438437

Surrogate Assessments In some circumstances researchers may need to obtain information from sources other than the index subjects. With some study designs, for example a case-control study of lung cancer, some subjects are unavailable to answer questions because of illness or death. In cohort studies, or intervention studies with mortality endpoints, surrogate interviews are sometimes required to assess smoking during the interval preceding death. Failure to obtain surrogate reports can cause considerable bias in some instances. In a case-control study of oral cancer, Greenberg and coworkers (1986) obtained inter- views with 112 cases (67.9 percent) and surrogate reports for 23 cases (13.9 percent). Cases needing surrogate reports had more advanced stages of disease at the time of diagnosis and were more likely to be black and less educated than cases interviewed in person. Cigarette smoking and drinking hard liquor were more common among these cases. Therefore, failure to include surrogate reports would have resulted in under- estimates of the strength of association between cigarette exposure and hard liquor and the risk of oropharyngeal cancer. Pickle, Brown. and Blot (1983) found that siblings of index subjects provided the most complete data about smoking in the subject's family of origin and early life events. Spouses and offspring supplied the most complete data about smoking history during adult life. Incomplete data generally increased with the amount of detail requested, so that there were considerably higher nonresponse rates for a detailed smoking history (approximately 50 percent) than for the history of a broad smoking status, such as never smoker (approximately 15 percent). Surrogates beyond a spouse or close relative provided much higher nonresponse rates for almost all questions in all statuses. McLaughlin and colleagues (1987) examined the reliability of retrospective surrogate reports obtained 10 years after initial reports and compared these with retrospective self-reports using data from the NHANES-I (Cornoni-Huntley et al. 1983). Correct identification of previous smoking status was generally provided by most types of surrogates, except siblings of male decedents. The combined level of agreement for all surrogates ranged from 85 to 95 percent and was remarkably similar to that from self-reports of living subjects. Thirty-five percent of the surrogates could not provide data on when smoking began compared with 1 percent in self-reports. Surrogates who responded tended to provide a later age for starting. Surrogates did, however, provide estimates of years smoked that were comparable to the original reports. In this study, siblings and other surrogates provided less reliable reports than spouses, offspring, or parents of subjects. Lerchen and Samet (1986) interviewed widows of lung cancer patients who had supplied their own smoking histories while alive. They found that of 77 wives of current smokers, all supplied information about the cases' cigarette smoking status (ever/never) that was in perfect agreement with the information supplied by the cases themselves. Sixty-six (86 percent) were able to supply complete responses about their husbands' smoking behavior. For those who responded, however, mean values reported by cases and their wives were not significantly different for age at which cases started smoking, years smoked, or average number of cigarettes smoked per day. Wives tended to report 30 TIMN 438440

requires some period of "regular" smoking for an individual to be classified as an ever smoker, 128 of 252 individuals reported being never smokers. However, when assessed concurrently with another questionnaire in which regular smoking was not defined and the respondent self-defined smoking, 7 percent fewer subjects (119 of 252) reported being never smokers. Thus, the use of more clearly defined questions, such as specifying 100 cigarettes in a lifetime, or I cigarette per day for 1 year, or 5 cigarettes per week for 1 year, will reduce misclassification. However, some misclassification will still occur for those individuals who smoked for relatively brief periods during their lives but cannot accurately remember how long they smoked or accurately estimate the number of cigarettes they smoked. Attention also must be paid to defining current or former smokers. Some studies, such as the Cancer Prevention Study I(CPS-I) (Hammond and Garfinkel 1969), define current smokers as those who respond affirmatively to the question "Have you smoked within the past year?" Other studies use smoking in the past 6 months as the guideline for current smokers (Coultas et al. 1988). The criteria for questions identifying current smokers can range from having smoked in the past year, to the past 6 months, to the past week, or to an unspecified period. A few additional questions will enhance the specificity of the definitions of current smokers and former smokers. These items, or comparable ones, have been used in previous surveys, for example, the 1988 Baseline Prevalence Survey for the Community Intervention Trial for SmokinQ Cessation, funded by the National Cancer Institute: "At what age did you start smoking on a regular basis?"; "On the average, about how many cigarettes did you smoke per day during the last 12 months you smoked?"; and for former smokers, "When did you quit smoking cigarettes?" (recorded to exact date if possible). These items provide addi- tional information for defining ever smokers, or stratifying by levels of exposure, and for determining the period of abstinence. The dynamic nature of smoking cessation hiQhlights the importance of being aware that any categorical definition of former smoker in relation to the health effects of smokinQ cessation will include former smokers who have been abstinent for varying periods of time. Optimally, questions on smoking history should ascertain the duration of abstinence for former smokers, and if possible, abstinence periods should be treated as continuous or cateaorized variables in an analysis, thus avoiding the problem of treating former smokers as a single group. However, benefits of cessation are still clearly observed in spite of the limitations of using categorical data. The most common minimum periods of abstinence used for defining former smoking status are 24 hours, 7 days, and 30 days. The National Interagency Council on Smoking and Health (1974) recommended using a minimum of 7 days of abstinence for detining, cessation. However, because of the nature of smoking, using a short abstinence period to define former smokers is not optimal in epidemioiogic studies. The degree of misclassification of former smokers will depend on the minimum duration of abstinence used to define former smokers and the criterion used to consider determine relapse. Many studies do not specify a minimum duration of abstinence for individuals classified as former smokers at a particular point in time. Data from such studies on the association of smoking cessation with health and disease outcomes must be 26 rrgNgN 438436

20 cigarettes smoked daily even when their husbands smoked substantially more or less. Pershagen and Axelson (1982) also reported perfect agreement regarding smoker/nonsmoker status when information was obtained from a close relative (parent, wife, or child) for 141ung cancer cases compared with information that had previously been obtained from the cases by the physician. Blot, Akiba, and Kato (1984) also interviewed next of kin in a case-control study of lung cancer among atomic bomb survivors who had previously provided information regarding their own smoking behavior while they were alive. The investigators found that only 1 percent of surrogates reported that a subject had never been a smoker while the subject reported that he or she had smoked, suggesting that the identification of never smokers by next of kin is very accurate. There was poorer agreement regarding those who smoked, with 13. percent of surrogates indicating that a subject had smoked while the subject had reported never smoking. Sandler and Shore (1986) examined the quality of data provided by adult offspring on parents' smoking and drinking. The data were from 518 cancer cases and 518 healthy controls aged 15 to 59. When possible, mothers provided data on their own smoking and their husbands' smoking. Of 982 subjects who had lived with their natural mother, 97 percent provided data on their mothers' smoking status. Of those whose mothers reported never having smoked cigarettes, 2.7 percent were reported as ever having smoked by the adult child. Of those mothers who reported ever having smoked, 8.8 percent were reported as never smokers. Of those fathers reported by the mother as never smokers. 17.2 percent were reported by subjects as ever smokers. Of those fathers reported as ever having smoked ciaarettes, 21.1 percent were reported as never smokers by their adult children. Even with the quantity of cigarettes collapsed into categories to include answers of less than 1 pack, I pack, and more than 1 pack, the proportion of mothers and subjects whose responses exactly agreed was 82.0 percent for mothers and 49.2 percent for fathers. Humble, Samet, and Skipper (1984) interviewed 46 subject-spouse pairs, with 2 people in each of 38 of these pairs acting as the subject and as a surrogate for his/her spouse, thus producing 84 total subject-surrogate pairs. For the 30 current or previou~ cigarette smokers whose spouses gave complete smoking data regarding the subjects, the subjects reported a mean use of 17.8 cigarettes per day compared with 14.3 reported by their spouses. The difference was not significant. Investigations indicate that useful information on smoking can be obtained in epidemiologic investigations that must rely on surrogate information (McLaughlin et al. 1987). Although greater misclassification occurs when surrogate reports are used compared with self-reports, consideration of variables such as the relationship of the informant, lensth of time he or she had known the case, the topic of the questions, and complexity of the data Qathered from the informant can add to the validity of the data (Rogot and Reid 1975). Nonbehavioral Measures Methods other than self-report have been used to assess smoking status. Some researchers have expressed concern that self-report when used alone can be an in- 31 TIMN 438441

who quit smoking were at greater risk for death than those who did not quit because of more severe underlying disease (Vlietstra et al. 1986; Hermanson et al. 1988). Analytic Issues in Observation Studies Complex associations among disease risk, age, and duration of active smoking and abstinence further complicate assessment of the health consequences of cessation. Analytic approaches should represent these relationships in a biologically appropriate fashion. The risks of many cigarette-related diseases (e.g., cancer, CVD, and chronic obstructive pulmonary disease) increase with age (Figure 2). Following cessation, disease risk may change in diverse patterns, depending on the disease-specific mechanisms through which cessation alters disease occurrence. Disease risk may be unaltered (Curve A), decline quickly or slowly compared with that for never smokers (Curve C), or decline to a level between that of never and persistent smokers (Curve B) (Figure 2). Comparing the disease risk for former smokers with the risk for persistent AGE 0 FIGURE 2.-Hypothetical examples of disease incidence rates for current, former, and never smokers, by age TIMN 438465 55

CHAPTER 3 SMOKING CESSATION AND OVERALL MORTALITY AND MORBIDITY TIMN 438480 71

CONTENTS Introduction ........................................................ 75 Smoking Cessation and Overall Mortality in Cohort Studies .................. 75 Smoking Cessation and Overall Mortality in Intervention Studies .............. 84 Smoking Cessation and Medical Care Utilization ........................... 87 Population Projections .............................................. 87 Observational Studies ............................................... 87 Smoking Cessation and Health Status .................................... 87 Conclusions ........................................................ 92 Chapter 3 Appendix .................................................. 93 References ......................................................... 99 TIMN 438481 73

TABLE 2.-Measures of false reports from studies using CO as a marker Reference Population ToW to give up Criterion for false reports of not smoking % (n/N) False reports Comments Part I. Volunteer samples Jones, Commins, London taxi drivers No 6.6%, COt lb 4.8 (1 /21) Cernik(1972) Petitti, Californians having health checkups, No 8ppmCO 0.6(I/181) Friedman, Kahn (1981) 176 female twins and 91 males Jarvis et al. (1987) Clinic outpatients No I O ppm CO (expired air) 16 (19/121) No 1.7%CO(Hb) 18(22/121) ~-j Bauman, Koch, Adolescent nonclinic selling No 6 ppm CO (expired air) 0 Bryan (1982) 8 ppm CO (expired air) 3 ~ ~e Stookey el al. (1987) Cessation study Yes 8 ppm 0 (0/20) .p. W Fortmann et al. Representative sample for No 8 ppm 4.2 (37/890) 00 (1984) cardiovascular risk study CA Part II. Intervention samples - P-A Delarue (1973) Canadians attending voluntary Yes 2% COI 1b 20.6 (22/107) I-yr followup antismoking clinic 4% COl lb 9.3 (10/107) 6"Ye COI lb 4.7 (s/107)

TABLE L-CoMinued ,Reference Part 111. High-risF:/medical patients Population Told to give tip Criterion for false reports of not smoking % (n/N) False reports Wilcox, Flughes, Roland ( 1979) Nottingham MI patients Yes 2 Ng/I(H) mL urinary nicotine or 10 pg/i (H) mL urinary cotinine 16.3 (8/49) Jarvis et al. (1987) Clinic outpatients No 13.7 ng/mL serum cotinine 19 (23/121) No 14.2 ng/mL salivary cotinine 18(22/121) No 49.7 ng/mL urinary cotinine 17 (21/121) No 21.8 ng/mL salivary nicotine 14(17/121) No 2.3 ng/mL plasma nicotine 14(17/121) No 58.6 ng/mL urinary nicotine 16(19/121) Haddow et al. (1987) US pregnant women No 10 ng/mL serum 4.9 (142/2,871) Comments An additional 5 subjects had detectable levels in concentrations below the cutpoint Unpublished data NOTE: n/N=numbcr of individuals reporting not smoking but with levels of biochemical marker Lxcculing cutpoint divided by all individuals reporting not smoking; M1=myocardial infarction. "IIritish United f'rovidence Association Medical Center in London. b Studies classified as owliers due to low criterion cutoffs. SOURCE: Adapted from l.ec (1988).

TABLE 1.-Continued Former smokers All durations Study Current Temporary Persistent smokers quitters quitters California HMOt 1.82 1.51 1.13 (Friedman et al. 1981) NQl'E: All monalily ratios are relative to never snwkers. ACS CPS-I=Amcrican Cancer Society Cancer Prevention Study I; IIMO=ltealth Maintenance Organization. 'Age-aJjusteJ. °AgeJ 50-74. `Aged 54-64. d Tobacco consumption in g/day. °Furmer smokers are those with sustained abstinence. ~ (Persistent quitters are those with sustained abstinence.

t TABLE 2.-Continued Told to Criterion for false % (n/N) Reference Population give up reports of not smoking False reports Comments Vogt et al. (1977) San Francisco Center of MRFIT Yes 8 ppm CO 4.4 (2/45) Sillett et al. UK study in 2 groups: (A) survivors of Yes 1.7% COl lb A: 21.6 (11/51) ~(1978)" Ml and (B) volunteers in nicotine gum B: 40.2 (33/82) trial Ronan et al. Irish post-MI patients Yes 1.6%, COI lb 8.8 (5/57) Mean 8.6-yr followup (1981) Research UK patients with smoking-related All groups 1.6% COI lb and 27 27% false reports rate Committee of the diseases in 4 group intervention ' 73 Ntnol/L SC.N- 25 ut 6-mo followup; British Thoracic trials involving advice, booklet, in plasma 25% false reports rate Society ( I983)a placebo, and nicotine polacrilex gum al I-yr followup NOTE: CO=carbon monoxide; n/N=number of individuals reporting not smoking bw with levels of biochemical marker exceeding cwpoint divided by all individuals reporting not smoking; COlib=carboxyhemoglobirr ppm=parts/million; SCN-=lhiocyanate; MRFIT=Multiplr Risk Factor Intervention Tri:d; MI=myocardial infarction. "Studies classified as outliers due to low criterion cutoffs. °May be same group as (l3) in Silleu et al. (1978). SOURCE: AJapted from Lxe (1988).

CHAPTER 3 APPENDIX 93 TIMN 438500

maintained cessation rates were significantly greater in the special intervention than in the usual care group, to date the difference has not been large enough to provide adequate statistical power to assess the effect of smoking cessation alone on differences in morbidity and mortality between the intervention and control groups (Chapter 3). However, MRFIT was designed as a multifactor trial and did not assess the impact of smoking cessation alone. Because MRFIT results indicated the greatest difference in smoking cessation between special intervention and usual care subjects compared with any other clinical trial and still lacked the power to detect outcome differences from smoking cessation, it is unlikely that smaller trials would have sufficient power to demonstrate an effect of cessation on morbidity and mortality (Chapter 3) (US DHHS 1983). Compared with observational studies which place few demands directly on subjects, the use of interventions for smoking cessation in clinical trials increases the probability of misreporting smoking status at postintervention followup because of the expectations of the participants and the investigators. Typical periodic followup in clinical trials, however, reduces the chances of misclassification related to relapses or to delayed action to quit smoking-phenomena that are often not adequately recorded in observa- tional studies. Routine followup also allows for more accurate measurements of the duration of prolonged or continuous abstinence and the opportunity to validate with biochemical testing. Intervention trials other than clinical trials also provide information on the health consequences of smoking cessation. A number of studies are in progress involving interventions of varying intensity within a community. The North Karelia project conducted in Finland is such a community trial; a comprehensive, community-based intervention program was condticted to reduce cardiovascular disease (CVD) (Tuomilehto et al. 1986). Mortality rates in North Karelia were compared with those in other areas of Finland. Methodologic Issues Introduction Epidemiologic studies have been the principal source of information on the health benefits of smoking cessation. Although the resulting data have provided strong evidence for the benefits of cessation, the data need to be interpreted with consideration of potential sources of bias and of other methodologic issues. This Section considers the methodologic issues potentially affecting interpretation of studies of the health consequences of smoking cessation. The criteria for causality have served as a basis for evaluating all of the evidence relevant to a particular association (US PHS 1964; US DHHS 1982. 1989). However, associations found in individual studies must also be assessed carefully. In any epidemiologic or clinical study, association may result by chance, as the result of bias, or through a causal mechanism. Thus, this Section presents an overview of statistical considerations relevant to studies of smoking cessation and the most prominent sources of bias in such studies-information bias and 51 TIMN 438461

TABLE 1.-Measures of false reports of not smoking from studies using nicotine and cotinine as a marker Reference Population Told to give up Criterion lor I:dse reports of not smoking % (n/N) False reports Pan L Volunteer samples Russell and London smokers, nonsmokers, and No Urinary nicotine 0(0/27) Feyerabend ( 1975) Williams et al. heavy passive smokers Students health screening No Plasma cotinine 2 (2/98) (1979) Haley, Axelrad, New York nonsmoking volunteers No Salivary or plasma cotinine 0(0/18) Tilton (1983) Wald et al. (1984) Nonsmokers attending BUPA', and No 256 ng/mL urinary cotinine 0.9 (2/221) Haddow, Oxford colleagues US women attending well-women No 30 ng/mL serum colinine 1.3 (3/232) Palomaki, Knight screening 10 ng/mL serum cotinine 2.2 (5/232) (1986) Coultas et al. New Mexico Ilispanic children and No 50 ng/mL salivary cotinine 3.2 (43/1,360) (1987) adults in household survey Lee(1987) Representative UK sample providing No 30 ng/ntL salivary cotinine 2.5 (20/808) saliva, without prior warning, after 10 ng/ntL salivary cotinine 4.2(34/808) Nanji and smoking data Lab sample No I µg/mL skin nicotine Q (0/43) Lawrence (1988) Pierce et al. (1987) Sydney, Melbourne smokers No 250 nmol/L salivary cotinine 4.0 (25/622) Comments No overlap between range of urinary nicotine levels of smokers (N= 18) and nonsmokers (N=27) No culpoint established; no cotinine detected in nonsmokers Cutpoint based on distribution 46.3% of sample below age 18 yr

In large-scale studies, use of biochemical assessments is generally not feasible; thus, mandatory use of such assessments and subsequent classification of refusers as smokers (as suggested by some investigators involved in clinical intervention studies e.g., Windsor and Orleans 1986) would result in an unacceptable distortion of the outcome data. In addition, some subjects may drop out if validation is required. The effect of lost subjects on study results may be difficult to estimate. In contexts other than intensive intervention trials, self-reported smoking status at the time of measurement and concurrent biochemical assessment have been demonstrated to be highly concor- dant (Fortmann et al. 1984; Petitti, Friedman, Kahn 1981) (Tables 1 and 2). This high concordance supports the use of self-report as a valid measure of smoking status in observation studies of the health effects of smoking cessation. PART II. ASSESSING THE CONSEQUENCES OF SMOKING CESSATION Study Designs Used to Assess the Consequences of Cessation Overview of Study Design Most evidence on the health benefits of smoking cessation derives from studies of human populations and not from animal studies or other types of research. Research on humans can be classified as experimental (the investigator assigns subjects to be exposed or not exposed to the risk factors or preventive factors of interest) or observa- tional (the investigator does not determine whether subjects are exposed or not exposed to the factors of interest; exposure reflects the subjects' choices or some other process). Intervention studies include randomized or nonrandomized community-based inves- tigations and clinical trials. The clinical trial, involving randomization of subjects to be exposed or not exposed to an intervention, has been used to investigate the effects of smoking cessation in patient groups and in populations: The observational designs include the ecologic study, the cross-sectional study, the cohort study, and the case- control study. The biases potentially affecting these studies can be broadly classified as selection bias, information bias, and confounding bias (Table 3) (Kleinbaum, Kupper, Mor- genstern 1982). Selection bias refers to distortion of an exposure-disease relationship by the mechanism through which subjects are selected. Information bias arises from the incorrect categorization of subjects as exposed or not exposed or as diseased or not diseased. The resulting misclassification of subjects on exposure or disease status may occur in a random or nonrandom fashion (Chapter 2, Part I). Confounding bias refers to the distortion of the apparent effect of an exposure on risk caused by association with other factors that affect outcome (Last 1988). In the subsequent review of the study designs used to assess the benefits of smoking cessation, sources of bias most relevant to each design are highlighted. 46 rrIMN 438456

TABLE 2.-Continued Reference Population Told to give tip Criterion for false reports of not smoking % (n/N) False reports Comments Russell et al. (1987) UK smokers attending general practitioners Some groups 7 ppm CO About 22 4-mo to I-yr followup Glasgow et al. US worksite smoking control study Yes 10 ppm CO 0 (0/4) 6-mo followup (1984) Jamrozik, Fowler UK smokers in trial of nicotine gum Yes 12 ppm CO 28.0 (7/25) 6-mo followup et al. (1984)' Clavel et al. French trial of acupuncture and Yes 5 ppm CO 0 (0/24) Sample of study (1985) nicotine gum participants (N=24); Lando and McGovern (1985) Richmond and Webster (1985) Abrams et al. (1987) US subjects undergoing various treatments for eliminating smoking Australian smokers in a general practice; randumized trial of effects of advice to give tip Worksite cessation Yes Test group es CO COf Ib, SCN-, cotinine in plasma, and reports by fumily and friends <9 ppm CO (expired air) 2 cases out of at most 90 5.7 (2/35) I.I (1/9) 1-yr followup Up to 2-mo followup Criteria not stated; 6-mo followup 8-wk followup r~ Glynn, Gruder, Chicago Lung Association Yes 10 ppm CO (expired air) 15.6 (7/45) 4-wk followup ~+ legerski (1986) cessation study w Part Ill. High-risk/medical paticnts - ~ Li et al. (1984) US asbesios-exposed smokers Yes 9 ppm CO I: 22.2 (4/18) 1 I-mo followup V1 receiving (I) behavioral counseling or 2: 23.1 (3/13) W (2) minimal warning 4~- W .

TABLE 1.-Continued Reference Population Part II. Intervention samples Russell et al. (1979) London smokers attending general practices in intervention trial Paxton (1980) Jamrozik, Vessey et aI. (1984)h Russell et al. (1987) b UK smokers assigned to various stop treatments , UK smokers attending general practitioners in trial of various antismoking interventions UK smokers attending general practitioners in trial of effects of brief intervention and support of a smokers' clinic ~ Abrams et al. Smokers/nonsmokers in worksite I~~I (1987) cessation program Stookey et al. (1987) Cessation study Told to give ttp Criterion for false reports of not smoking % (n/N) False reports Yes Salivary nicotine 7.1 ( t/14) Yes Urinary nicotine n=2, N<60 Some groups 1(H) ng/mL urinary cotinine 23.9 (11/46) Some groups 50 pg/L urinary cotinine 38.8 (57/147) Yes 10 ng/mL salivary cotinine 9.1 (1/I I ) Yes IO ng/mL salivary cotinine Nonsmokers 0 (0/20) Former smokers 45.1 (46/102) Comments No cutpoint established; length of followup not stated Study began with 60 subjects; 2 false reports of not smoking detected; cutpoint not established; 6-mo followup If nonparticipants considered as false reports of not.smoking, then 39.7% (23/58)'gave erroneous reports; I-yr followup 1-yr followup Self-reported abstainers; 8-wk followup Length of followup not stated

In this analysis, subjects who had quit smoking were assigned to the duration of abstinence category appropriate for when they enrolled in the study. This method of assignment tends to blunt the rate of decline of mortality risk according to duration of abstinence when compared with never smokers because former smokers do not change categories as duration of abstinence lengthens. No attempt was made in this study to determine smoking status after enrollment, and persons who had quit at enrollment but had resumed smoking were still considered former smokers. Likewise, persons who smoked at enrollment but subsequently quit remain assigned to the current smoker category. This probably leads to some degree of misclassification and affects relative risk estimates (Chapter 2). Like ACS CPS-I and other cohort studies, mortality ratios were substantially lower among former smokers than continuing smokers for all durations of abstinence except that of 1 to 3 years. With the exclusion of those subjects who had a history of cancer, heart disease, or stroke and those who said they were "sick" at the time of recruitment, mortality ratios were lower among former than continuing smokers for all durations of abstinence, among males at all prior levels of cigarette consumption, and among females who smoked fewer than 20 cigarettes per day before they quit. The difference in the pattern of decline in overall mortality between all subjects and the subset of subjects who were healthy at recruitment provides strong evidence that recent quitters disproportionately include those who have quit because they are ill. In contrast with ACS CPS-I, which was conducted in the early 1960s, mortality ratios among both heavy and light smokers in ACS CPS-II remained substantially elevated in comparison with those of never smokers 10 years after quitting. This increase was evident in all subjects and in the subset of subjects who did not have a history of cancer, heart disease, or stroke and who did not state that they were "sick" when recruited. Sixteen years after quitting, the mortality risk among male former smokers of fewer than 21 cigarettes reached that of never smokers but remained elevated among former smokers of 21 cigarettes or more. Among female former smokers in both categories, mortality was comparable with that of never smokers after 16 years of abstinence. The results of ACS CPS-II are broadly in agreement with those of the British Physicians Study (Doll and Peto 1976; Doll and Hill 1964a,b) and the U.S. Veterans Study (Kahn 1966; Rogot and Murray 1980). In both, the overall mortality risk among former smokers remained elevated in comparison with that of never smokers up to 15 years after quitting, although the risk was substantially less than among continuing smokers. An Australian study of petrochemical workers (Christie et al. 1987) appears to differ from the other cohort studies in finding that overall mortality risk among former smokers reached that of never smokers 5 years after quitting. This study is unique in that subjects classified as former smokers were all persistent abstainers. The differences among other studies in estimates of the duration of abstinence needed for a former smoker to have the same overall mortality risk as a never smoker are likely to be due to other smoking-related factors, such as age at smoking initiation, that differ among study populations and over time (Chapter 2). Irrespective of the duration of abstinence needed to reach the mortality risk of never smokers, former smokers have substantially lower mortality when compared with continuing smokers. TIMN 438486 79

mortality risk was still higher than that of never smokers even after 10 years of abstinence. The more recent ACS study, ACS CPS-II, is designed similarly to CPS-I. Re- searchers enlisted 77,000 volunteers, who then solicited their friends, neighbors, and relatives to participate in the study. Those enrolled completed a four-page confidential questionnaire on medical history, health behaviors, medication use, and occupational exposures (Stellman and Garfinkel 1986; Garfinkel and Stellman 1988). A total of 521,555 men and 658,748 women were enrolled; 4-year followup data (1982-86) on the cohort were included in the 1989 Surgeon General's Report (US DHHS 1989). In this Report, mortality rates for all causes of death from the ACS CPS-II were calculated using updated data for the same 4-year followup period (Table 2). Rates were calculated by gender in 5-year age groups for current and former smokers according to level of cigarette consumption (1-20 cig/day, ?21 cig/day for males; 1-19 cig/day, -20 cig/day for females). Rates for former smokers were further stratified by years since smoking cessation (<1, 1-2, 3-5, 6-10, 11-15, and -16). Slightly different strata were used for men and women with respect to daily cigarette consumption in order to provide suitable distributions of subjects across categories of smokers and ex-smokers. y TABLE 2.-Overall mortality ratios among current and former smokers, relative to never smokers, by sex and duration of abstinence at date of enrollment, ACS CPS-II Former smokers ~ Duration of abstinence at enrollment (yr) Current smokers <1 • 1-2 3-5 6-10 11-15 ?16 Males 1-20 ciglday 2.22 2.49 2.38 2.03 1.63 1.38 1.06 >-21 cig/day 2.43 2.77 2.64 2.25 2.04 1.77 1.27 Females 1-19 cig/day 1.60 1.58 1.96 1.41 1.14 1.10 1.01 >30 cig/day 2.10 3.39 2.58 2.03 1.60 1.38 1.15 Former smokers excluding those with cancer, heart disease. or stroke and those "sick" at interview C Duration of abstinence at enrollment (yr) urrent smokers <1 1-2 3-5 6-10 11-15 116 Males 1-20 cig/day 2.34 2.06 2.05 1.89 1.48 1.29 1.01 -21 ciQ/day 2.73 1.85 2.15 1.90 1.77 1.65 1.19 Females l-t9 cig/day 1.82 0.76 1.26 1.42 1.01 1.09 1.00 -20 cig/day 2.46 3.33 2.15 1.44 1.46 1.18 0.95 NOTE: Mortality ratios are relative to those of never smokers. ACS CPS-11=American Cancer Society Cancer Prevention Study lI. SOURCE: Unpublished tabulations, American Cancer Society. 78 TIMN 438485

Cross-sectional studies may be affected by selection bias. Because cigarette smoking is a strong cause of disease and death, groups studied cross-sectionally may not accurately reflect the natural history of smoking, smoking cessation, and the develop- ment of smoking-related illness. The proportion of heavier smokers and more suscep- tible smokers may be reduced compared with the original birth cohorts giving rise to the cross-sectional study population (McLaughlin et al. 1987). Former smokers who stopped because of the development of disease may be underrepresented, whereas those who stopped to reduce the risk of illness may be overrepresented. Information bias is also of potential importance in cross-sectional studies. Pre- existing conditions in survey participants may affect recall of past smoking or may alter the approach used by interviewers to gather smoking information. However, as summarized in Tables I and 2, cross-sectional surveys generally demonstrate low rates of misreporting of smoking status when compared with cotinine and CO levels. As mentioned previously, a single observation on smoking behavior may lead to misclassification of smokers because of the dynamic nature of smoking behavior. Former smokers are typically a heterogeneous group with periods of abstinence ranging from days to years. For example, in the 1986 Adult Use of Tobacco Survey (US DHHS 1989), the subjects' responses were classified in 10 cateQories, 4 of which included former smokers. Of the former smokers, 12.5 percent had quit within the past 3 months, 7.8 percent had quit in the past 3 to 12 months, 22.3 percent had quit in the past I to 5 years, and 57.4 percent had quit 5 or more years earlier. Cohort Studies In a cohort study, the subjects are selected on the basis of exposure status (e.g.,. smoking behavior) and observed for development of disease. Observation may be forward in time (prospective), backward in time (historical or retrospective), or both. Correct conclusions can usually be made about the temporal relation between exposure (smoking cessation) and outcome (reduction of morbidity or mortality). With the cohort design, multiple health outcomes can be considered simultaneously. For ex- ample, the CPS-I and CPS-II conducted by the American Cancer Society (ACS) examined the effect of smoking behavior on total mortality and specific causes of death. In a study of smoking cessation, selection bias could affect the findings of cohort studies if subjects lost to observation were more or less likely to benefit from smoking cessation than subjects remaining under observation (Greenland 1977). For interven- tion studies and cohort studies, the rate of subject loss provides an index of the potential selection bias. In a cohort study of smoking cessation, some misclassification of exposure may be introduced if the classification of smoking status is based on a single assessment. Although the categorization of smoking status may be correct at the time the informa- tion is collected, inevitably some former smokers will resume smoking and some current smokers will stop. The extent of the resulting error will increase with the duration of followup. The resulting misclassification will tend to underestimate the effects of quitting because those who relapse to become current smokers would not be expected to experience beneficial effects attributable to quitting. 48 TIMN 438458

confounding bias. It also considers the potentially complex problem of analyzing data on the effects of smoking cessation. Statistical Considerations Statistical significance testing addresses the likelihood that an observed association has occurred by chance if, in fact, exposure and disease are unassociated (the null hypothesis). By convention, probability (p) values less than 0.05 are generally accepted as "statistically significant"; that is, chance is considered an unlikely explanation for the association. For example, if the p value is less than 0.05, the probability that chance explains the association is less than 5 percent. Confidence intervals describe the range of effects compatible with the data at some specified level of probability, for example 95 percent. • Some studies find associations that do not attain statistical significance. "Negative" investigations must be interpreted in the context of an investigation's sample size; a small sample size may not provide sufficient information to test associations in the range of interest. Such small sample sizes often provide inadequate statistical power to test for the anticipated effects of smoking cessation, and such studies are uninforma- tive as a result. In interpreting associations not achieving statistical sisnificance, confidence limits describe the range of effect compatible with the data. ~ Bias In any epidemiologic study, associations may be affected by bias. Biases from misclassification and from confounding need to be considered in interpreting the findings of studies of the consequences of smoking cessation. This Section focuses on the effects of these biases in studies of smoking cessation. CategorizinQ the dynamic process of smoking cessation poses a substantial challenge toepidemiologic researchers (Chapter?, Part I). Moreover. subjects may not accurately report their own smoking behavior, and reliance on surrogate sources of information on smoking, as may be necessary in case-control studies, may also introduce error. The consequences of misclassification in observation studies have received substan- tial consideration in the epidemiologic literature (Copeland et al. 1977; Greenland 1980: Fleiss 1981; Kleinbaum, Kupper. Morgenstern 1982: SchiesseIman 1982; Rothman 1986). Misclassification can occur in classifying either exposure or outcome. Only exposure misclassification, that is smoking status, will be considered in this Section (Chapter 2, Part I). ~ Misclassification may be classified as nondifferential (or random) or as differential: both types of misclassification are potentially relevant to studies of smoking cessation. Nondifferential misclassification occurs randomly in relation to disease or outcome status, whereas differential misclassification affects exposure information in a pattern that varies with outcome status. For example, differential misclassification would occur in a case-control study of lung cancer if cases tended to minimize the extent of past smoking in comparison with the information given by controls; elderly cases and 52 TIMN 438462

For example, in ACS CPS-I involving nearly 1 million people, Hammond and Garfinkel (1969) studied changes in smoking status over a 2-year period. Male fotmer cigarette smokers in 1959-60 who reported that they were smoking in 1961-62 varied according to duration of prolonged abstinence reported in the 1959-60 survey. For respondents abstinent less than I year in 1959-60, 37.3 percent reported smoking 2 years later; of those reporting abstinence for 1 to 2 years, 19.1 percent were smoking 2 years later, and of those reporting abstinence of more than 2 years, 4.6 percent were smoking 2 years later. For all males who were former smokers in 1959-60, 11.3 percent reported smoking 2 years later. For all female former smokers in 1959-60, 6 percent reported smoking 2 years later. In the U.S. Veterans Study (Rogot and Murray 1980; Kahn 1966), male veterans in a cohort of 248,846 were classified based on responses to questionnaires administered in 1954 or in 1957 (if the 1954 questionnaire was not returned) and then followed for 16 years to determine the relationship between tobacco use and mortality. Undoubtedly, many of the original current smokers became former smokers as a result of the strong trend of smoking cessation among U.S. males during the followup period (US DHHS' 1989). Repeated assessment of smoking status in a cohort study can mitigate misclassifica- tion due to chanaes in smokina status over time (Chapter 2, Part I). Repeated measures are often feasibly made in cohort studies to minimize the effects of misclassitication. Alternatively, validation substudies can be conducted within the cohort to quantify misclassification errors (Greenland 1988). Case-Control Studies Case-control studies involve selection of study subjects based on the presence (cases) or absence (controls) of a disease. Exposure and other attributes of cases and controls (e.g., smoking status or lifetime cigarette consumption) are then measured. The groups are compared with respect to the proportion having the attribute of interest to calculate the exposure odds ratio, which estimates the relative risk associated with exposure. Case-control studies can generally be conducted in less time than cohort studies or intervention studies and are less expensive to perform. Case-control studies are well suited for evaluation of diseases with low incidence rates. Case-control analyses may be affected by information bias and selection bias. Case-control studies are prone to information bias if lifetime exposure histories are collected by interview (Schlesselman 1982). Retrospective lifetime histories of smok- ing or other exposures obtained from ill or elderly subjects may introduce misclassifica- tion. Similarly, studies that rely on reports from surrogates to assess smoking may misclassify exposure. If individuals classified as cases recall more accurately or less accurately than those classified as controls, differential misclassification results (Gordis 1982). Differential misclassification may also be introduced ifrespondents deliberately falsify answers or if interviewers differentially gather information from cases and controls (interviewer bias); interviewers not blinded to case-control status may probe more intensely for a putative causal exposure in cases than in controls (Sackett 1979). Blinding is often not feasible, and meticulous attention must be directed to training interviewers and to designing questionnaires to remove the possibility of interviewer 49 TIMN 438459

Data from case-control studies can be used for the same analytic objectives. Infor- mation on age at starting to smoke, duration of smoking, duration of abstinence, and number of cigarettes smoked can be obtained retrospectively. Conventional analytic methods enable calculation of odds ratios by time since quitting, which estimate the ratios of incidence rates; the reference group for former smokers can be either never smokers or current smokers. Risk of disease for former smokers changes because exposure to active smoking ceases; for some diseases, the exposure of interest in assessing the health consequences of cessation is the subsequent tobacco exposure experienced by continuing users but avoided by former smokers. Some analytic methods may not address adequately this avoided exposure. For example, using variables for cumulative exposure combines the additional exposure for the continuing smoker with the consumption to the point of cessation for the abstinent smoker. If repair processes affect disease risk after cessation, then the interval of abstinence is also a relevant exposure parameter. Thus, reaardless of the type of data analyzed, the method of analysis should properly represent the underlying biologic process. SUMMARY Correct classification of smoking status is important to determine accurately the effects of cessation. Smoking cessation is a dynamic process in which smokers progress through a series of stages in an effort to quit smoking. These stages have been labeled differently by various investigatprs. The model generating the most research refers to the stages as precontemplation, contemplation, action, and maintenance and/or relapse. Very few smokers progress through these stages linearly, because most smokers relapse and recycle through .the staaes three or four times before attainina long-term main- tenance. Four common types of studies for assessing the health consequences of smoking cessation are vulnerable to various sources of information bias leading to misclassitica- tion of smoking status. Cross-sectional surveys have a relatively low frequency of misreporting; however, recall of duration of abstinence is vulnerable to error. A case-control study, because of its retrospective nature, is possibly more likely to have misreporting of smoking status in diseased cases than in nondiseased controls. Cohort studies are likely to have low rates of misreporting of initial smoking status but hiah rates of misclassification due to chan-es in smoking status over time. Clinical trials are likely to have high rates of misreporting for subjects receiving intensive clinical interventions. However, such trials should have relatively little misclassification of smoking status over time and provide more accurate assessment of duration of abstinence when regular followups are maintained. Ylisclassification~of smokers as former smokers will have the effect of under- estimating the benefits of smoking cessation when a true effect exists. The extent of the bias is proportional to the degree of misclassification. Any specificity added to measurement by validation measures will diminish the misclassification bias. TIMN 438467 57

bias. Although selection bias may affect any case-control study that is not population- based, it is unlikely to be of particular importance in most case-control studies of smoking cessation. Intervention Trials Intervention trials are designed to test a hypothesized cause-effect relationship or the benefits of a preventive program by modifying the putative causal or preventive factor and measuring the effect on relevant outcome measures. Intervention trials may be directed at individuals or groups, such as communities. Regardless of the unit of observation, the trials may be conducted with (e.g., a clinical trial) or without ran- domization to the intervention. Clinical trials are most commonly used to assess therapeutic interventions, but this design has also been used to evaluate preventive interventions, such as smoking cessation. A clinical trial includes one or more comparison groups in which subjects receive the control intervention; subjects are randomly assigned to the treatment and comparison groups to ensure that the groups are comparable with respect to charac- teristics potentially affecting the outcomes of interest. Individuals or groups such as communities can be the units of randomization. Within the limits of chance, random assignment makes the intervention and control groups similar at the onset of study. Although widely used to test smoking cessation methods, clinical trials have been used infrequently to assess the health benefits of smoking cessation. In comparison with observation studies, the clinical trial desiQn offers the potential for eliminating or more tightly controllinQ.bias from the selection of subjects and from confounding. However, for many health outcomes, both a large sample size and a lenothy followup period may be needed to have sufficient statistical power. Moreover,~in a study of smokinQ cessation, the power of the trial also depends on the extent of the reduction in smoking in the intervention group, in comparison with the control group. In the reported smoking intervention trials, only a minority of participants attained continuous or prolonged abstinence following most cessation interventions (Hunt, Barnett, Branch 1971, Hunt and Bespalec 1973; Ockene et al. 1990). Even with intensive, prolonged interventions, as in MRFIT, only 42 percent of smokers within the special intervention group were not smoking at 6-year followup, and only 26 percent of baseline smokers had been continuously abstinent from cigarettes over this prolonged period (Ockene et al. 1990). Only a few clinical trials provide information relevant to the health benefits of cessation (Chapter 3). In the Whitehall Civil Servants Study (Rose et al. 1982), the investigators randomly intervened in smoking with advice from a physician in a group of men at hish risk for cardiopulmonary disease. In MRFIT, smoking intervention was one component of the risk factor intervention program directed at the special interven- tion group (MRFIT Research Group 1982). In most clinical trials that assess the effect of cessation on disease outcomes, such as the Whitehall Civil Servants Study (Rose et al. 1982), the investigators did not monitor longitudinally the persistence of quitting or levels of biochemical markers. The only clinical trial that has provided these measures is MRFIT (Ockene et al. 1990). Although 50 TIMN 438460

CONCLUSIONS 1. Most former smokers have cycled several times through the process of smoking cessation and relapse before attaining long-term abstinence. Any static measure of smoking status is thus a simplification of a dynamic process. 2. In studies of the health effects of smoking cessation, persons classified as former smokers may include some current smokers. Consequently, the health benefits of smoking cessation are likely to be underestimated. 3. In contexts other than intervention trials, self-reported smoking status at the time of measurement and concurrent biochemical assessment are highly concordant. This high concordance supports self-report as a valid measure of. smoking status in observational studies of the health effects of smoking cessation. 58 TIl'IN 438468

TAI31.E 1.-Summary of longitudinal studies of overall mortality ratios relative to never smokers among male current and former smokers according to duration of abstinence (when reported) Current Former smokers Study smokers Duration of abstinence (yr) All 1-4 5-9 10-15 > 15 British Physicians" (Doll and Peto 1976) 1.8 1.5 1.5 1.3 1.1 ACS CPS-Ih 1-I9 cig/day 1.72 1.44 1.34 1.01 (Hammond 1966) 2(T`39cig/day 1.92 1.96 1.48 1.31 U.S. Veterans` I(>-20 cig/day 1.82 1.87 1.24 1.47 (Kahn 1966) 21-39cig/day 2.04 2.08 1.88 1.22 Swedish study I-7 g/day`t 1.21 1.08 (Carstensen, Pershagen, 8-15 g/day`t 1.35 Eklund 1987) >15 g/day`~ 1.70 <5 ?5 Australian I--19 cig/dny 1.45 1.60 0.93 petrochemical 20-29 cig/day 2.09 1.55 0.90 workers` (Christie et al. >_30 cig/day 2.10 1.58 0.92 192i7) Framingham' (Gordon, Kannel, McGee 1974) I _47 0.84

TABLE 6.-Continued Results Ilealth status Reference Population measure Current smokers Former smokers Never smoker Cig/day Quit Quit II-t5 21-25 >31 <_Iyr >tyr Segovia, Telephone survey of Setf-report of "good health" 4.18`° 2.00'' 1.46e 3.42e 5.13e 6.14e Bartlett, representative sample Edwards US adults (1989) Gallop Workers in the Work absences 1.25( 1.09( I OOc (1989) pulp/paper industry °Referrent. bRatio compared wilh current smokers. `Mean ridit score adjusted for age and sex. dOdds ratio compared with never smokers and adjusted for age, sex, und socirreconomic status. `Log odds of self-report of good health. 'Ratio of absences compared with never smokers.

TABLE 6.-Relation of smoking cessation to various measures of general health status Reference Population US DI II IS Representative sample of (1985) US population Reed 450 employees offered (1983) subscription to an I1MO Balarajan, Household survey of Yuen, Bewley residents of Great Britain (1985) Seidell et al. 1,245 persons in a (1986) morbidity registry Results liealth status measure Current smokers Former smokers Never smokers Days of work lost due to illness Females -20 yr I.O0" 0.826 0.8e 20-44 yr I.00" 0.79 0.79 45-64 yr I.(Hr 0.91 1.00 Males ?20 yr I .00" 1.03b 0.79b 20-44 yr I.(x)p 0.92 0.86 45-64 yr 1.00" 1.05 0.66 General physical 0.50` 0.52` 0.49c health status Cig/day Quit Quit 1-9 1O-19 ?20 zl yr <1 yr Self-report of illness and Chronic illness 1.07`t 1.31`t 1.76`t 1.43d 1.26' 1.0" physician visits . Acute illness 1.03 1.09 1.29 1.11 1.48 1.0° Outpatient visit 1.46 1.46 1.43 1.40 1.25 10 Physician 1.12 1.08 1.09 1.19 1.47 1.0" consultation Cig/day <1O ?10 Number of health Females 9.6 11.6 10.2 9.0 complaints Males 9.0 9.6 6.8 7.3

CONCLUSIONS 1. Former smokers live longer than continuing smokers, and the benefits of quitting extend to those who quit at older ages. For example, persons who quit smoking before age 50 have one-half the risk of dying in the next 15 years compared with continuing smokers. 2. Smoking cessation at all ages reduces the risk of premature death. 3. Among former smokers, the decline in risk of death compared with continuing smokers begins shortly after quitting and continues for at least 10 to 15 years. After 10 to 15 years of abstinence, risk of all-cause mortality returns nearly to that of persons who never smoked. 4. Former smokers have better health status than current smokers as measured in a variety of ways, including days of illness, number of health complaints, and self-reported health status. 92 TI?qN 438499

For three representative age groups (50-54, 60-64, and 70-74 yr), Figure I shows the relative risk of death among current and former smokers compared with never smokers based on recent ACS CPS-1I data for the subjects who did not have cancer, heart disease, or stroke and were not "sick" at recruitment. Complete data from ACS CPS-II on mortality in current, former, and never smokers aged 50-74 years are presented in Table 7 of the Chapter Appendix. Data are not presented for those aged less than 45 years and greater than 80 years because there were fewer than 10 deaths in almost all of the categories of former smokers. In each of the age subgroups shown in Figure 1, among both sexes and among former light and heavy smokers, mortality risk relative to continuing smokers decreased with increasing duration of abstinence. Using a method described by Kleinbaum, Kupper, and Morgenstern (1982), the data from ACS CPS-II were also used to estimate the effects of quitting at various ages on the cumulative risk of total mortality in a fixed interval after cessation. Several assumptions have been made in conjunction with CPS-II age-specific mortality data in order to estimate as many as 16.5 years' risk of death from all causes for individuals who continue to smoke and those who stop smoking. The first assumption is that age-specific mortality rates measured from 1982-86 CPS-II data remain constant for the next 16.5 years. The f rst category of smoking cessation is 1-2 years: that is, the individual gave up smoking 1 to 2 years ago. It is assumed that, on average, respondents in the 1-2-year category gave up smoking 1.5 years ago. Similarly, for the cessation categories 3-5, 6-10, and 1 1-15 years, the average durations of abstinence are 4, 8, and 13 years, respectively. It is further assumed that respondents are exposed to the age-specific mortality rates of the age interval in which quitting occurs for 1.5 yeafs and to each of the next three age intervals for 5 years each, making a total of '16.5 years. For example, a quitter of the 40-44-year interval would be exposed to the age-specific mortality rates of the 40-44-year-olds for 1.5 years. to those of 45-49-year-olds for 5 years, to those of 50-54-year-olds for 5 years, and to 55-59-year-olds for 5 years. The results of this analysis, presented in Table 3 and in greater detail in Table 8 of the Chapter Appendix, show that the benefits of cessation for total mortality extend to quitting at older ages. For example. a healthy man aged 60-64 years who smokes 21 cigarettes or more per day is estimated to have a chance of dying in the next 16.5 years of 56 percent if he continues to smoke and 51 percent if he quits. Quitting smoking at younger ages confers even greater proportionate increases in survival (see Figure 2 of the Chapter Appendix). Framingham investigators recently analyzed data from their cohort (D'Agostino et al. 1989) and also found that the benefits of quitting apply to those who quit at more advanced ages. These researchers estimated that mean additional life expectancy for those who quit at ages 35 to 39 was 5.1 years for males and 3.2 years for females. For those who quit at ages 65 to 69, additional life expectancy was estimated to be 1.3 years for males and 1.0 year for females. As discussed in detail in Chapter 2 and other chapters, smokers differ from non- smokers in a variety of social, behavioral, and psychological characteristics, and successful quitters differ from those who continue to smoke (Rode, Ross. Shephard 1972; Blair et al. 1980; Haines, Imeson, Meade 1980; McManus and Weeks 1982: Billings and Moos 1983; Gottlieb 1983; Brod and Hall 1984; Seltzer and Oechsli 1985; 80 TIMN 438487

former smokers and would need to be strong predictors of mortality. There is no substantial evidence that this is the case. SMOKING CESSATION AND OVERALL MORTALITY IN INTERVENTION STUDIES Five studies, four of which were randomized trials, evaluated overall mortality in relation to interventions that included smoking cessation as a component. The results of these studies are summarized in Table 4. TABLE 4.-Summary of overall mortality ratios in intervention studies in which smoking cessation was a component Study Intervention Subjects (age) Difference in smoking Mortality ratio Whitehall Civil Servants' (Rose et al. 1982) Smoking Males (40-59) -l4%b 0.98 North Karelia Smoking. BP, Both sexes (35-64) Mates -4'7eh 1.00` (males) (Tuomilehto et al. 1986) diet Females-3%s 0.9-)` (females) Oslo' (Hjermann et al. Smoking. BP. Males (40-59) -4 ciJdayh 0.68`t 1981) diet WHO' (WHO European Smoking. BP, diet Males (40-59) --8.9% 0.97`t Collaborative Group 1983) MRFIT' Smokina. BP, Males (35-57) '-l3°l0 1.02d (7 vr) (MRFIT Research Group diet 0.92 (10.5 yr) 1982. 1990) NOTE: BP=blood pressure: WHO=World Health Organization: MRFIT=Multiple Risk Factor Intervention Trial. "Randontized trial. hintervention minus control. `Change in mortality in rest of Finiand/change in mortality in North Karelia. dMortality in intervention/mortality in control. Only one study examined smokinQ intervention alone (Rose and Hamilton 1978: Rose et al. 1982). Of 1,445 male smokers, aged 40 to 59 and at high risk of coronary heart disease (CHD) or chronic bronchitis, 714 were randomly assigned to an interven- tion group and 73 1 to a normal care group. Men in the intervention group were given individual advice to quit smokina, and if interested in quitting, up to four additional visits over 12 months. At the 9-year followup. 55 percent of responders in the intervention reported abstinence compared with 41 percent in the normal care group. After 10 years of followup, there were 123 deaths in the intervention group and 128 in the normal care group. The proportionate difference in total mortality between the intervention group and normal care group (-2 percent) was not statistically significant, but the confidence interval was wide (-22 percent to +23 percent). There were 81 84 TIMN 438491

TABLE 5.-Summary of studies of medical care utilization among smokers and former smokers Reference Population Measure of medical care utilization Ashford 75.500 residents of Physician (1973) Exeter visits, home visits, hospitalization Oakes et at. 2,557 HMO members Physician (1974) in California visits, hospitalization Marsden. Bray. 1985 worldwide Herbold (1988) survey of alcohol and drug use by military personnel Vogt and 2.582 HMO Schweitzer members in OreQon (1985) Newcomb and 654 adults aged 2 i?4. Bentler (1987) in Los Aneeles Physician visits. days hospitalized Results No consistent differences in any measure of utilization between former smokers and current smokers. Male former smokers have more physician visits than current smokers: female former smokers have more physician visits than current smokers. Male former smokers are less likely than current smokers to be hospitalized: hospitalization among female former smokers compared with current smokers varies with age. Physician vtsits Days hospitalizeda Nonsmokers 2.41 0.64 Smokers <_0.5 ppd 2.37 0.82 1 ppd 2.56 0.68 _1.5 ppd 3.16 0.99 Days Former smokers have lower mean hospitalized, number of hospital days than current physician visits smokers after adjustment for age. sex. duration of membership. and alcohol use. Total physician visits are higher among former smokers than current smokers after adjustment for age. sex, duration of membership. and alcohol use. Nishts Adolescent smokins is related to hospitalized, spending more nights in the hospital and physician visits having more physician visits for illness during early adulthood. Freeborn et al. 3l2 adults aged ?65 in Ambulatory Smokers consistently are more often in (1990) an HMO in Oregon care use upper tertile of care utilization. NOTE: ppd=pack.s/day: HMO=Health Maintenance Oreanization. ' Vtean. identify the reasons for quitting, the relation between quitting and health status may be obscured in studies that classify persons as former and current smokers (Chapter?). A few studies differentiate between short-term abstainers (<1 yr) and lonQ-term abstainers (> 1 yr), and these studies are highlighted. 88 TIMN 438495

controls might introduce nondifferential misclassification from errors in recall of past smoking. The consequences of nondifferential and differential misclassification have been addressed in the epidemiologic literature. Bross (1954) is credited with demonstrating that random misclassification in a 2x2 contingency table diminishes an association that exists between two variables; in general for such cross-classified data, nondifferential misclassification of exposure biases toward the null value, indicating no effect of eposure (Rothman 1986). For exposures classified into three or more levels, the consequencs of nondifferential misclassification are not exclusively directed toward reducing the degree of association. Differential misclassification may either strengthen or weaken associations, depending on the direction of the bias in reporting exposure (Kleinbaum, Kupper, Morgenstern 1982; Rothman 1986). The information presented in prior sections of this Chapter describes the directions that bias may take and allows some generalizations. First, some degree of nondifferen- tial misclassification may affect studies of active smoking and of smoking cessation: the extent of misclassification depends on the type of information collected, the choice of respondents (index subject or surrogate), and the health and ase of the respondents. Second, because disease is present at the time of interview,~nondifferential mis- classification is particularly likely to affect exposure information collected in cross- sectional studies and case-control studies, but little empirical evidence is available. Third, because of the dynamic nature of smoking cessation, some current and former smokers will be misclassified in cohort studies and clinical trials unless smoking behaviors are measured with sufficient frequency during followup. For example, MRFIT data illustrate the potential for misclassification of current and former smokers as smokinQ status chanQes over time if smokina status is not lonai- tudinally assessed (Ockene et al. 1990). The usual care group included 4,091 smokers at baseline with 12.7 percent reporting quitting by the first annual followup visit. Of those first-year quitters, only about half or 6.3 percent of all usual care smokers maintained abstinence for the entire 6-year followup period ("continued stoppers"). However in each year of foilowup, additional smokers quit ("new stoppers") at a maximum rate of 7.5 percent between the first and second years, decreasing to the lowest rate of 4.2 percent between the fifth and sixth years. Simultaneously, smokers who quit and relapsed during the trial succeeded in quitting in subsequent followup periods ("recycled stoppers"). Recycled stoppers increased from 5.3 percent of the usual care baseline smokers in the third year to 15.3 percent at the end of the sixth year. By the sixth year of the study. 25.8 percent of the usual care group were classified as former smokers; 6.3 percent stopped during the first year and maintained abstinence for the remaining 6-year followup period: 15.3 percent stopped. relapsed, and stopped again: and 4.2 percent stopped for the first time in the last year of followup. Although the usual care group is not representative of adult male smokers, these data illustrate the dynamics of smoking behavior and the potential for misclassification. Incorrect categorization of some current smokers as former smokers and of some former smokers as current smokers, if nondifferential, would tend to reduce the apparent benefit of smoking cessation, as disease occurrence is reduced in the category of apparent current smokers by the inclusion of former smokers and is increased in the 53 TIMN 438463

nonintervention (i.e., cohort) studies described in the previous Section clearly indicate a benefit of smoking cessation on overall mortality. SMOKING CESSATION AND MEDICAL CARE UTILIZATION Population Projections The relationship between smoking cessation and medical care utilization is a complex issue. Data on differential disease and mortality rates comparing smokers and abstainers are abundant, and many investigators have used these data to project the savings in dollars attributable to smoking cessation (Weinkam. Rosenbaum, Sterling 1987; Leu and Schaub 1983; Luce and Schweitzer 1978; Oster, Colditz, Kelly 1984). Generally, these projections produce results that depend on the many assumptions of the models that create them. For example, Luce and Schweitzer (1978) projected that the total 1976 dollar cost of smokinQ in the United States was about $27.5 billion and that excess medical care costs accounted for about $8.2 billion of those costs. Weinkam, Rosenbaum, and Sterling (1987) and Leu and Schaub (1983). both using population simulation approaches, concluded that smoking does not, over a lifetime, lead to increased medical care utilization. This is because the short-term higher levels of utilization of smokers are approximately balanced by shorter longrevity and the resultino reduced need for medical care. ~ Oster. Colditz. and Kelly (1984) used population projections to estimate the medical care costs of smoking and the proportion of those costs that are potentially recoverable depending on the age at which smoking is Qiven up and the level of smoking prior to quitting. Male light smokers (<I pack/day) who quit between aQes 35 and 39 were estimated to recover about 59 percent of their lifetime excess medical care.costs. Even if quitting was delayed until ages 75 to 79. light smokers were estimated to recover one-third of the costs. For heavy smokers, quitting earlier was estimated to have somewhat more benefit. For both sexes and all levels of smokin2, medical care cost savinQs from smokina cessation were estimated to be substantial. ~ Observational Studies Table 5 summarizes studies that directly measured utilization of medical services by current smokers, former smokers. and never smokers. These studies suggest that smoking is associated with higher utilization of hospital services and that former smokers experienced a brief period of increased utilization of hospital services just after quitting followed by declines in utilization to levels of never smokers. Modest increases in outpatient utilization by smokers are to some degree offset by a decreased propensity to use preventive care services (Marsden, Bray, Herbold 1988; Vogt and Schweitzer 1985; Oakes et al. 1974). ~ SMOKING CESSATION AND HEALTH STATUS Table 6 summarizes studies of smoking cessation and health status. The variety of measures used makes direct comparison across studies problematic. Furthermore, in most cases, only a comparison of measures for never, current, and former smokers is available. Because some smokers quit due to illness and because most studies fail to 87 TIMN 438494

VOGT, T.M., SELVIN, S., WIDDOWSON, G., HULLEY, S.B. Expired air carbon monoxide and serum thiocyanate as object measures of cigarette exposure. American Journal of Public Health 67(6):545-549, June 1977. WALD, N.J., BOREHAM, J., BAILEY, A., RICHIE, C., HADDOW, J.E., KNIGHT, G. Urinary cotinine as marker of breathing other people's tobacco smoke. (Letter.) Lancet I (8370):234-23I, January 28, 1984. WARNER, K.E. Possible increases in the underreporting of cigarette consumption. Journal of the American Statistical Association 73(362):314-318, June 1978. WILCOX, R.G., HUGHES, J., ROLAND, J. Verification of smoking history in patients after infarction using urinary nicotine and cotinine measurements. British Medical Journal 2: I026-1028, October 27, 1979. WILLIAMS, C.L., ENG, A., BOTVIN, G.J., HILL, P., WYNDER, E.L. Validation of students' self-reported cigarette smoking status with plasma cotinine levels. American Journal of Public Health 69(12):1272-1274, December 1979. WINDSOR, R.A., CUTTER, G.. MORRIS, J., REESE, T., MANZELLA, B., BARTLETT, E.E., SAMUELSON, G., SPANOS, P. The effectiveness of smoking cessation methods for smokers in public health maternity clinics: A randomized trial. American Journal of Public Health 75(12):1389-1392, December 1985. WINDSOR, R.A.. ORLEANS, C.T. Guidelines and methodological standards for smoking cessation intervention research among pregnant women: Improving the science and art. Health Education Quarterly 13(2):131-161, Summer 1986. ZEIDENBERG, P., JAFFE, J.H., KANZLER, M., LEVITT, M.D., LANGOIVE, J.J., VAN VANAKIS. H. Nicotine: Cotinine levels in blood during cessation of smoking. Comprehen- sive Psychiatry 18(1):93-101, January-February 1977. TIMN 438479 69

category of apparent former smokers by the inclusion of current smokers. Stratification by the duration of abstinence may provide some control of this type of misclassification. The category of never smokers in an epidemiologic study may include some persons who smoked in the past (Britten 1988; Persson and Norell 1989). In general, former smokers who reported themselves as never smokers consumed fewer cigarettes than those correctly categorizing themselves as former smokers. Nevertheless, the bias resulting from the inclusion of some former smokers in the category of never smokers would tend to reduce the apparent benefit of cessation when former smokers are compared with never smokers. The consequences of misclassification must be considered in the context of the disease under investigation. For example, in studying lung cancer and smoking cessation, the failure of long-term former smokers to report a brief period of relapse has little relevance. In contrast, unreported periods of relapse would be relevant in assessing smoking cessation and occurrence of myocardial infarction or of respiratory symptoms, conditions for which cessation has some short-term benefit. Bias from confounding is also of concern in studies of the health consequences of smoking cessation. Former smokers tend to differ from continuing smokers in the earlier intensity of cigarette smoking and in other aspects of lifestyle that may determine disease risk. Former smokers tend to have smoked fewer cigarettes per day and to have started smoking at an older age than continuing smokers (Friedman et al. 1979; Garvey et al. 1983; Myers et al. 1987; Volume Appendix). Thus, at any age, former smokers have had less cumulative exposure to cigarette smoke, on average. than continuing smokers. Failure to account appropriately for differences in cumulative exposure between former smokers and continuing smokers may exaggerate the benefits of cessation. Misclassification of smoking measures may limit the degree-to which confounding can be controlled (Greenland 1980: Rothman 1986). Other differences between former smokers and current smokers may also influence disease risk. Former smokers are more likely to be of higher socioeconomic status than continuing smokers and tend to follow a healthier lifestyle than persistent smokers (Chapter 11 and Volume Appendix). Former smokers generally drink less alcohol and less coffee, are more physically active, and experience less stress, although their relative body weight tends to be greater (Friedman et al. 1979; Kaprio and Koskenvuo 1988; Chapters 10 and 11). However, some persons may stop smoking because a personal combination of risk factors places them at increased risk for disease. In the British Regional Heart Study, former smokers had higher blood pressure and total serum cholesterol at entry than current or never smokers (Cook et al. 1986). In fact, observed mortality rates for many diseases have been higher for former smokers than current smokers during the first few years following cessation. Persons with symptoms of incipient illness or with newly diagnosed illness may stop smoking (Hammond and Garfinkel 1966). Consequently, mortality rates for former smokers immediately following cessation may exceed those for current smokers. In studies of the effect of cessation on the course of established disease, consideration must be given to the severity of the underlying disease in former smokers and persistent smokers. For example, in a study of mortality following myocardial infarction, persons 54 TIMN 438464

ROSE, G., HAMILTON, P.J. A randomised controlled trial of the effect on middle-aged men of advice to stop smoking. Journal of Epidemiology and Community Health 32(4):275-28 1, December 1978. ROSE, G., HAMILTON, P.J.S., COLWELL, L., SHIPLEY, M.J. A randomised controlled trial of anti-smoking advice: 10-year results. Journal of Epidemiology and Community Health 36(2):102-108, June 1982. ROSEN, T.J., SHIPLEY, R.H. A stage analysis of self-initiated smoking reductions. Addictive Behaviors 8(3):263-272, 1983. ROTHMAN, K.J. Modern Epidemiology. Boston: Little, Brown and Company, 1986. RUSSELL, M. Cigarette consumption and biochemical measures of smoke intake. (Letter.) British Medical Journal 285(6340):507-508, August 14, 1982. RUSSELL, M.A.H., FEYERABEND, C. Blood and urinary nicotine in non-smokers. Lancet 1(7900):179-181, January 25, 1975. RUSSELL, M.A.H., JARVIS, M.J., DEVITT, G., FEYERABEND, C. Nicotine intake by snuff users. British Medical Journal 283(6295):814-817, September 26, 1981. RUSSELL, M.A.H., STAPLETON, J.A., JACKSON, P.H., HAJEK, P., BELCHER, M. District programme to reduce smoking: Effect of clinic supported brief intervention by general practitioners. British Medical Journal 295(6608):1240-1244, November 14, 1987. RUSSELL, M.A.H., WILSON, C., TAYLOR, C., BAKER, C.D. Effect of general practitioners' advice against smoking. British Medical Journa12:231-235, July 28, 1979. RUTH, K., NEATON, J. Evaluation of two biological markers of tobacco exposure used in the MRFIT. Preventive Medicine, in press. SACKETT, D.L. Bias in analytic research. Journal of Chronic Diseases 32(1-2):51-63, 1979. SANDLER, D.P., SHORE, D.L. Quality of data on parents' smoking and drinking provided by adult offspring. American Journal of Epidemiology 124(5):768-778, November 1986. SCHACHTER, S. Recidivism and self-cure of smoking and obesity. American Psychologist 37(4):4364 44. April 1982. SCHLESSELMAN, J.J., STOLLEY, P.D. Case Control Studies: Design. Conduct, Analysis. Monographs in Epidemiology and Biostatistics: No. 2 New York: Oxford University Press, 1982. SEPKOVIC, D.W., HALEY, N.J. Biomedical applications of cotinine quantitation in smoking related research. American Journal of Public Health 75(6):663-665, June 1985. SEPKOVIC, D.W., HALEY, N.J., HOFFMANN, D. Elimination from the body of tobacco products by smokers and passive smokers. (Letter.) Journal of the American Medical Association 256(7):863. August 1986. SHANNON, L, SUDDICK, R., DOWD, F. JR. Saliva: Composition and secretion. In: Meyers, H. (ed.) Monographs in Oral Science, Volume 2. New York: S. Karger, 1974. SHIPLEY, R.H. Maintenance of smoking cessation. Effects of follow-up letters, smoking motivation, muscle tension and health locus of control. Journal of Consulting and Clinical Psychology 49(6):982-984, December 1981. SHUMAKER, S.A., GRUNBERG, N.E. Proceedings of the National Working Conference on Smoking Relapse. Health Psychology 5(Supplement): 1-99, 1986. SILLETT, R.W., WILSON, M.B., MALCQLM, R.E., BALL. K.P. Deception among smokers. British Medical Journal 2:1185-1186, October 28, 1978. SLATTERY, M.L., HUNT, S.C., FRENCH, T.K., FORD, M.H.. WILLIAMS, R.R. Validity of cigarette smoking habits in three epidemiologic studies in Utah. Preventive Medicine 18:1 1- 19, 1989. STEINMAN, G.D. Thiocyanate vs. cotinine as a marker to identify smokers. (Letter.) Clinical Chemistry 31(8.):1406. August 1985. STEW ART, R.D. The effect of carbon monoxide on humans. Annual Review of Pharmacology 15:409-425, 1975. ' 67 TIMN 438477

PERSSON, P.G., NORELL, S.E. Retrospective versus original information on cigarette smok- ing. Implications forepidemiologic studies. American Journal ofEpidemiology 130(4):705- 712, October 1989. PETITTI, D.B., FRIEDMAN, G.D., KAHN, W. Accuracy of information on smoking habits provided on self-administered research questionnaires. American Journal of Public Health 71(3):308-311, March 198 1. PICKLE, L.W., BROWN, L.M., BLOT, W.J. Information available from surrogate respondents in case-control interview studies. American Journal of Epidemiol'ogy I I8(1):98-108, July 1983. PIERCE, J.P., DWYER, T., DIGIUSTO, E., CARPENTER, T., HANNAM, C., AMIN. A., YONG, C., SARFATY, G., SHAW, J., BURKE, N., QUIT FOR LIFE STEERING COM- MITTEE. Cotinine validation of self-reported smoking in commercially run community surveys. Journal of Chronic Diseases 40(7):689-695, 1987. POJER, R., WHITFIELD, J.B., POULOS, V., ECKARD, I.F., RICHMOND, R., HENSLEY, W.J. Carboxyhemoglobin, cotinine, and thiocyanate assay compared for distinguishing smokers from non-smokers. Clinical Chemistry 30(8):1377-1380, 1984. PROCHASKA, J., DICLEMENTE, C. C. The Transtheoretical Approach: Crossing Tradition- al Boundaries of Therapy. Pacific Grove, California: Brooks/Cole Publishing Company, 1984. PROCHASKA. J., VELICER, W., DICLEMENTE, C., GUADAGNOLI, E.. ROSSI, J. Patterns of change: Dynamic typology applied to smoking cessation. Multivariate Behavioral Re- search, in press. PROCHASKA. J.O., DICLEMENTE, C.C. Stages and processes of seif-change of smoking: Toward an integrative model of change. Journal of Consulting and Clinical Psychology 51(3):390-395, 1983. PROCHASKA, J.O., DICLEMENTE. C.C. Toward a comprehensive model of change. In: Miller, W., Heather, N. (eds.) Treating Addictive Behaviors. Processes of change. New York: Plenum Press, 1986, pp. 3-27. PROCHASKA. J.O., DICLEMEivTE. C.C., VELICER, W.F., GINPIL, S., NORCROSS, J.C. Predicting change in smoking status for self-changers. Addictive Behaviors 10:395=~06, 1985. RADFORD, E.P., DRIZD, T.A. Blood carbon monoxide levels in persons 3-74 years of age: United States.1976-80. National Center for Health Statistics. In: Viral and Health Statistics No. 76, DHHS Publication No. (PHS) 82-1250, March 17, 1982. RAW, M., JARVIS, M.J., FEYERABEND, C., RUSSELL, M.A.H. Comparison of nicotine chewing-gum and psychological treatments for dependent smokers. British Medical Journal 281:481-482, 1980. RESEARCH COMMITTEE OF THE BRITISH THORACIC SOCIETY. Comparison of four methods of smoking withdrawal in patients with smoking related diseases. British Medical Journal286(6366):595-597, February 1983. RICHMOND, R., WEBSTER, I. Evaluation of general practitioners' use of a smoking inter- vention programme. International Journal of Epidemiology 14(3):396-401, 1985. ROGOT. E., MURRAY, J.L. Smoking and causes of death among U.S. veterans: 16 years of observation. Public Health Reports 95(3):213-222, May-June 1980. ROGOT, E., REID, D.D. The validity of data from next-of-kin in studies of mortality among migrants. International Journal of Epidemiology 4(1):51-54, 1975. RONAN, G., RUANE, P., GRAHAM, I.M., HICKEY, N., MULCAHY, R. The reliability of smoking history amongst survivors of myocardial infarction. British Journal of Addiction 76:425--428, 1981. 66 . TIMN 438476

CHAPTER 4 - SMOKING CESSATION AND RESPIRATORY CANCERS TIMN 438509 103

LANDO, H.A. A factorial analysis of preparation, aversion, and maintenance in the elimination of smoking. Addictive Behaviors 7:143-154, 1982. LANDO, H.A., MCGOVERN, P.G. Nicotine fading as a nonaversive alternative in a broad- spectrum treatment for eliminating smoking. Addictive Behaviors 10:153-161, 1985. LANGER, P., GREER, M.A. Antithyroid Substances and Naturally Occurring Goitrogens. Basel: S. Karger, 1977. LANGONE, JJ., GJIKA, H.B., VAN VUNAKIS, H. Nicotine and its metabolites. Radio- immunoassays for nicotine and cotinine. Biochemistry 12(24):5025-5030, November 20, 1973. LAST, J.M. (ed.) A Dictionary ofEpidemiology, Second Edition. New York: Oxford University Press, 1988. LEE, P.N. Passive smoking and lung cancer association: A result of bias? Human To.ricology 6:517-524, 1987. LEE, P.N. Misclassification of Smoking Habits and Passive Smoking. A Review of the Evidence. Berlin: Springer-Verlag. 1988. LERCHEN, M.L., SAMET, J.M. An assessment of the validity of questionnaire responses provided by a surviving spouse. American Journal of Epidemiology 123(3):481-489, March 1986. ~ LI, V.C., KIM. Y.J.. EWART, C.K.. TERRY, P.B., CUTHIE, J.C., WOOD, J.. EMMETT, E.A., PERMUTT, S. Effects of physician counseling on the smoking and behavior of asbestos- exposed workers. Preventive Medicine 13(5):462-476, September 1984. LICHTENSTEIN, E., BROWN, R.A. Smoking cessation methods: Review and recommenda- tions. In: Miller, W.R. (ed.) The Addictive Behaviors: Treatment of Alc•ohnlism. Drug Abuse. Smoking and Obesity. New York: Pergamon Press, 1980. LIPINSKI, D., BLACK. J.L.. NELSON, R.O.. CIMINERO, A.R. Influence of motivational variables on the reactivity and reliability of self-recording. Journal ofConsultinQ and Clinical Psychrilo,Qy 43(5):637-646, October 1973. LUEPKER. R.V., PALLONEN, U.E., MURRAY, D.M.. PIRIE, P.L. Validity of telephone surveys in assessing cigarette smoking in young adults. American Journal of Public Health 79(2):202-204. February 1989. MALCOLM, R.E.. SILLET, R.W.. TURNER. J.A.M.C.M., BALL, K.P. The use of nicotine chewing gum as an aid to stopping smoking. Psychopharmacology 70(3):295-296, 1980.. MARLATT, G.A.. CURRY. S., GORDON, J.R. A longitudinal analysis of unaided smoking cessation. Journal of Consulting and Clinical Psychology 56(5):715-720, October 1988. MARLATT. G.A.. GORDON. J.R. Relapse Prevention: Maintenance Strategies in the Treat- ment of Addictive Behaviors. New York: Guilford Press, 1985. MARSH. G.M., SACHS. D.P.L., CALLAHAN, C., LEVITON. L.C., RICCI. E.. HENDER- SON, V. Direct methods of obtaining information on cigarette smoking in occupational studies. American Journal of Industrial Medicine 13:71-103,1988. MCFALL, R.M. Smoking-cessation research. Journal of Consulting and Clinical Psychology 46(4):703-712, August 1978. MCLAUGHLIN. J.K., DIETZ, M.S.. MEHL, E.S., BLOT, W.J. Reliability of surrogate information on cigarette smoking by type of informant. American Journal of Epidemiology 126(1):144-146, July 1987. MULTIPLE RISK FACTOR INTERVENTION TRIAL RESEARCH GROUP. Multiple Risk Factor Intervention Trial. Risk factor changes and mortality results. Journal of the American Medical Association 248(12):1465-1477, September 24, 1982. MURRAY, D.M., O'CONNELL. C.M., SCHMID, L.A., PERRY, C.L. The validity of smoking self-reports by adolescents: A reexamination of the bogus pipeline procedure. Addictive Behaviors 12(t):7-15, 1987. 64 TIMN 438474

HALEY, N.J., HOFFMANN, D. Analysis for nicotine and cotinine in hair to determine cigarette smoker status. Clinical Chemistry 31(10):1598-1600, October 1985. HALEY, N.J., SEPKOVIC, D.W., LOUIS, E., HOFFMANN, D. Absorption and elimination of nicotine by smokers, nonsmokers, and chewers of nicotine gum. In: Rand, R.J., Thurau, K. (eds.) The Pharmacology of Nicotine, ICSU Symposium Series 9, Washington, DC: IRL Press, 1987, pp. 20-21. HALEY, N.J., SEPKOVIC, D.W., HOFFMANN, D. Elimination of cotinine from body fluids: Disposition in smokers and nonsmokers. American Journal of Public Health 79(8):1046- 1048, August 1989. HAMMOND, E.C., GARFINKEL, L. The influence of health on smoking habits. In: Haenszel, W. (ed.) Epidemiological Approaches to the Study of Cancer and Other Chronic Diseases. NCI Monograph No. 19. U.S. Department of Health, Education, and Welfare, Public Health Service, National Cancer Institute, January 1966, pp. 269-285. HAMMOND, E.C., GARFINKEL, L. Coronary heart disease, stroke and aortic aneurysm. Factors in etiology. Archives of Environmental Health 19:167-182, August 1969. HANSEN, W.B., MALOTTE, C.K.. FIELDING, J.E. The bogus pipeline revisited: The use of the threat of detection as a means of increasing self-reports of tobacco use. Journal ofApplied Psychology 70(4):789-792, November 1985. HATZIANDREU, E.J.. PIERCE, J.P., FIORE. M.C., GRISE, V., NOVOTNY, T.E.. DAVIS, R.M. The reliability of self-reported cigarette consumption in the United States. American Journal of Public- Health 79(8):1020-1023, August 1989. HERMANSON, B., OMENN, G.S., KRONMAL. R.A., GERSH, B.J. Beneficial six-year outcome of smoking cessation in oldermen and women with coronary artery disease. Results from the CASS Registry. New EnglandJournal of Medicine 319(21):1365-1369. November 24,1988. HORAN, J.J., HACKETT, G., LINBERG, S.E. Factors to consider when using expired air carbon monoxide in smoking assessment. Addictive Behaviors 3(1):25-28. 1978. HORN, D. A model for the study of personal choice health behaviour. International Journal of Health Education 19:89-98, 1976. HUGHES, G.H., HYMOWITZ, N., OCKENE. J.K.. SIMON, N., VOGT. T.M. The Multiple Risk Factor Intervention Trial (MRFIT). V. Intervention on smoking. Preventive Medicine 10(4):476-500, July 1981. HUGHES, J., FREDERIKSEN, L.. FRAZIER. M. A carbon monoxide analyzer for measure- ment of smoking behavior. Behavior Therapy 9:293-296. 1976. HUMBLE, C.G., SAMET, J.M., SKIPPER, B.E. Comparison of self- and surrogate-reported dietary information. American Journal of Epidemiology 119(1):86-98. 1984. HUNT, W.A., BARNETT, L.W.. BRANCH. L.G. Relapse rates in addiction programs. Journal of Clinical Psychology 27(4):455-456, October 1971. HUNT, W.A., BESPALEC, D.A. An evaluation of current methods of modifying smoking behavior. Journal of Clinical Psychology 30:431--138. 1973. HUNTER, S.M., WEBBER, L.S., BERENSON. G.S. Cigarette smoking and tobacco usage behavior in children and adolescents: Bogalusa Heart Study. Preventive Medicine 9(6):701- 712, November 1980. ISACSSON, S.-O.. JANZON, L. Results of a quit-smoking research project in a randomly selected population. Scandinavian Journal of Social Medicine 4:25-29, 1976. JAMROZIK. K., FOWLER, G., VESSEY, M., WALD, N. Placebo controlled trial of nicotine chewing Qum in general practice. British Medical Journa1289(6448):794-797, September 29, 1984. JAMROZIK, K., VESSEY, M., FOWLER, G., WALD, N., PARKER. G., VAN VUNAKIS, H. Controlled trial of three different antismoking interventions in general practice. British Medical Journal 288(6429):1499-1503, May 19, 1984. 62 TIMN 438472

TABLE 3.-Examples of potential methodologic problems in investigating the health consequences of smoking cessation Problem Current smokers developing symptoms of disease quit smoking Consequences Apparent benefits of cessation are reduced Self-reported former smokers are actually smoking (information bias) Former smokers tend to have smoked less than persistent smokers (confounding bias) Former smokers tend to have a healthier lifestyle than persistent smokers (confounding bias) . Smoking practices and the presence of smoking-related diseases affect participation in studies (selection bias) Small number of subjects in a study Apparent benefits of cessation are reduced Failure to account for the difference may exaggerate the apparent benefits of cessation Failure to account for the difference may exaggerate the apparent benefits of cessation Apparent benefits of cessation may be increased or decreased A beneficial effect of cessation may not reach statistical significance Ecologic Studies Ecologic studies represent a descriptive approach for examining the relation between risk factors and disease. Groups, rather than individuals, are the unit of analysis in ecologic studies. For example, changes,in lung cancer mortality rates for selected countries have been examined for correlation with changes in measures of smoking for those countries, such as the percentage of smokers or per capita cigarette consumption (US PHS 1964; Cairns 1975; Cummings 1984; Doll and Peto 1981). Ecologic studies often have the advantage of being performed inexpensively and feasibly by using already available data. This design has well-described limitations related to the estimation of exposure and control of confounding, and may yield seriously biased data on exposure-disease relationships (Kleinbaum, Kupper, Morgenstern 1982; Rothman 1986). Cross-Sectional Studies In a cross-sectional or prevalence study, exposure and outcome are assessed at the same point in time among individuals in a population. Because cross-sectional studies measure exposure and outcome variables simultaneously, the true temporal relation between exposure and disease may be obscured (Rothman 1986). However, cross- sectional studies can be readily performed and have supplied much of the evidence on smoking cessation and nonmalignant respiratory diseases (Chapter 7). TIMN 438457 47

INTRODUCTION The overall risk of mortality among smokers has been discussed in several prior reports of the Surgeon General (US PHS 1964, 1969; US DHEW 1979; US DHHS 1989). The 1989 Report estimated that approximately 390.000 Americans died in 1985 from diseases attributable to smoking (US DHHS 1989). Another source (Mattson, Pollack, Cullen 1987) estimated that 36 percent of heavy smokers aged 35 will die before age 85, and 28 percent before age 75, from a disease caused by smoking. Prior reports of the Surgeon General (US PHS 1968; US DHEW 1979; US DHHS 1989) have reviewed the association of smoking with overall morbidity, concluding that overall morbidity is increased among smokers. Quantitative estimates of the amount of morbidity attributable to smoking vary because of differences in the measures of morbidity used. Data from the aggregate of studies of overall mortality and morbidity among smokers and former smokers show that smoking causes increased risk of morbidity and mor- tality. However, the temporal pattern of the reduced all-cause mortality after quitting and the effects on mortality risk of quitting at various ages have not been ful ly described. In addition, questions about the benefits of smoking cessation for mortality have arisen because of the results of studies involving interventions to promote'smokinQ cessation. The association of smokina with medical care utilization is a topic that has not been addressed in detail in previous reports of the Surgeon General. This Chapter reviews studies of overall mortality among former smokers, with particular attention to the temporal pattern of decline in mortality after quitting and the association of age at quitting with decline in mortality. Overall mortality in intervention studies that include smoking cessation is discussed with attention to problems of inferring the benefits of smoking cessation forthe individual from these studies. Studies of medical care utilization by and health status of former smokers are described. SMOKING CESSATION AND OVERALL MORTALITY IN COHORT STUDIES Table 1 summarizes the results of major cohort studies comparing overall mortality among never, current, and former smokers. The studies consistently showed a substan- tially lower risk of mortality among former smokers in comparison with continuing smokers. Compared with continuing smokers, former smokers had a progressive decline in mortality risk as duration of abstinence increased, although risk in some studies was increased for 1 to 3 years after cessation, almost certainly because some people quit due to ill health (Chapter 2). The durations of abstinence required for former smokers to reach the mortality risk of never smokers differ among studies. The American Cancer Society (ACS) study of 1 million American volunteers (Hammond 1966), also known as the 25-State Study and as the Cancer Prevention Study I (ACS CPS-I), found that after 10 years, mortality rates among former smokers of fewer than 20 cigarettes per day reached levels equivalent to those of never smokers. Among former smokers of 20 cigarettes or more per day, TIMN 438482 75

JANZON, L., LINDELL, S.-E., TRELL. E., LARME, P. Smoking habits and carboxy- haemoglobin. A cross-sectional study of an urban population of middle-aged men. Journal of Epidemiology and Community Health 35(4):271-273, December 1981. JARVIS, M., TUNSTALL-PEDOE, H., FEYERABEND, C., VESEY, C., SALLOOJEE, Y. Biochemical markers of smoke absorption and self reported exposure to passive smoking. Journal of Epidemiology and Community Health 38(4):335-339, December 1984. JARVIS, M., TUNSTALL-PEDOE, H., FEYERABEND, C., VESEY, C., SALOOJEE, Y. Comparison of tests used to distinguish smokers from nonsmokers. American Journal of Puhlic Health 77(11):1435-1438, November 1987. JARVIS, M.J., RAW, M., RUSSELL, M.A.H., FEYERABEND, C. Randomised controlled trial of nicotine chewing-gum. British Medical Journal 285(6341):537-540, August 21, 1982. JARVIS, M.J., RUSSELL, M.A.H., FEYERABEND. C., EISER, J.R., MORGAN, M. Passive exposure to tobacco smoke: Saliva cotinine concentrations in a representative population sample of nonsmoking schoolchildren. British MedicalJourna1291(6500):927-929. October 5, 1985. JONES, E.E., SIGALL, H. The bogus pipeline: A new paradigm for measuring affect and attitude. Psychological Bulletin 76:349-364,1971. JONES, R.D., COMMINS, B.T., CERNIK. A.A. Blood lead and carboxyhaemoglobin levels in London taxi drivers. Lancet 2:302-303, August 12, 1972. KAHN, A., RUTLEDGE. R.B., DAVIS. G.L., ALTES. J.A., GANTNER. G.E.. THORNTON. C.A.. WALLACE. N.D. Carboxyhemoglobin sources in the metropolitan St. Louis popula- tion. Archives of Environmental Health 29(3 )): 127-135September 1974. KAHN, H.A. The Dom study of smoking and mortality among U.S. veterans: Report on eight and one-half years of observations. In: Haenszel, W. (ed.) Epidemiological Approaches to the Study of Cancer and Other Chronic• Diseases. NCI Monograph No. 19. U.S. Department of Health, Education, and Welfare, Public Health Service. National Cancer Institute. January 1966, pp. 1-125. KAPRIO, J., KOSKENVUO, M. A prospective study of psychological and socioeconomic characteristics, health behavior and morbidity in cigarette smokers prior to quitting compared to persistent smokers and non-smokers. Journal of Clinical Epidemiology 4l(2):139-150, 1988. KIRSCHT, J.P., JANZ. N.K., BECKER. M.H., ERAKER. S.A., BILLI. J.E.. WOOL- LISCROFT, J.O. Beliefs about control of smoking and smoking behavior: A comparison of different measures in different groups. Addicrire Behaviors 12.(2):205-208, 1987. KLEINBAUM, D.G., KUPPER, L.L.. MORGENSTERN, H. Epidemiologic Research: Prin- ciplesand QuantitativeMethnds. Belmont, California: Lifetime Learning Publications. 1982. KNIGHT, G.J., WYLIE, P., HOLMAN. M.S.. HADDOW, J.E. Improved radioimmunoassay for cotinine by selective removal of bridge antibodies. Clinical Chemistry 31(1):118-121, January 1985. ~ KORNITZER, M., VANHEMELDONCK, A., BOURDOUX, P.. DE BACKER. G. Belgian heart disease prevention project: Comparison of self-reported smoking behaviour with serum thiocyanate concentrations. Journal nfEpidemiolvgyand Community Health 37(2):132-L36, June 1983. KOZLOWSKI, L.T. Pack size, reported smoking rates and public health. AmericanJournal of Public Health 76(1 1):1337-1338. November 1986. KRALL. E.A.. VALADIAN, I., DWYER. J.T., GARDNER. J. Accuracy of recalled smoking data. American Journal of Public Health 79(2):200-206, February 1989. LANDO, H.A. Effects of preparation, experimenter contact, and a maintained reduction alternative on a broad-spectrum program for eliminating smoking. Addictive Behaviors 6:123-133, 1981. 63 TIMN 438473

References ABRAMS, D.B., FOLLIC, M.J., BIENER, L., CAREY, K.B., HITTI, J. Saliva cotinine as a measure of smoking status in field settings. American Journal of Public Health 77(7):846- 848, July 1987. ABRAMS, D.B., WILSON, G.T. Self-monitoring and reactivity in the modification of cigarette smoking. Journal of Consulting and Clinical Psychology 47(2):243-25 i,1979. ASSAF, A.R., MCKENNEY, J.L., BANSPACH, S.W., CARLETON, R.A. Validation of self-reported smoking practices. Paper presented at the 10th Annual Convention of the Society of Behavioral Medicine, April 1989. AUGER. T.J., WRIGHT, E. JR., SIMPSON, R.H. Posters as smoking deterrents. Journal of Applied Psychology 56(2):169-171, April 1972. BAUMAN, K.E., KOCH, G.G., BRYAN, E.S. Validity of self-reports of adolescent cigarette smoking. International Journal of the Addictions 17(7):113 I-t 136, 1982. BENOWITZ, N.L. The use of biologic fluid samples in as"sessing tobacco smoke consumption. In: Grubowski, J., Bell, C.S. (eds.) Measurement in the Analysis and Treatment of Smoking Behavior. NIDA Research Monograph 48. U.S. Department of Health and Human Services, Public Health Service, Alcohol, Drug Abuse, and Mental Health Administration, National Institute on Drug Abuse. 1983, pp. 6-26. BIGLAN, A., GALLISON, C., ARY, D., THOMPSON, R. Expired air carbon monoxide and saliva thiocyanate: Relationships to self-reports of marijuana and cigarette smokino. Addic- tive Behaviors 10(2):137-144, 1985. ~ BLISS. R.E.. O'CONNELL, K.A. Problems with thiocyanate as an index of smokinQ status: A c.cal review with suggestions for improving the usefulness of biochemical measures in sinokin- cessation research: Health Pstichologr 3(6):563-581, 1984. BLOT, W.J., AKIBA. S.. KATO. H. Ionizing radiation and luncr cancer: A review including preliminary results from a case-control study among A-bomb survivors. In: Prentice, R.L., Thompson. D.J. (eds.) Atomic Bomb Survivor Data: Utili:ation andAnalysis. Philadelphia: Siam Institute for Mathematics and Society, 1984. BOXER, G.E.. RICKARDS, J.C. Studies on the metabolism of the carbon of cyanide and thiocyanate. Archives of Biochemistry and &oph' ysics 39(1):7-26, July 1952. BRESLOW, N.E., DAY. N.E. Statistical Methods in Cancer Research. Volume 11-The Design and Analysis of Cohort Studies. [ARC Scientific Publications No. 82. International A~encv for Research on Cancer, Lyon, France. 1987. ~ BRITTEN. N. Validity of claims to lifelong non-smoking, at age 36 in a longitudinal study. /nternatianal Journal of Epidemiology 17(3 ):525-529. 1988. BROSS, I.D.J. Misclassification in 2x2 tables. Biometrics 10:478--186. 1954. BROWN, C.C., CHU, K.C. Use of multistage models to infer stage affected by carcinogenic exposure: Example of lung cancer and cigarette smoking. Journal of Chronic• Diseases 40(Supplement 2):171S-179S, 1987. BROWNELL. K.D., MARLATT, G.A.. LICHTENSTEIN. E.. WILSON. G.T. Understandine and preventing relapse. American Psti•chologist 41(7):765-782, July 1986. v BROWNER. W.S., NEWMAN. T.B.. CUMMINGS. S.R. Designing a new study. III. Diag- nostic tests. In: Hulley, S., Cummines. S. (eds.) Desi,qning a New Sttcdy. 1988. Chapter 9. CAIRNS, J. The cancer problem. Scientific American 233(5):64-72. 77-78. November 1975. CAREY, K.B., ABRAMS. D.B. Properties of saliva cotinine in young adult light smokers. American Journal of Puhlic Health 78(7):842-843, July 1988. CLAVEL, F., BENHAMOU, S., COMPANY-HUERTAS, A., FLAMANT. R. Helping people to stop smoking: Randomised comparison of groups being treated with acupuncture and nicotine gum with control group. British Medical Journa1291(6508)1538-1539, November 30, 1985. 59 TIMN 438469

has been used to validate self-reports of not having smoked in the past 7 days and may be useful to validate up to 3 to 4 weeks. However, specificity of this measure is low compared with cotinine and CO. Bogus Pipeline The bogus pipeline, an assertion to subjects that biochemical assessments will be used to assess smoking status when they will actually only be collected but not evaluated, is used mostly in research with adolescents. One of the reasons given by researchers for continuing to use biochemical verification for at least some proportion of the total subjects is the assertion that if the subjects believe biochemical validation will occur, they will be more likely to provide valid responses to self-report measures. This "bogus pipeline effect" was first presented by Evans, Hansen, and Mittelmark (1977) from the work of Jones and Sigall (1971) concerning smoking among adolescents. It is believed that there is great pressure among adolescents to misreport smoking activities. Murray and coworkers (1987) provided an extensive review of this aspect. Murray and Perry (1987) attempted to determine the conditions under which a bogus pipeline will be effective by manipulating conditions of anonymity. They demonstrated that a bogus pipeline for adolescents is more likely to have an effect if there is an expectation that subjects would otherwise perceive large amounts of pressure to report not smoking and there is a credible pipeline message. However, their findings suQgest that an effective procedure to ensure anonymity can reduce this pressure and likewise reduce the need for the pipeline. Contextual Issues Affecting Biochemical Assessment The accuracy of self-report measures, the desirability for behavioral or biochemical validation of self-report, and the type of assessment needed are issues that need to be considered in the context of the type of study, the nature and size of the study sample, and possible refusal problems. The nature of the subject sample can affect the likelihood of misreporting and therefore the desirability of validation by biochemical assessment. In Table 1, studies demonstrating misreporting rates for individuals who report cessation but who are assessed to be smokers by cotinine or nicotine measurement are classified into three types of subjects: untreated volunteer samples, intervention samples, and high-risk for disease and/or medical patients. Table 2 presents a similar classification of studies demonstrating misreporting with CO validation. The tables are adapted from Lee's work (1988) with the inclusion of additional studies. In cases where multiple cutoff criteria are recorded, the values closest to the optimal cutoff are reported. Several studies should be viewed as outliers and are noted in the tables. These studies reported unusually high rates of individuals who reported not smoking but were above the cutpoint and also employed cutoff criteria far below optimum cutpoints (aand Richard 1988). (Cummings For untreated volunteer samples, the mode for individuals classified as smokers by biochemical assessment who reported not smoking is zero, and no sample exceeds 5 TIMN 438447 37

TABLE 3.-Estimated probability of dying in the next 16.5-year interval for quitting at various ages compared with never smoking and continuing to smoke, by amount smoked and sex Males Age at 1-20 cig/day ?21 cig/day quitting or at start of Never Continuing Former Continuing Former interval smokers smokers smokers smokers smokers 40-44 0.05 0.11 0.05 0.14 0.07 45-49 0.07 0.18 0.10 0.22 0.11 50-54 0.11 0.27 0.17 0.31 0.21 55-59 0.18 0.39 0.28 0.46 0.33 60-64 0.30 0.54 0.46 0.56 0.51 65-69 0.46 0.68 0.59 0.67 0.64 70-743 0.40 0.61 0.55 0.58 0.52 Females Aee at 1-19 cig/day j20 cis/day quitting or at start of Never Continuina Former Continuins Former interval smokers smokersZ smokers smokers smokers 40-44 0.03 0.06 0.03 0.08 0.04 45-49 0.04 0.09 0.06 .0.13 0.05 50-54 0.07 0.14 0.07 0.19 0.09 55-59 0.11 0.21 0.12 0.27 0.15 60-64 0.18 0.30 0.19 0.38 0.32 65-69 0.30 0.46 0.39 0.52 0.32 70-74" 0.26 0.41 0.27 0.45 0.31 NOTE: Based on American Cancer Society Cancer Prevention Study 11 data for persons without a history of cancer, heart disea.se, or stroke who were not "sick" at enrollment. 'Estimates for quitting at this age are estimates of the probability of dying in the next 12.5-yr interval. SOURCE: Unpublished tabulations. American Cancer Society. Kaprio and Koskenvuo 1988). These differences may exist among adolescents prior to initiation of smoking (Seltzer and Oechsli 1985). For these reasons, interpretations of studies comparing these self-selected groups (never smokers, smokers, and quitters) must consider the problem of confounding (Chapter 2). Misclassification, which is discussed in detail in Chapter 2, also must be considered. However, studies of smoking cessation predominantly misclassify persons who are still smoking cigarettes as former smokers, and this would tend to obscure the benefits of cessation in comparison with continued smoking. Further, although the possibility of uncontrolled confounding needs to be considered in epidemiologic studies of smoking cessation and mortality, the totality of data must be interpreted with consideration of its consistency. To account for the evidence of a benefit of quitting that derives from nonexperimental cohort studies, confounders would need to be distributed quite differently among current and TIMN 438490 83

TABLE 7.-Age- and sex-specific mortality rates among never smokers, continuing smokers, and former smokers by amount smoked and duration of abstinence at time of enrollment for subjects in ACS CPS-II study who did not have a history of'cancer, heart disease, or stroke and were not sick at enrollment Males Former smokers (1-2O cig/day) N C Duration of abstinence (yr) Age ever smokers urrent smokers <1 1-2 3-5 6-10 11-15 L> 16 45-49 186.0 439.2 234.4 365.8 159.6 216.9 167.4 159.5 50-54 255.6 702.7 544.7 431.0 454.8 349.7 214.0 250.4 55-59 448.9 1,132:4 945.2 728.8 729.4 590.2 447.3 436.6 60-64 733.7 1,981.1 1,177.7 1,589.2 1,316.5 1,266.9 875.6 703.0 65-69 1.119.4 3,003.0 2,244.9 3,380.3 2,374.9 1.820.2 1,669.1 1,159.2 70-74 2,070.5 4,697.5 4,255.3 5,083.0 4,485.0 3,888.7 3,184.3 2,194.9 75-79 3,675.3 7,340.6 5,882.4 6,597.2 7,707.5 4,945.1 5,618.0 4,128.9 Males Former smokers (>2I cig/day) C Duration of abstinence (yr) Age urrent smokers <1 1-2 3-5 6-10 I1-15 ?16 45-49 610.0 497.5 251.7 417.5 122.6 198.3 193.4 50-54 915.6 482.8 500.7 488.9 402.9 393.9 354.3 55-59 1,391.0 1,757.1 953.5 1,025.8 744.0 668.5 537.8 60-64 2,393.4 1,578.4 1,847.2 1,790.1 1,220.7 1,100.0 993.3 65-69 3,497.9 2,301.8 • 3,776.6 2,081.0 2,766.4 2,268.1 1,230.7 70-74 5,861.3 3,174.6 2,974.0 3,712.9 3,988.8 3,268.6 2,468.9 75-79 6,250.0 4,000.0 4,424.8 7,329.8 6,383.0 7,666.1 5,048.1

,smokers describes the disease burden removed by cessation; whenever possible, this Report provides this comparison. For many diseases, risks for former smokers do not revert to those for never smokers. Relative risks for former smokers compared with never smokers describe the persisting consequences of past active smoking. Thus, in studies concerning the consequences of smoking cessation, the analytic focus is on describing disease incidence after cessation in relation to either the incidence of disease in never smokers or in smokers who do not stop smoking. Interest centers on addressing several questions: In a population that started smoking at a given age, smoked at the same rate, and then quit at a given age, how does the disease rate evolve as a function of time since quitting? In particular, how does the disease rate compare with that of a population of lifelong nonsmokers of the same age or with that of a population of smokers who continue to smoke at the same rate? How does the disease rate after cessation depend on such factors as duration of smoking, number of cigarettes smoked daily, age at starting, or other factors? These analytic questions are generally addressed by estimating either the attributable risk (the difference between the risks for exposed and nonexposed) or the relative risk (the ratio of the risks in exposed and nonexposed) and comparing former smokers with either never smokers or current smokers. A cohort study that observed subjects from birth to death could supply the data requisite for meeting these analytic goals. Observations could be made concerning the age at starting smoking, the amount smoked, the age at stopping smoking, the duration of time since stopping smokinQ, and the occurrence of disease. Incidence rates could be calculated and the attributable risk or relative risk considered as a function of time since quitting. To assess the effects of such factors as duration or amount of smoking, smoking cohorts with different durations and rates could be analyzed. Typically, however, cohort studies enroll subjects at various aQes, and the smoking histories of the subjects span a broad range of ages at starting smoking, durations of smoking, amounts of smokina, aQes at stopping smoking, and ages at observation. In analyzing data from a cohort study, stratification and multivariate modeling are used to describe the disease occurrence in former smokers in relation to the time interval since cessation. New statistical methods have facilitated the analysis of longitudinal data on cancer and other diseases (Breslow and Day 1987; Thomas 1988). The analytic approach should provide control for the effect of changing disease risk with increasing age; as duration of smoking abstinence increases, age and disease risk should be compared with that of never or current smokers in the same age stratum. However, some analytic approaches may introduce overadjustment for the time- related dimensions of smoking history and of age and obscure the benefits of cessation. Age at starting smoking, age at observation, duration of smoking, and duration of abstinence are interdependent: specification of any three of these variables fixes the fourth. Assuming that current and former smokers of a aiven attained age started smoking at about the same age, the duration of smoking among former smokers must be less than for current smokers. Thus, adjustment for duration of smoking in compar- ing current and former smokers is incorrect. Methods that attempt to allow each of these four time-dependent factors to vary freely are inappropriate and provide biased descriptions of the variation in risk following cessation (Brown and Chu 1987). 56 TIMN 438466

MALES Current Smokers Current Smokers <1 <1 1-2 3-5 1-2 3-5 6-10 6-10 Former Smokers Duration of Abstinence (yr) a 1-20 cig/day Q ?21 cig/day 11-15 11-15 -16 -16 FIGURE 1.--Compared with never smokers, relative risk of mortality in current and former smokers aged 50-54, 60-64, and 70-74 at enrollment, by amount smoked and duration of abstinence SOURCE: Unpublished tabulations. American Cancer Society. TIMN 438488 81

TABLE 8.-Estimated probability of dying in the next 16.5-year interval (95% CI) for quitting at various ages compared with never smoking and continuing to smoke, by amount smoked and sex Males Age at 1-20 cig/day ?21 cig/day quitting or at start Never Continuing Former Continuing Former of interval smokers smokers smokers smokers smokers 40-44 0.05 0.11 0.05 0.14 0.07 (0.04-0.05) (0.10-0.12) (0.04-0.06) (0.13-0.15) (0.06-0.09) 45-49 0.07 0.18 0.10 0.22 0.11 (0.07-0.08) (0.17-0.19) (0.08-0.11) (0.21-0.23) (0.10-0.13) 50-54 0.11 0.27 0.17 0.31 0.21 (0.11-0.12) (0.26-0.28) (0.15-0.19) (0.30-0.33) (0.18-0.23) 55-59 0.18 0.39 0.28 0.46 0.33 (0.17-0.19) (0.38-0.41) (0.25-0.31) (0.43-0.48) (0.30-0.37) 60-64 0.30 0.54 0.46 0.56 0.51 (0.28-0.31) (0.52-0.57) (0.42-0.50) (0.51-0.61) (0.48-0.57) 65-69 0.46 0.68 0.59 0.67 0.64 (0.43-0.48) (0.64-0.72) (0.51-0.67) (0.57-0.78) (0.51-0.77) 70-74r 0.40 0.61 0.55 0.58 0.51 (0.38-0.43) (0.56-0.65) (0.45-0.64) (0.44-0.71) (0.32-0.72) Females Age at quitting or at start Never of interval smokers 1-19 cig/day _20 cig/day Continuing Former Continuing Former smokers smokers smokers smokers 40-44 0.03 0.06 0.03 0.08 0.04 (0.03-0.03) (0.05-0.06) (0.02-0.04) (0.08-0.09) (0.03-0.05) 45-49 0.04 0.09 0.06 0.13 0.05 (0.04-0.04) (0.08-0.09) (0.04-0.07) (0.12-0.13) (0.04-0.07) 50-54 0.07 0.14 0.07 0.19 0.09 (0.06-0.07) (0.13-0.15) (0.05-0.09) (0.18-0.20) (0.07-0.11) 55-59 0.11 0.21 0.13 0.27 0.15 (0.11-0.11) (0.19-0.22) (0.09-0.16) (0.25-0.29) (0.12-0.19) 60-64 0.18 0.30 0.19 0.38 0.32 (0.18-0.19) (0.27-0.33) (0.13-0.25) (0.34-0.41) (0.24-0.39) 65-69 0.30 0.46 0.39 0.52 0.32 (0.29-0.31) (0.41-0.52) (0.26-0.52) (0.45-0.59) (0.17-0.47) 70-74" 0.26 0.41 0.27 0.45 0.31 (0.25-0.27) (0.35-0.47) (0.09-0.46) (0.37-0.53) (0.13-0.50) NOTE: Based on American Cancer Society Cancer Prevention Study [1 data for persons without a history of cancer, heart disease, or stroke who were not "sick" at enrollment. Cl=confidence interval. 'Estimates for quitting at this age are estimates of the probability of dying in the next t'_.5-yr interval. SOURCE: Unpublished tabulations. American Cancer Society. TIMN 438503 97

FEMALES u Current Smokers < Current Smokers Current Smokers <1 <1 1-2 1-2 3-5 3-5 6-10 11-15 6-10 >16 11-15 -16 1-2 3-5 6-10 11-15 >_16 Former Smokers Duration of Abstinence (yr)  1-19 cig/day El ?20 cig/day FIGURE 1. (Continued)-Compared with never smokers, relative risk of mortality in current and former smokers aged 50-54. 60-64, and 70-74 at enrollment, by amount smoked and duration of abstinence ~ SOURCE: Unpublished tabulations. American Cancer Society. 82 TIMN 438489

Data from the National Center for Health Statistics (US DHHS 1980) suggest that former smokers have fewer illness days than continuing smokers, particularly among younger women. Gallop (1989) found that former smokers have absentee rates between those of current smokers and never smokers. Segovia, Bartlett, and Edwards (1989) conducted a telephone survey of 3,300 adults and found a strong relation between smoking status and the reporting of good health. Persons who had quit smoking for more than 1 year reported good health with about the same frequency as persons who smoked only I to 5 cigarettes per day, whereas those who had quit for less than 1 year reported good health at a frequency comparable with smokers of 16 to 20 cigarettes per day. Balarajan, Yuen, and Bewley (1985) examined the associations among various levels of smoking, recent and former cessation, and presence of acute and chronic illness, medical office visits, and doctor consultations. Current smokers had a higher prevalence of acute and chronic illness, and rates varied in relation to the amount smoked. Former smokers who had quit in the year prior to the survey had higher rates of illness compared with continuing smokers, and former smokers who quit more than 1 year prior to the survey had rates between those of never smokers and smokers of 20 cigarettes or more per day. Reed (1983) found no difference in general physical health status between current, former, and never smokers, not otherwise defined. Seidell and colleagues (1986) examined the number of reported health complaints, out of an inventory of 51 possible complaints, by smoking status and found that male, but not female, former smokers reported fewer health complaints than smokers. Astrand and Isacsson (1988) found that male employees of a pulp and paper plant who smoked retired at' an earlier age than nonsmokers. Data from the 1979 National Health Interview Survey indicate that smokers have more restricted activity days, more bed disability days, more hospital days, more physician visits, and an increased probability of being unable to work or keep house, than nonsmokers (Rice, Hodgson, Sinsheimer 1986). Analyses of data for the 1976-80 Health Interview Surveys showed that smokers have a 55 to 75 percent excess in days with respiratory conditions associated with reduced activity (Ostro 1989). Smokers experience more school absences (Chariton and Blair 1989; Alexander and Klassen 1988) and work absenteeism (Andersson and Malmgren 1986; Coughlin 1987; Hendrix and Taylor 1987; Gallop 1989) than do never smokers. None of these studies reported information on former smokers. These studies are extremely heterogeneous, with some methodologic shortcomings (Chapter 2). Furthermore, smoking is associated with other behaviors that may affect health (Pearson et al. 1987; Stephens 1986), and the studies do not adjust for changes in otherrisk variables, such as increased exercise, that might be associated with smoking cessation. Taken together, however, the studies are consistent with the hypothesis that smoking cessation produces improvements in health status. This conclusion is evident particularly when considering that smoking-reiated morbidity is a powerful motivation to quit smoking and that recent quitters are likely to be sicker than continuing smokers. TIMN 438496 89-

percent for either cotinine or CO. For intervention studies, values are typically 2 to 5 percent for cotinine and 0 to 10 percent for CO. High risk/medical samples appear to have the highest rates of misclassification of former smokers with the rates exceeding 20 percent. For example, as shown in Table 1, Jarvis and colleagues (1987) reported very low rates (1 percent) of false reporting in vascular patients who were not advised to quit compared with the rate in high-risk patients who were advised to quit (17 percent). It is likely that the pressure to stop smoking influenced the accuracy of patient reporting. Observation studies in which no intervention occurs, or intervention studies in which there is minimal intervention or interaction with smokers, are less likely to prompt false reports of smoking cessation than studies in which intensive intervention does occur. In the former types of studies, in which no or low-intensity intervention occurred, there was a much lower prevalence of subjects reporting a 24-hour quit attempt during the past 6 months or current abstinence (Prochaska et al. 1985) than in intensive interven- tion studies, making misreporting less likely. A greater tendency to misreport in no or low-intensity intervention studies might occur with adolescents, for whom pressures to report not smoking may be omnipresent (Pechacek, Murray et al. 1984; Chapter 2, see section on Bogus Pipeline). A similar pressure might occur in some other instances, such as worksites in which a ban has been placed on smoking, where no intervention occurs but there may still be pressure on individuals to misreport. However, no studies have looked at the possibility of misreporting in such instances. The context in which the study takes place is likely to influence the degree of misreporting. Data currently being collected from smQking cessation programs in a wide variety of contexts may help to clarify this issue. Clinic interventions and intensive interventions, on the other hand, typically ask participants to set a quit date. Close relationships are developed with the counselors, and self-reports of quitting are often given initially in a peer group. Under these higher demand conditions, biochemical verification may be needed to decrease the mis- reporting of current smokers as former smokers. For example, in MRFIT, special intervention subjects claiming to be former smokers at fol lowup examinations had mean SCN- levels between those of never smokers and continuing smokers (Ockene et al. 1982). Similar discrepancies between reported and validated cessation rates did not occur for the usual care men who had not received intensive intervention. The use of biochemical tests for validating self-reports in epidemioloQic studies has a number of limitations. The tests do not have perfect sensitivity and specificity; their half-lives do not necessarily fit the timeframe to be covered: and not all subjects are willing to provide the necessary samples for assessment. A very sensitive test may misclassify subjects as smokers if they have heavy passive smoke exposure (DiGuisto and Eckhard 1986; Haddow. Palomaki, Knight 1986; Haley et al. 1989; Jarvis et al. 1985), smoke occasionally (i.e., 1 or 2 cigarettes on isolated occasions) (Williams et al. 1979), and/or use nicotine in some other form, such as nicotine polacrilex gum or smokeless tobacco (Cohen et al. 1988; Slattery et al. 1989). Biochemical markers are also limited because they assess relatively short-term cessation (less than 2 weeks), and in studies concerned with the impact of cessation on health, there is more interest in evaluating consequences of long-term cessation. ~ITAN 438455 45

TABLE 3.-Continued Reference Population Smoking status and yr since stopped smoking Mortality ratios (N)° Comments I lammond (1966) ACS CPS-1 males 1-19 2!20 1959-63. 3.5-yr followup. cig/day cig/day men aged 50-69 Never smokers 1.002) 1.002) Current smokers 6.5 (8.0) 13.7 (351) Former smokers <I 7.20) 29.1 (33) 1-4 4.6(5) 12.0 (33) 5-9 1.0(1) 7.2(22) ?10 0.4(I) I.I (5) ACS (unpublished ACS CPS-11 males tabulations) 1-20 2t21 cig/day cig/day Never smqkers 1.0(81) 1.0(81) Current smokers 18.8 (6(18) 26.9 (551) Former smokers <I 26.7 (33) 50.7 (64) 1-2 22.4(71) 33.2(117) 3-5 16.5 (82) 20.9 (96) 6-iO 8.7(80) t5.0(t06) I I-I 5 6.0(69) 12.6 (95) _ 16 3.1 (144) 5.5(112)

TABLE 4.-Relative risks of lung cancer among former smokers, by number of years since stopped smoking, and current smokers, from selected case-control studies Reference Population Delinition of former smoker Smoking status and yr since stopped Results Adjustment' Graham and Levin (1971) New York At hospital admission Never smokers Current smokers Former smokers 0-0.5 >0.5-1 > 1-3 >3-10 >10 Males 1.0 8.8 42.2 23.3 10.0 3.3 1.3 Crude Wigle Mao Grace cancer Alberta Canada At interview Males Females Age and , , (1980) , , patients Neversmokers 0.1 0.2 cumulative Current smokers 1.0 1.0 smoking Former smokers <2 2.4 0.9 2-9 0.7 0.5 ~ 10-14 0.7 0.5 ~P. W >IS 0.2 0.4 00 (1984) Correa et al Louisiana NR • Males and females Sex and age ~ . Never smokers 0 1 F~+ ~ Current smokers . 12.6 Former smokers 3-5 7.7 6-2O 7.0 >20 3.9 tA

TABLE 7.-Continued Females . Former smokers (1-19 cig/day) Never Current Duration of abstinence (yr) Age smokers smokers <1 1-2 3-5 6-10 II-I5 Z16 45-49 125.7 225.6 0 433.9 212.0 107.2 135.9 91.0 50-54 177.3 353.8 116.8 92.1 289.5 200.9 121.3 172.1 55-59 244.8 542.8 287.4 259.5 375.9 165.8 202.2 247.2 60-64 397.7 858.(1 1,016.3 365.0 650.9 470.8 570.6 319.7 65-69 692.1 1,496.2 1,108.() 1,348.5 1,263.2 864.8 586.6 618.0 70-74 1,160.0 2,084.8 645.2 1,483.1 1,250.0 1,126.3 1,070.5 1,272.1 75-79 2,070.8 3,319.5 0 2,580.6 2,590.7 3;)60.4 1,666.7 1,861.5 Females Former smokers (>_20 cig/day) Current Duration of abstinence (yr) ASe smokers <1 1-2 3-5 6-10 11-15 2!16 ~ 45-49 277.9 266.7 102.7 178.6 224.7 142.1 138.8 ~ 50-54 517.9 138.7 466.8 270.1 190.2 116.8 83.0 I~el 55-59 823.5 473.6 602.0 361.0 454.5 412.2 182.1 ~ 60-64 1,302.9 1,114.8 862.1 699.6 541.7 373.1 356.4 ~ 65-69 1,934.9 2,319.6 1,250.0 1,688.0 828.7 797.9 581.5 OC 70-74 2,827.0 4,635.8 2,517.2 1,687.3 2,848.7 1,621.2 1,363.4 ~ 75-79 4,273.1 2,409.6 5,769.2 3,125.0 2,978.7 2,803.7 2,195.4 ~ NOTE: Moriality rates are per IINI,(X)0 persons. ACS CPS-I1=American Cancer Sucicty Cancer Prevention Snidy II. SOURCE: Unpublished tnbul:uions, American C:uicer Society.

TABLE 2-Continued Reference Ohlin, Lundh, Westling (1976)" Isacsson and Janzon(1976) Lando(1982) Malcolm et al. (1980)" b Raw et al. (1980) Population Swedish patients with smoking-related diseases attending antismoking clinic and given nicotine gum Swedish heavy smokers in yuit-smoking research project US smokers in muhigroup smoking cessation study UK trial of nicotine chewing gum UK smokers attending a smokers' clinic in comparison of psychologic treatment and use of nicotine gum US smokers in multigroup smoking cessation study Jarvis et al. (1982) UK smokers attending a smokers' clinic in trial of nicotine gum Told to give tip Criterion for false reports of not smoking % (n/N) False reports Comments Yes 0.8% COHb 19.2 (25/130) 19.2% false reports at 32.1 (35/109) I -wk followup; es %,COllb .8(3/34) 32.1% false reports of not smoking at 6-mo fol lowup 8-9-wk followup Yes CO 0 (0/22 to 60) I-yr followup Yes 1.6%, COlIb 41.6 (47/113) 1-mo followup Yes CO or COI Ib 0 (0/33) I-yr followup Yes CO Between 1.4 ( I /14) and Not clear when I "deceiver" withdrew 4.2 (1/24) from study; 1-wk es O or COl lb (0/26) (In4)to I-yr(In4) followup; abstinence status also based on reports of informants I-yr followup

rn TABLE a.-Continued Reference Population Definition of former smoker Smoking status and yr since stopped Results Adjustment" Alderson, Lee, Wang United Kingdom AI hospital admission Males Females Age (1985) Never smokers 0.1 0.2 Current smokers 1.0 1.0 Former smokers 1-3 1.8 2.1 5-10 0.4 0.7 > 10 0.3 0.3 Gao et al. (1988) Shanghai NR Never smokers Males 1.0 Females 1.0 Age and education Current smokers 3.9 2.9 Former smokers 1-4 6.9 7.2 5-9 3.1 3.9 >10 1.1 2.2 Higgins, Mahan, 6 US cities Wynder(1988) At least I yr at time of interview Joly, Lubin, Cuba NR Caraballoso (1983) Males Never smokers 1.0 Former smokers <10 11.9 10-19 6.1 20-29 3.7 ~30 1.9 Males Females Duration of Current smokers 1.0 1.0 smoking Former smokers 1-4 1.2 2.0 ?5 0.6 0.9

observed. Nevertheless, the mortality rate in the intervention group was one-third lower than in the control group (one-sided p value=0.12). Because there were changes in both smoking and cholesterol levels, the difference in mortality cannot be attributed entirely to smoking cessation. The World Health Organization (WHO) European Collaborative Group conducted an intervention study in factories in four European countries (WHO European Col- laborative Group 1983). The study involved random allocation of 66 factories that employed 49,781 men aged 40 to 59 to an intervention program targeting smoking, cholesterol level, and blood pressure or to a control group. After 4 years, the net reduction in mean cigarettes per day in the intervention factories was 8.9 percent (WHO European Collaborative Group 1983). At 6 years, overall mortality in the intervention factories was 4.04 percent; in the control factories, it was 4.15. The difference was not statistically significant. The Multiple Risk Factor Intervention Trial (MRFIT) was a randomized study of more than 12,000 American men, aged 35 to 57 at entry, who were at high risk for CHD on the basis of their smokina behavior, blood pressure, and cholesterol levels (MRFIT Research Group 1982). Men in the special intervention group received an intensive intervention aimed at reducing blood pressure and cholesterol and encouraging smok- ing cessation. Men in the usual care group were referred to their physicians and examined annually. The interventions continued over the entire course of the study. At 6 years, 44.4 percent of special intervention smokers and 25.8 percent of the usual care smokers reported cessation. In the 7-year followup data reported in 1982, there was no difference in total mortality between the special intervention and usual care groups (MRFIT Research Group 1982). However, in the 10.5-year followup data of MRFIT participants, overall mortality for the special intervention participants was 7.7 percent lower than for the usual care group (one-sided p value=0.10; 90-percent confidence'interval (CI),-16.6 to +2.3) (MRFIT Research Group 1990). A subaroup of MRFIT special intervention participants, who were hypertensive, had resting electrocardiogram abnormalities, and comprised 3 1 percent of the special intervention sroup, may have suffered excess mortality as a result of an unanticipated adverse effect of one of the antihypertensive drugs (Cutler. MacMahon, Furbersz 1989). This has recently been suggested as an explanation for the absence of an~overall difference in mortality between the special intervention and usual care groups at the 7-year followup (MRFIT Research Group, submitted for publication). Furthermore, Ockene and coworkers (1990) recently reported that at 10.5 years, MRFIT participants who quit smoking had significantly lower death rates than those who continued to smoke in both special intervention and usual care groups. Most important, like the other multifactor intervention trials, it is difficult to infer a benefit or a lack of benefit of smoking cessation for total mortality from this study. In summary, studies involving smoking cessation interventions include a randomized trial in which smoking cessation was the sole intervention and three intervention studies in which it was a component. The small size of the former and the mixing of a smokinQ intervention with other interventions in the latter make it impossible to reach con- clusions about the benefits of smoking cessation from these studies alone; however. 86 TIMN 438493

MEN ~ KREYBERG II (N-204) ~ 15 a ~ 10 T WOMEN KREYBERG I (N-95) 40.7 35 30 ~ 25~ 20 ~ 15+ 10~ 10 F 15 KREYBERG 11 (N-100) 5+ 01-4 0 5-9 ®10-19 ® 20-29 > 30 1 - 11 - 21 - 31 - >41 NUMBER OF CIG/DAY Yr of abstinence FIGURE 1.-Risk of lung cancer by number of cigarettes smoked per day before quitting, number of years of abstinence, sex, and histologic types SOURCE: Higgins and Wynder (1988). ~IMN 438525 121 -

smoking-related deaths in the intervention group and 92 in the normal care group. The proportionate difference in smoking-related deaths was -9 percent. Again the con- fidence interval was wide (-31 percent to +20 percent). Twenty percent of the men in the intervention group who quit smoking cigarettes took up pipe or cigar smoking compared with 3 percent of the men in the normal care group, and to the extent that pipe and cigar smoking are mortality risk factors, any benefit of cessation of cigarette smoking is obscured. This trial is largely uninformative as to the benefit or lack of benefit of smoking cessation for total mortality because of the small number of subjects. The trial was further compromised by the relatively poor compliance of the subjects with the intervention: the net reduction in mean cigarette consumption over the 10 years of the followup among the intervention group compared with the normal care group was only 7.6 cigarettes per day. Other intervention studies that allow assessment of the relation of smoking cessation to overall mortality have involved multiple interventions aimed at reducing several different factors for CHD. The ability to draw conclusions about the effect of smoking cessation on overall mortality from these studies is quite limited for this reason. ~ The North Karelia study targeted a region of Finland that had the world's highest CHD death rate at the time of the study's initiation (Tuomilehto et a]. 1986) and was aimed at modifying smokine, cholesterol levels, and blood pressure. The rest of Finland was used for comparison. In the 10 years after initiation of an aggressive risk reduction program, there was a 35-percent decrease in smoking in North Karelia compared with a 2-percent reduction in the rest of Finland (Salonen et al. 1989). Blood pressure and cholesterol levels did not change significantly in the intervention area compared with the rest of Finland. Total mortality in the intervention area in the 10 years after the start of the study declined more rapidly than in the rest of Finland, although the difference in the rate of decline in overall mortality was not statistically significant. For at least two reasons, interpretation of the North Karelia study is problematic with respect to the effect of smoking cessation on overall mortality. First, the study was nonexperimental, with conclusions based on a comparison of total mortality in the study area with that of Finland. DurinQ the study period, overall mortality also declined in the rest of Finland, perhaps because of secular changes in other factors related to mortality and to changes in medical care (Salonen et al. 1989). Second, the study was not designed to investigate smoking cessation alone. Because of the mixing of inter- ventions for three CHD risk factors, it was difficult to isolate the impact of the smoking cessation component. The Oslo study (Hjermann 1980; Hjermann et al. 1981; Holme 1982) involved 1.232 normotensive men at high risk for CHD because of their smoking behavior and cholesterol levels. The men were randomly assigned either to receive interventions aimed at reducing both CHD risk factors or to a control Qroup. Tobacco consumption. including pipe and cigar smoking, fell 45 percent more in the intervention group than in the control group. ~ There was also a mean difference of 13 percent in serum cholesterol between the intervention and control groups over 5 years (Hjermann et al. 1981). The study was small, and it was not designed to examine total mortality endpoints; only 42 deaths were 85 TIMN 438492

TABLE 4.-Continued Reference Population Detinition of former smoker Smoking status and yr since stopped Results Adjustment' Lubin et al. (1984a) European case-control At interview Males - Females Duration of study Current smokers 1.0 1.0 smoking Former smokers 1-4 1.1 0.9 5-9 0.7 0.7 10-14 0.6 0.4 15-19 0.4 0.5 20-24 0.4 0.5 >25 0.3 0.3 Pathak et al. (1986) New Mexico At least I yr before Males Number of interview • <_65 >65 cig/day Current smokers 1.0 1.0 Former snwkers 5 0.5 0.7 10 0.2 0.5 20 0.1 0.3 Damber and Larsson Swedenh NR (1986) NOTE: NR=not rcporled. "Factors adjusted for in anatysis by yr of smoking abstinence. °Etitimated from figure 4 of reference. Males Age Current smokers 9.5 Former smokers 1-5 7.5 6-10 3.0 >10 2.0 J

CONTENTS Lung Cancer ....................................................... 107 Pathophysiologic Framework ........................................ 107 Smoking and Histopathology of the Airways ......................... 108 Other Changes ................................................. 109 Smoking Cessation and Lung Cancer Risk .............................. 110 Pattern of Changing Risk After Cessation ............................. 110 Effect of Antecedent Smoking History .............................. 122 Duration of Smoking ............................................ 122 Daily Cigarette Consumption ................................... 124 Inhalation Practices ........................................... 124 Different Tobacco Products ..................................... 124 Effect of Age at Cessation ...................................... 125 Multistage Modeling .............................................. 126 Cessation After Developing Disease .................................. 129 Cessation After Diagnosis of Lung Cancer ............................. 129 Laryngeal Cancer .................................................... 131 Pathophysiologic Framework ........................................ 131 Smoking Cessation and Laryngeal Cancer Risk ......................... 131 Conclusions ....................................................... 135 References ........................................................137 105 TIMN 438510 51®

TABLE 6.-Relative risks of lung cancer among never, former, and current smokers by types of tobacco products smoked Smoking status Reference Population Tobacco product Never smokers Former smokers Current smokers I liggins, Mahan, Wynder (1988) 6 US cities Cigarettes only 1.0 6.9 16.0 Cigars only 1.0 2.5 3.1 Pipes only 1.0 0.7 1.9 Cigars and pipes 1.0 2.4 2.5 Mixed smokers 1.0 5.1 10.5 Lubin. Richter, Blot (1984) European case-control Yr since stopped 1-4 ?5 study Cigars only 0.6 0.7 1.0 Mixed cigars and cigareues 4.4 0.9 1.0 Pipes only 2.0 0.9 1.0 Mixed pipes and cigarettes 1.2 0.8 1.0 Damber and Larsson (1986) Sweden Yr since stopped 1-10 >10 Cigarettes only° 5.0 1.2 9.5 Pipes only 5.0 4.5 8.0 "Estinwted from figure 5 uf re/erenct: relerence group is never snuikCrs.

.A TABLE 3.-Confinued Smoking status Reference Population and yr since Mortality ratios (N)° Comments stopped smoking ACS (unpublished ACS CPS-I1 females tabulations) 1-19 Z20 cig/day cig/day Never smokers 1.0(181) 1.0(181) Current smokers 7.3(145) 16.3(434) Former smokers <I 7.9(5) 34.3 (31) 1-2 9.103) 19.5 (42) 3-5 2.9(7) 14.602) 6-1O 1.00) 9.102) 11-15 1.5(6) 5.9(20) _16 1.4(23) 2.6(18) NOTE: ACS CPS-I .md -I1=Amcrican Cancer Six:iety C:incer Prtvemion Siudies I anJ II. ''Numtxer of observations. b hictudes data only for ex-cigarcne smokers who stopped for reusons other th:m physician's order.

GALLOP, B. Sickness absenteeism and; smoking. (Letter.) New Zealand Medical Journal 102(863):112, March 8, 1989. GARFINKEL, L., STELLMAN, S.D. Smoking and lung cancer in women: Findings in a prospective study. Cancer Research 48(23):6951-6955, December 1, 1988. GORDON, T., KANNEL, W.B., MCGEE, D. Death and coronary attacks in men after giving up cigarette smoking. Lancet 1345-1348, December 7, 1974. GOTTLIEB, N.H. The determination of smoking types: Evidence for a sociological- pharmacological continuum. Addictive Behaviors 8(1):47-51, 1983. HAINES, A.P., IMESON, J.D., MEADE, T.W. Psychoneurotic profiles of smokers and non-smokers. British Medical Journa1280(6229):1422, June 14, 1980. HAMMOND, E.C. Smoking in relation to the death rates of one million men and women. In: Haenszel, W. (ed.) Epidemiological Approaches to the Study of Cancer and Other Chronic Diseases. NCI Monograph 19. U.S. Department of Health, Education, and Welfare, Public Health Service, National Cancer Institute. January 1966, pp. 127-204. HENDRIX, W.H., TAYLOR, G.S. A multivariate analysis of the relationship between cigarette smoking and absence from work. American Journal of Health Promotion 2(2):5-11, Fall 1987. HJERMANN, I. Smoking and diet intervention in healthy coronary high risk men. Methods and 5-year follow-up of risk factors in a randomized trial. The Oslo Study. Journal of the Oslo City Hospitals 30(1):3-17, January 1980. HJERMANN, L, HOLME, I., VELVE BYRE, K., LEREN, P. Effect of diet and smoking intervention on the incidence of coronary heart disease. Lancet 2(8259):1303-1310, Decem- ber 12, 1981. HOLME, 1. On the separation of the intervention effects of diet and antismokina advice on the incidence of major coronary events in coronary high risk men. The Oslo Study. Journal of the Oslo City Hospitals 32(3l4):31-54, March-April 1982. KAHN, H.A. The Dom study of smoking and mortality among U.S. veterans: Report on eight and one-half years of observation. In: Haenszel. W. (ed.) Epidemiological Approaches to the Study of Cancer and Other Chronic Diseases. NCI Monograph 19. U.S. Department of Health, Education, and Welfare, Public Health Service, National Cancer Institute. January 1966, pp. 1-125. KAPRIO, J., KOSKENVUO, M. A prospective study of psychological and socioeconomic characteristics, health behavior and morbidity in cigarette smokers prior to quitting compared to persistent smokers and non-smokers. Journal of Clinical Epidemiology 41(2):139-150, 1988. KLEINBAUM, D.G., KUPPER, L.L., MORGENSTERN, H. Epidemiologic Resear•c•h. Bel- mont, California: Lifetime Learning Publications, 1982. LEU, R.E., SCHAUB, T. Does smoking increase medical care expenditure? Social Science and Medicine 17(23):1907-1914, 1983. LUCE, B.R., SCHWEITZER, S.O. Smoking and alcohol abuse: A comparison of their economic consequences. New England Journal of Medicine 298(10):569-571, March 9, 1978. MARSDEN, M.E., BRAY, R.M., HERBOLD. J.R. Substance use and health among U.S. military personnel: Findings from the 1985 worldwide survey. Preventive Medicine 17(3):366-376, May 1988. MATTSON, M.E., POLLACK, E.S., CULLEN, J.W. What are the odds that smoking will kill you? American Journal of Public Health 77(4):425-43 1, April 1987. MCMANUS, I.C., WEEKS, S.J. Smoking, personality and reasons for smoking. Psychological Medicine 12(2):349-356, May 1982. 100 1 TIMN 438506

STOOKEY, G.K., KATZ, B.P., OLSON, B.L., DROOK, C.A., COHEN, S•J. Evaluation of biochemical validation measures in determination of smoking status. Journal of Dental Research 66(10):1597-1601, 1987. SWAN, G.E., PARKER, S.D., CHESNEY, M.A., ROSENMAN, R.H. Reducing the confound- ing effects of environment and diet on saliva thiocyanate values in ex-smokers. Addictive Behaviors 10(2):187-190, 1985. THOMAS, D.C. Models for exposure-time-response relationships with applications to cancer epidemiology. Annual Review of Public Health 9:451-482, 1988. TUOMILEHTO, J., GEBOERS. J., SALONEN, J.T., NISSINEN, A., KUULASMAA, K., PUSKA, P. Decline in cardiovascular mortality in North Karelia and other parts of Finland. British Medical Journal 293:t068-1071, October 25, 1986. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking: Cancer. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Office on Smoking and Health. DHHS Publication No. (PHS) 82-50179, 1982. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking: Cardiovascular Disease. A Report of the Siu geon General. U.S. Department of Health and Human Services, Public Health Service, Office on Smoking and Health. DHHS Publication No. (PHS) 84-50204, 1983. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Involuntary Smoking. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service. Centers for Disease Control. DHHS Publication No. (CDC) 87-8398, 1986. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. Reducing the Health Conse- quences of Smoking: 25 Years of Progress. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. DHHS Publication No. (CDC) 89-8411, 1989. U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE. Smoking and Health. A Report of the Surgeon General. U.S. Department of Health, Education, and Welfare, Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health. DHEW Publication No. (PHS) 79-50066,1979. U.S. PUBLIC HEALTH SERVICE. Smoking and Health. Report of the Advisoiy Committee to the Surgeon General of the Puhlic Health Service. U.S. Department of Health. Education, and Welfare, Public Health Service, Center for Disease Control. PHS Publication No. 1103, 1964. VESEY, C. Thiocyanates and cigarette consumption. In: Greenlaugh, R.M. (ed.) Smoking and Arterial Disease. London: Pitman Press, 1981. VLIETSTRA, R.E., KRONMAL, R.A., OBERMAN, A., FRYE, R.L., KILLIP, T. III. Effect of cigarette smoking on survival of patients with angiographically documented coronary artery disease. Report from the CASS Registry. Journal of the American Medical Assocation 255(8):1023-1027, February 28, 1986. VOGT, T.M. Smoking behavioral factors as predictors of risks. In: Jarvik. M.E., Cullen, J.W., Gritz, E.R., Vogt, T.M., West, L.J. (eds.) Research on Smoking Behavior. NIDA Research Monograph 17. U.S. Department of Health, Education, and Welfare, Public Health Service, Alcohol, Drug Abuse, and Mental Health Administration, National Institute on Drug Abuse. DHEW Publication No. (ADM) 78-581, December 1977, pp. 98-110. VOGT, T.M. Questionnaires vs. biochemical measures of smoking exposure. (Letter.) American Journal of Public Health 72(1):93, January 1982. 68 TIMN 438478

LUNG CANCER Epidemiologic studies have provided overwhelming evidence for a causal association of cigarette smoking with lung cancer (US PHS 1964; US DHEW 1979; US DHHS 1989). The plausibility of this association is supported by the presence of numerous carcinogens in tobacco smoke. Compared with the risk among never smokers, the risk of lung cancer for smokers may be increased twentyfold or more for heavy smokers (US DHHS 1989). Risk of lung cancer increases with the number of cigarettes smoked daily and the duration of cigarette smoking; risk declines after cessation (US DHHS 1982, 1989). For example, in an analysis of data from the British Physicians Study, Doll and Peto (1978) indicated that among subjects who persisted in smoking, lung cancer incidence increased with the fourth or fifth power of the duration of smoking and with approximately the square of daily cigarette consumption. In 1985, estimated attributable risks of lung cancer from cigarette smoking were 90 percent for males and 79 percent for females in the United States (US DHHS 1989). This Section considers the effects of cigarette smoking on the epithelium of the airways of the lungs, the site from which most lung cancers stem, and the evolution of the smoking-related changes after cessation. The epidemiologic evidence on lung cancer risk after smoking cessation is comprehensively reviewed; the change in risk over time following cessation is described; and factors modifying the effect of cessation are considered. The Section includes discussion of the application of multistage modeling to data on smoking cessation. Pathophysiologic Framework Previous Surgeon General's reports have provided extensive reviews on carcinogenic components of tobacco smoke and on experimental carcinogenesis with tobacco smoke (US DHEW 1979; US DHHS 1982, 1986). Tobacco smoke contains numerous carcinogenic agents with both initiating and promoting activity. Although the specific mechanisms of respiratory tract carcinogenesis by tobacco smoke are not yet fully characterized, the plausibility of the smoking-lung cancer relation has been considered to be well supported by the available information (US PHS 1964; US DHHS 1982). Carcinogenesis in the respiratory tract is widely considered to be a multistep process involving sequential changes in a cell from the normal to the malignant state. Extensive experimental and human evidence is consistent with the multistage hypothesis, and application of the new molecular and cellular biology techniques to the study of lung cancer is providing further insights into the genetic mechanisms underlying the development of this disease (Birrer and Minna 1988). Experiments with animals have shown that agents may initiate or promote cancer. In animal experiments involving a sequence of exposures to agents, those agents that cause cancer when administered initially are referred to as initiators, whereas agents that promote the growth of initiated cells are referred to as promoters. Diverse multistep models of carcinogenesis have been developed (Farber 1984). The age-incidence patterns for epithelial cancers such as lung cancer, which show that the rates usually increase as a power of age, are also consistent with a multistage process TI1121V 438511 107

CHAPTER 5 SMOKING CESSATION AND NONRESPIRATORY CANCERS

TABLE S.-Relative risks of lung cancer among never, current, and former smokers, by nuniber of years since stopping smoking and histologic type Reference Wynder and Stellman (1979) Benhamou et al. (1985) Lubin and Blot(1984) Population 6 US cities French males, European case-control study European case-control study Smoking st:uus .md yr since stopped Males Kreyberg type Histologic type Females Kreyberg type I It 1 II Never smokers _ 1.0 1.(1 1.0 1.0 ('turent smokers 32.3 10.7 10.5 4.4 Former smokers 1-3 53.8 14.2 13.6 6.7 4-6 24.9 5.9 6.2 3.6 7-10 17.2 6.6 5.1 4.1 II-15 13.7 5.4 8.8 5.6 ?16 5.0 1.2 0.9 Never smokers Former smokers 1-3 4-6 7-10 11-19 ?20 Current smokers Former smokers 1-4 5-9 10-14 15-19 >20 Males Kreyberg type I II 34.6 12.2 10.9 6.3 4.2 6.7 2.1 1.0 Maleti SQ ADENO 1.0 1.0 1.1 1.0 0.7 0.8 0.6 0.6 0.4 0.6 0.4 0.5 Females SQ ADENO 1.0 1.0 1.1 0.9 0.4 0.4 0.3 0.7 1.0 0.4 1.2 0.3 NOTE: SQ=squamous cell carcinoma of the Iung: AI)GNO=adentx:arcinoma of ihe lung.

TABLE 9.-Relative risks of laryngeal cancer by smoking status Reference Population Smoking status Relative risks Kahn (1966) US veterans Never smokers 1.0 Current smokers 9.5 Former smokers 7.2 Wigle, Mao, Grace Alberta, Canada, cancer Never smokers 1.0 (1980) patients Current smokers 7.8 Former smokers 6.3 ACS (unpublished ACS CPS-11 Males Females tabulations) Never smokers 1.0 1.0 Current smokers 12.8 9.5 Former smokers 6.7 6.5 Falk et al. (1989) Texas Never smokers 1.0 Curient smokers 9.0 Former smokers 3.2 (yr since stopped)" 1-10 11-20 Cig/day 21-30 31-40 >40 3-9 3.0 3.6 4.0 7.2 0.9 >_ 10 2.8 1.2 1.0 3.1 3.5

1 0.9-i 0.8-I 0.7 - X Cn .. 0.6 - 1 1 ~ WITH ADJUSTMENT FOR ~ SMOKING DURATION ~ WITHOUT ADJUSTMENT FOR~. . SMOKING DURATION 11 . . . . . ~ cG 0.5 uJ > 0.4 r Q 0.3 Ul 02 Cd 0.1 b 0 5 10 15 20 25 30 35 YR SINCE QUIT SMOKING FIGURE 2.-Relative risk of lung cancer among ex-smokers compared with continuing smokers as a function of time since stopped smoking, estimated from logistic regression model, pattern adjusted for smoking duration compared with pattern unadjusted for duration SOURCE: Brown and Chu (1987). time than men who had stopped for a shorter time. The relative risk of lung cancer continued to decrease sharply with increasing years of abstinence without adjusting for smoking duration, whereas the decreasing relative risk plateaued when adjusted for duration of smoking (Figure 2). The difference in this pattern was most noticeable for increasing years of smoking abstinence. For those who had stopped smoking for 27 years or more, the relative risk compared with continuing smokers was 0.40 when adjusted for duration, but 0.17 when no adjustment was made. However, control for previous duration of smoking (or cumulative previous smoking history) in determining the risk of lung cancer among former smokers may constitute overadjustment if age and duration of cessation also are included in the model (Chapter 2). ~ In summary, only limited analyses address the effect of duration of previous smoking on the decline in risk following cessation. The data point to less decline of relative risk following cessation, comparing longer term with shorter term studies, but additional investigation is needed. 123 TIMN 438527 `

TABLE 2.-Relative risks of lung cancer among never, former, and current smokers in selected epidemiologic studies Smoking status Reference Population Subgroup Never smokers Former smokers Current smokers - 1-19 cig/day _20 cigklay 1-19 cig/day ?20 cig/day I lanmnond (1966) ACS CPS-I 1.0 2.0 7.9 6.5 13.7 Kahn (1966) US veterans 1.0 4.7 10.9 Canadian Department of Canadian males 1.0 6.1 14.9 National I lealth and Welfare (1966) Cederlof et al. (1975) Males 1.0 6.1 7.8 Females 1.0 1.5 4.5 Doll and Peto (1976) British male physicians 1.0 4.3 10.4 Doll et al. (1980) British female physicians 1.0 3.3 6.4" Wigle, Mao, Grace Alberta (Canada) cancer Males 1.0 6.5 10.4 (1980) patients Females 1.0 2.1 5.2 ~ Wu et al. (1985) Los Angeles (CA) whites Syuamous 1.0 7.7 35.3 Adunoaarcinoma 1.0 1.2 4.1 .2 'Carstensen, Pershagen, Swedish males 1.0 1.1 7.5h Eklund(1987) ACS ACS CPS-II Males 1.0 8.9 21 3 U4 (unpublished Females 1 0 4 8 . 12 1 00 tabulations) . . . ~ ~ NOTG: ACS CPS-1 and II=American Cancer Society Cancer Prevemion SwJies I and 11. ~ ' 15-24 cig/day. h8-15 cig/day.

TABLE 3.-Lung cancer mortality ratios among never, current, and former smokers by number of years since stopped smoking (relative to never smokers), prospective studies Reference Population Smoking status and yr since stopped smoking Mortality ratios (N)" Comments Doll and Peto (1976) British male physicians Never smokers 1.00) .1951-71, 20-yr followup; Current smokers 15.8 (123) data on former smokers in Former smokers summary form 1-4 16.0 (15) 5-9 5.9(12) 10-14 5.3(9) - 15 2.0(7) Rogot and Murray (1980) US veteransh Current smokers 11.3 (2,609) 1954-69, 16-yr followup Former smokers 1-4 18.8 (47) 5-9 7.7(86) 10-14 4.7 (100) 15-19 4.8(115) _20 2.1 (123) US DHIIS (1982) Japanese males Current smokers 3.8 Former smokers 1-4 4.7 5-9 2.5 >_I0 1.4

CONCLUSIONS 1. Smoking cessation reduces the risk of lung cancer compared with continued smok- ing. For example, after 10 years of abstinence, the risk of lung cancer is about 30 to 50 percent of the risk for continuing smokers; with further abstinence, the risk continues to decline. 2. The reduced risk of lung cancer among former smokers is observed in males and females, in smokers of filter and nonfilter cigarettes, and for all histologic types of lung cancer. 3. Smoking cessation lowers the risk of laryngeal cancer compared with continued smoking. 4. Smoking cessation reduces the severity and extent of premalignant histologic changes in the epithelium of the larynx and lung. 135 TIMN 438539

Characteristic profiles were found in the laryngeal and lung tissues; levels of adducts tended to increase with the amount of cumulative smoking. The study included only three long-term former smokers with duration of abstinence ranging from 10 to 14 years. These subjects had low levels of adducts compared with current smokers. Smoking Cessation and Lung Cancer Risk Pattern of Changing Risk After Cessation Numerous cohort and case-control studies have documented a reduction in the relative risk of lung cancer among former smokers compared with current smokers. The findings of selected studies are presented in Table 2. Former smokers in these studies experienced a 10- to 800-percent increase in risk of lung cancer compared with never smokers; however, compared with current smokers, former smokers showed a 20- to 90-percent reduction in risk. The relative risk estimates provided in Table 2 group former smokers with varying durations of abstinence from smoking. However, the number of years since cessation has a strong effect on risk of lung cancer among former smokers: in studies assessing risk by duration of abstinence, the reduced risk has been evident within 5 years of cessation compared with continued smoking, and the benefit of cessation has increased as the duration of abstinence lengthened. However, in most of the studies, the risk of lung cancer among former smokers remained elevated above the risk among never smokers, even in the longest peridds of abstinence evaluated. In many studies, risks among former smokers were higher than among continuing smokers during the first few years after stopping smokihg. This pattern of risk reflects cessation by individuals who quit smoking because of symptoms and illness before the clinical diaanosis of lung cancer (Chapter 2; Haenszel, Loveland, Sirken 1962; Doll and Hill 1964; Kahn 1966). Table 3 summarizes standardized mortality ratios of lung cancer among former smokers by years of abstinence, as reported in five cohort studies: British physicians, U.S. veterans, Japanese males, and the American Cancer Society Cancer Prevention Studies, ACS CPS-I and ACS CPS-II. These studies varied in the length of followup, the extent of information obtained on smoking history, and the number of lung cancer cases. Compared with never smokers, former smokers who had been abstinent for 10 to 20 years or more showed a varying extent of risk reduction among the studies. In the British Physicians Study, U.S. Veterans Study, and ACS CPS-II, former smokers who had been abstinent for 15 years or more showed an 80- to 90-percent reduction in risk compared with current smokers. The percentage reduction in risk was slightly lower amona the Japanese cohort and higher in ACS CPS-I. y Results from selected case-control studies are shown in Table 4. As in the cohort studies, former smokers who had been abstinent the longest experienced increased risk compared with never smokers, but substantially reduced risk in most studies compared with current smokers. Thus, reduction in risk of lung cancer after smoking cessation has been observed in numerous cohort and case-control studies conducted in the United Kingdom (Doll and Peto 1976; Alderson, Lee, Wang 1985), the United States (Kahn 1966; Hammond 1966; 110 TIMN 438514

(Doll 1971; Doll and Peto 1978; Peto 1984; Day 1984). The bronchial epithelia of sustained smokers show a progression of abnormality (Saccomanno et al. 1974). The pseudostratified, ciliated epithelium becomes metaplastic and then dysplastic. Car- cinoma in situ may develop and eventually become invasive (McDowell, Harris, Trump 1982). To the extent that cigarette smoking affects late as well as early stages in this process, smoking cessation would be expected to have beneficial consequences on lung cancer incidence. The epidemiologic evidence provides strong support for the an- ticipated benefits of smoking cessation. Cigarette smoking is associated with changes in the large and small airways, in the respiratory epithelium and parenchyma, and in the numbers, type, and functional capacities of inflammatory cells. The reversibility of these changes after smoking cessation is germane to respiratory carcinogenesis and to the health consequences of smoking cessation. This Section focuses on studies that have examined the effect of smoking on the respiratory epithelium and on the cells in the lungs of current, former, and never smokers. Additional relevant information is reviewed in Chapter 7 and in previous reports of the Surgeon General (US DHHS 1984, 1986). Smoking and Histopathology of the Airways Extensive histopathologic evidence is available on the effects of smoking on the airways of the lung. The association between smoking and premalignant changes in the bronchial epithelium has been addressed by many investigators (US DHHS 1982). Based on sequential examinations of exfoliative cytologic specimens from uranium miners over -a period of many years, Saccomanno and colleagues (1974) reported evidence of squamous metaplasia progressing through increasing atypia to carcinoma in situ and invasive bronchogenic carcinoma. Detailed observations have been made on the histopathology of lung specimens obtained at autopsy (Auerbach et al. 1957, 1962a,b, 1963, 1964, 1972; Auerbach, Garfinkel, Hammond 1974). In 1962, Auerbach and coworkers (1962a) reported that the frequency and intensity of epithelial changes increased with the number of cigarettes smoked daily. In addition, these investigators assessed changes following smoking cessation in postmortem bronchial epithelial specimens from 72 ex-smokers and controls matched individually with 2 controls per case (Auerbach et al. 1962b). One control was a cutrent smoker matched with an ex-smoker on age, occupation, residence, and smoking history. The second control was a lifetime nonsmoker also matched with an ex-smoker on aQe, occupation, and residence. Some type of epithelial abnormality was found in 98 percent of histologic sections from current smokers, 67 percent from ex-smokers, but only 26 percent from never smokers. This pattern persisted for many specific types of epithelial abnormalities including absence of ciliated cells, presence of atypical cells, and presence of hyperplasia and goblet cells in glands (Table 1). The occurrence of unciliated atypical cells, the most severe change before invasive carcinoma, was similar among ex-smokers and never smokers but was considerably greater among current smokers. The number of cells with atypical nuclei was reported to decrease with increasing number of years since smoking cessation. When current smokers were matched with former smokers of the same age at time of cessation, former smokers 108 TIMN 438512

TABLE 7.-Standard mortality ratios of lung cancer antiong former smokers in ACS-CPS II (relative to never smokers) by years of smoking abstinence, daily cigarette consumption at time of cessation, and history of chronic disease No history of chronic disease" All respondents 1-20 -21 1-20 ?21 ' cir./day cig/day cig/day cig/day Males Current smokers Former smokers (yr since stopped) <1 1-2 3-5 6-10 l 1-15 23.5 16.8 16.7 19.7 8.6 6.3 31.5 23.4 25.3 20.5 14.2 13.6 18.8 26.7 22.4 16.5 8.7 6.0 26.9 50.7 33.2 20.9 15.0 12.6 -16 3.3 5.3 3.1 5.5 No history of chronic disease° All respondents 1-19 cig/day >-20 cig/day 1-19 cig/day ?20 cig/day Females Current smokers 10.5 24.1 7.3 16.3 Former smokers (yr since stopped) <1 3.4 21.1 7.9 34.3 , 1-2 9.0 18.2 9.1 19.5 3-5 2.5 13.2 2.9 14.6 6-10 1.1 12.0 1.0 9.1 1 I-15 1.1 2.9 1.5 5.9 -16 1.6 2.4 1.4 2.6 ''No history of cancer, heart diseytie, or stroke. SOURCE: Unpublished tabulations. American Cancer Society. extent, performance status, and type of protocol treatment. Similarly, statistical sig- nificance was maintained after simultaneous adjustment for both thymosin and radia- tion therapy. The study by Bergman and Sorenson (1988) involved 154 small cell lung cancer patients who received combination chemotherapy. Thirty-two had stopped smoking; at least 6 months before the initiation of treatment or had never smoked, 5 1 patients stopped smoking less than 6 months prior to the start of treatment, and 71 patients continued to smoke during the treatment period; the median survival was 39, 42, and 40 weeks, respectively. Reasons for differences in results between the two studies are not clear. Overall, patients in the study by Bergman and Sorenson (1988) had smoked fewer pack-years, but the median survival and performance.status of each of the three 130

Graham and Levin 1971; Pathak et a1.1986), Canada (Wigle, Mao, Grace 1980), Europe (Lubin et al. 1984a; Damberand Larsson 1986), Asia (US DHHS 1982; Gao et al. 1988), and Latin America (Joly, Lubin, Caraballoso 1983). Although only a few studies had information on female former smokers, the pattern of risk reduction was similar to that observed for males. Decrease in risk after smoking cessation also has been reported for each of the major histologic types of lung cancer (Wynder and Stellman 1977; Lubin and Blot 1984; Benhamou et aI. 1985; Higgins and Wynder 1988) (Table 5 and Figure 1). Higgins and Wynder (1988). found that the decline in risk after cessation was more consistent for Kreyberg I tumors (primarily squamous cell, small cell, and large cell cancers) than for Kreyberg II tumors (primarily adenocarcinomas and bronchiolo- alveolar carcinomas) (Figure 1). Smokers of filter and nonfilter cigarettes (Wynder and Stellman 1979; Lubin et al. 1984b) and of other tobacco products (Joly, Lubin, Caraballoso 1983; Lubin et al. 1984b; Damber and Larsson 1986; Higgins, Mahan, Wynder 1988) have reduced lung cancer risk following cessation (Table 6). Although the findings of the reviewed studies uniformly indicate lower risk among former smokers, the magnitude and rapidity of the risk reduction with smoking cessation varies among the studies. This variation has several potential explanations. First, years of abstinence among those who stopped smoking for the longest time interval varied from 5 to 25 years or more. Second, although former smokers have a risk of lung cancer between those of continuing smokers and never smokers, the pattern of declining risk as duration of abstinence lengthens has not been fully characterized. The small number of former smokers in some studies limits the precision with which the decline in risk can be described, particularly for the longer durations of abstinence. Third, aspects of the active smoking history, includin.g cumulative smoking exposure up to the time- of quitting, age at initiation, years of smoking, number of cigarettes smoked per day, inhalation practices, types of cigarettes and other tobacco products smoked, age at smoking cessation, and the reason for stopping, may modify the risk of lung cancer after cessation (Chapter 4, see section on Effect of Antecedent Smoking History). The varying extent to which these factors have been considered in analyzing the effect of cessation may partially explain the differences in risk observed in former smokers among the studies. As discussed below, failure to adjust for previous smoking history may exaggerate the benefit of smoking cessation, but adjustment for cumulative smoking history also may result in overadjustment of the risk estimate (Chapter 2). Fourth, the studies vary in the definition of former or ex-smokers and in the analytic treatment of former smokers who have recently stopped smoking. In the case-control studies, former smokers have been defined as individuals who were abstinent at the time of interview, at the time of cancer diagnosis, or at some other reference point (e.g., 1 year before diagnosis of lung cancer and a comparable time for controls). To reduce the bias introduced by quitting because of illness, former smokers who stopped smoking after developing symptoms or disease may be excluded from analysis. Information on the reason for cessation was collected only in some studies, and persons with symptoms at cessation have not been handled uniformly in the published literature. Finally, results of the relevant studies are not totally comparable because the risks of former smokers were compared with those of never smokers in some studies and with continuing smokers in others. 118 TIAIN 43,8522

Although this review has emphasized the results of cohort and case-control studies, descriptive data on lung cancer mortality in the United States are consistent with a beneficial effect of the declining prevalence of cigarette smoking. Devesa, Blot, and Fraumeni (1989) described declining mortality rates for lung cancer at ages below 45 years. The decreases were greatest among white men but also occurred among white women and blacks of both sexes. Effect of Antecedent Smoking History The preceding Section reviewed epidemioiogic studies describing the pattern of lung cancer risk following smoking cessation. This Section considers factors related to smoking that plausibly could modify the effect of cessation on lung cancer risk; these factors include the duration of smoking, daily cigarette consumption, inhalation prac- tices, types of tobacco products smoked, and age at cessation. Duration of Smoking Duration of smoking prior to cessation is a potentially important modifier of the pattern of risk reduction in ex-smokers. Graham and Levin (1971) examined the risk of lung cancer associated with increasing durations of abstinence and with stratification by duration of smoking (<_30 or ?3 I years and <_40 or -41 years). The decline in risk associated with stopping was greater for those who had smoked for shorter periods than for those who had smoked for loneer periods. Similar results were reported by Lubin and colleagues (1984a), who determined the risk of developing lung cancer by time since stopping smoking (0, 1-4, 5-9, and -10 years) and total duration of smoking (1-19, 20-39, 40-49, and -50 years). In each category of smoking duration, the risk of developing lunQ cancer decreased as the number of years since stopping smoking increased, but the rate of decline was greater among those who had smoked for a shorter time. Among men who had smoked for 1 to 19 years. the risk of developing lung cancer after 10 years of abstinence dropped to less than one-third of that among current smokers. On the other hand, for men who had smoked 50 years or more and stopped for at least 10 years, the risk was still 90 percent of that for men who continued to smoke. This analysis, which matched for age and controlled for both duration of smoking and length of abstinence, introduces too many variables for the temporal dimensions of cigarette use (Chapter 2). By simultaneouslv considering attained age, duration of smoking, and length of abstinence, the analytic model incorrectly forces former smokers to have a younger aQe of starting to smoke than current smokers. In a case-control study in Sweden, Damber and Larsson (1986) also found higher relative risks amona former smokers of pipes and cigarettes who had smoked longer. Brown and Chu (1987) suggested that failure to adjust for previous duration of smoking may result in risk estimates for former smokers that are too low and thus exaggerate the benefits of smoking cessation. Based on reanalysis of data from the large European case-control study. Brown and Chu (1987) reported that the correlation between duration of smoking and time since stopping smoking for ex-smokers was--0.6, indicatinQ that men who had stopped smoking for many years had also smoked for less 122 TIMN 438526

Cessation After Developing Disease Individuals who stopped smoking are not a randomly selected group in most studies (Chapter 2). Often, smokers quit as a result of developing symptoms of a life- threatening disease or immediately after diagnosis of cancer. This phenomenon is evidenced by the increase in risk of lung cancer in the immediate period after cessation. Some studies have grouped these former smokers with the continuing smokers or have excluded them from the analysis. A few epidemiologic studies have assessed the risk of lung cancer among those who quit for health reasons and for non-health-related reasons. In the U.S. Veterans Study, about 10 percent of the smokers quit because of a doctor's orders; these smokers were presumably ill. The lung cancer mortality ratio relative to never smokers for ex- smokers who stopped because of non-health-related reasons was 4.43 compared with 5.83 among ex-smokers who stopped on a doctor's orders and 8.98 among continuing smokers (Kahn 1966). In the European case-control study, Brown and Chu (1987) reported that the relative risk of lung cancer for those who stopped smoking because of health reasons compared with those who stopped for reasons other than health was 1.3 (p<0.001). Moreover, the percentage who stopped for health reasons decreased with increasing years of abstinence. Among those who had stopped for 1 year or less, 95.8 percent stopped because of health reasons compared with 65.7 percent of longer term ex-smokers. In ACS CPS-II, men and women who did not have a history of heart disease, stroke, or cancer at the time of interview showed a decreased risk of luna cancer in the first 2 years after smoking cessation when compared with continuing smokers. In contrast, the risks for all subjects combined (i.e.. those with and without a history of previous chronic disease) were increased during the first 2 years after smoking cessation when compared with continuing smokers. The lower risks among the group with no history of previous disease compared with the total group persisted for subsequent periods of smoking abstinence (Table 7). ~ Cessation After Diagnosis of Lung Cancer Two studies examined the relationship between smoking status and treatment out- come of patients with small cell lung cancer. In the study by Johnston-Early and associates (1980), survival was prolonged in patients who were ex-smokers or who had stopped smoking at diagnosis, whereas no difference in survival by smoking status was detected in the study by Bergman and Sorenson (1988). The study by Johnston-Early and colleagues (1980) involved 112 patients with small cell lung cancer; 20 had stopped smoking before diasnosis; 35 had stopped at diaenosis; and 57 continued smoking. Therapies included chemotherapy with radiation therapy, with or without thymosin fraction V. The three patient groups were similar in disease extent, pretreatment performance status, pack=years smoked, and age and sex distribu- tion. The patients who had stopped smoking prior to diagnosis had the best survival, followed by those who had stopped at diagnosis, and finally by those who continued smoking; the median survival for the three groups was 70, 52, and 47 weeks, respec- tively. Overall survival differences remained after individually adjusting for disease TIMN 438533 129

smoking status groups were poorer than for the comparable smoking status groups in the study by Johnston-Early and associates (1980). LARYNGEAL CANCER Pathophysiologic Framework Smoking has been firmly established as a cause of laryngeal cancer (US DHHS 1982, 1989) based on numerous epidemiologic studies. These studies have employed diverse methodologies and have been performed in different countries and covered various time periods. Tobacco smoke exposure has been measured by number of cigarettes smoked per day, number of years of smoking, age when started to smoke, type of cigarettes smoked, and depth of inhalation (US DHHS 1982). In the larynx, as in the bronchus, a sequence of histologic changes occurs with continued smoking. These changes progress from cells with atypical nuclei, to car- cinoma in situ, to invasive carcinoma. Autopsy studies show that recovery of the laryngeal epithelium can follow smoking cessation. Auerbach, Hammond, and Gar- finkel (1970) studied postmortem specimens of laryngeal epithelium from 942 men (644 current cigarette smokers, 94 cigar and/or pipe smokers, 116 ex -cigarette smokers, and 88 never smokers). Ex-smokers in this study had stopped smoking for at least 5 years. Compared with current smokers, ex-smokers showed fewer histologic changes; 75 percent of ex-smokers and never smokers showed no cells with atypical nuclei, whereas almost all current smokers showed some cells with atypical nuclei. Similar findings were reported by Muller and Krohn (1980), who obtained laryngeal epithelial specimens from autopsy. Of the 148 cases in the study, 24 were never smokers and 24 were ex-smokers who had stopped smoking for at least 5 years. Table 8 shows the relative distribution of selected histologic features by smoking status. Occurrence of all histologic changes was lowest among never smokers, intermediate among ex-smokers, and highest among current smokers. However, the histologic findings of ex-smokers in this study were more similar to those of light current smokers (<10 cig/day) than to those of never smokers. Smoking Cessation and Laryngeal Cancer Risk A few studies provide data on the relationship between smoking cessation and risk of laryngeal cancer (Table 8). Former smokers are at less risk than current smokers, but have about six times the risk of never smokers. The relative risk of laryngeal cancer is higher immediately after smoking cessation (i.e., 1-3 years after quitting) compared with continuing smokers. However, after approximately 3 to 4 years of smoking abstinence, former smokers show lower relative risks with increasing years of smoking abstinence (Table 8). Based on a case-control study of laryngeal and hypopharyngeal cancer conducted in Europe, Tuyns and colleagues (1988) suggested that the benefit of smoking cessation seemed to appear sooner after cessation for cancer of the hypopharynx/epilarynx than for the larynx. TIMN 438535 131

50 40 30 20 MEN Continuing smokers a Former smokers I I Never smokers WOMEN - FIGURE 2.-Estimated probability of dying in the next 16.5-yr interval for quitting at ages 55-59 compared with never smoking and continuing to smoke, by sex NOTE: Continuing and former smokers include only those smoking _21 (men) or >20 (women) cig/day. Vertical bars represent 95% Cl; the interval for female never smokers is not shown because it is extremely narrow ( I I--I I%). Based on American Cancer Society Cancer Rrevention Study ll data for persons without a history of cancer, heart disease, or stroke who were not "sick" at enrollment. SOURCE: Unpublished tabulations, American Cancer Society, (see Table 8).

TABLE 8.-Histologic changes in laryngeal epithelium by smoking status Histologic change (% relative frequencies) Smoking status Normal squamous epithelium Keratinizing squamous epithelium Hyperplastic squamous epithelium Squamous metaplasia Never smokers 83 4 8 21 Ex-smokers 54 33 29 33 Current smokers Light 56 25 12 58 Moderate 46 36 26 46 Heavy 31 44 33 52 SOURCE: Abytracted from text and figures 2-5 in Muller and Krohn (1980). Risk reduction pattern by years of smoking abstinence and number of cigarettes smoked daily was examined in a few studies (Table 9). In the U.S. Veterans Study, the risk of death from laryngeal cancer was lower among ex-smokers who smoked 10 to 20 or 21 to 39 cigarettes per day than among current smokers, but it was not lower among those smoking I to 9 or 40 cigarettes or more per day. However, there were very few laryngeal cancer deaths in the lowest and highest consumption levels (two and one, respectively) (Kahn 1966). In ACS CPS-II, ex-smokers who smoked less than 21 cigarettes per day showed a greater reduction in laryngeal cancer mortality for all durations of smoking abstine;nce compared with ex-smokers who smoked 21 cigarettes or more per day relative to current smokers. In a case-control study conducted in the Texas Gulf Coast region (Falk et al. 1989), there was no consistent pattern of greater proportion of reduction in risk among those who had smoked fewer cigarettes per day prior to smoking abstinence. Moreover, there was still a threefold increased risk among those who had smoked more than 30 cigarettes daily after 10 years of smoking abstinence (Table 9). ~ The effect of smoking duration prior to smoking cessation was not considered in the studies mentioned above. There is some indication that the average age at which the ex-smoker developed clinical laryngeal cancer was about 10 years older (68.7) than that of the current smoker (Wynder et al. 1976). Alcohol has been shown to have an independent effect on risk of laryngeal cancer, but the relationship is weaker than the one between smoking and laryngeal cancer. The relative risks for joint exposure to alcohol and tobacco are consistent with a multiplica- tive interaction of the two agents (Flanders and Rothman 1982: Elwood et al. 1984; Olsen. Sabroe, Fasting 1985). In this review of the literature, no studies were found that accounted for the effects of alcohol intake in examining risk of laryngeal cancer after smoking cessation. 132 rrIMN 438536

fit of the multistage model to data from ACS CPS-I and the U.S. Veterans Study. These researchers observed that crude rates of lung cancer decreased with increasing years of smoking abstinence although the trend was less steep when average amount of smoking and ages when smoking started and stopped were considered in the analysis. Moreover, the observed lung cancer rates among ex-smokers were compared with the expected rates, which were computed in three ways-risk at the time of quitting, risk at current age with excess risk frozen at the time of quitting, and never smokers of the same age. For each comparison approach, the ratio of observed to expected rates decreased with increasing years of smoking abstinence. Freedman and Navidi (1989) concluded that this pattern was incompatible with the multistage model, which predicts stabilization of excess risk when an individual stops smoking. Gaffney and Altshuler (1988) reexamined data from the British Physicians Study and found that the best-fitting model among current smokers predicted an increase in the excess incidence among ex-smokers, which was inconsistent with the observed decreased rates. These researchers found that a two-stage model fit the incidence of lung cancer in both current smokers and ex-smokers. Gaffney and Altshuler (1988) then proposed a two-stage model with clonal growth in which cigarette smoke induced the initial transition and promoted clonal growth in these cells initiated by cigarette smoke. Moolgavkar, Dewanji, and Luebeck (1989) questioned the biologic plausibility of the proposal by Gaffney and Altshuler (1988) and noted that their model only fit part of the British physicians data set, did not consider each age-smoking level, and discounted the possibility that smoking affected two transition rates in the carcinogenic process. Mooigavkar, Dewanji, and Luebeck (1989) reanalyzed the British Physicians Study within the framework of the two-mutation, recessive oncogenesis model. Based on this model, the second-mutation rate would be affected by smoking, and a sudden decline in risk after cessation of smoking would be predicted. However, this model implies that smoking affects the last stage in a multistage process, contrary to current considera- tions. In summary, multistage models have been used to describe the interrelationships among number of cigarettes smoked daiiy, duration, time since exposure ended, and lung cancer incidence. Several investigators have interpreted the data on risk among former smokers in different ways. The epidemiologic data clearly indicate that the risk among former smokers is between that of continuing smokers and never smokers. Various models can be fit to the different data sets. The expected pattern of risk among former smokers is sensitive to the model selected and dependent on the relative magnitude of the effect of smoking on early versus late stages of the process of carcinogenesis. Usina multistaQe models, the data on former smokers are insufficient to allow precise quantification of the relative effects of smoking on the early and late stages of the carcinogenic process. which smoking is assumed to affect. Nevertheless, data indicate that smoking has an effect on the late stages of the carcinoQenic process and that cessation reduces lung cancer occurrence. 128 TIMN 438532

U.S. PUBLIC HEALTH SERVICE. Smcrkirq and Health. Report of the Adrisnrv Committee to the Sut-genrr General of the Public Health Service. U.S. Department of Health, Education, and Welfare, Public Health Service, Center for Disease Control. PHS Publication No. 1103, 1964. WHITTEMORE, A.S. Effect of cigarette smoking in epidemiological studies of lung cancer. Statistics in Medicine 7(1-2):223-238. January-February 1988. WIGLE. D.T., MAO, Y., GRACE, M. Relative importance of smoking as a risk factor for selected cancers. Canadian Journal of Public Health 71(4):269-275, July/August 1980. WU. A.H.. HENDERSON. B.E., PIKE, M.C., YU, M.C. Smoking and other risk factors for lung cancer in women. Journal of the National Cancer Institute 74(4):747-75 !, April 1985. WYNDER. E.L.. COVEY. L.S., MABUCHI, K.. MUSHINSKI, M. Environmental factors in cancer of the larynx. A second look. Cancer 38(4):1591-1601. October 1976. WYNDER, E.L., KABAT, G.C. The effect of low-yield cigarette smoking on lung cancer risk. Cancer 62(6):1223-1230, September 15, 1988. WYNDER. E.L., STELLMAN, S.D. Comparative epidemiology of tobacco-related cancers. Cancer Research 37(12):4608-4622, December 1977. WYNDER. E.L., STELLMAN, S.D. Impact of long-term filter cigarette usage on lung and larynx cancer risk: A case-control study. Journal nftheNational Cancerlnstitute 62(3):47 1 - 477, March 1979. Z'Il'VIIN 438544 141

COHEN, S., LICHTENSTEIN, E., PROCHASKA, J.O., ROSSI. J.S., GRITZ. E.R., CARR, C.R., ORLEANS, C.T., SCHOENBACH, V.J., BIENER, L., ABRAMS, D., DICLEMENTE, C., CURRY, S., MARLATT, G.A., CUMMINGS, K.M., EMONT. S.L., GIOVINO, G., OSSIP-KLEIN. D. Debunking myths about self-quitting. Evidence from 10 prospective studies of persons who attempt to quit smoking by themselves. American Psychologist 44(11):1355-1365, November 1989. COHEN, S.J., KATZ, B.P., DROOK, C.A., CHRISTEN, A.G., MCDONALD. J.L.. OLSON, B.L., CLOYS, L.A., STOOKEY, G.K. Overreporting of smokeless tobacco use by adolescent males. Journal of Beharioral Medicine 11(4):383-393, August 1988. COLE, P., MORRISON, A.S: Basic issues in population screening for cancer. Journal of the National Cancer Institute 64(5):1263-1272, May 1980. COOK. D.G.. POCOCK, SJ., SHAPER, A.G.. KUSSICK. S.J. Giving up smoking and the risk of heart attacks. A report from the British Regional Heart Study. Lanrer 2(8520):1376-1380, December 13, 1986. COPELAND, K.T., CHECKOWAY, H., MCMICHAEL, A.J., HOLBROOK, R.H. Bias due to misclassification in the estimation of relative risk. American Journal of Epidemiology 105(5):488-495, May 1977. CORNONI-HUNTLY, J., BARBANO. H.E.. BRODY. J.A.. COHEN, B.. FELDMAN. J.J., KLEINMAN. J.C., MADANS. J. National Health and Nutrition Examination I Epidemiol- ogy Follow-up Survey. Public Health Reports 98:245-25 l, 1983. COULTAS, D.B.. HOWARD, C.A.. PEAKE. G.T.. SKIPPER. B.J.. SAMET. J.M. Salivary cotinine levels and involuntary tobacco smoke exposure in children and adults in New Mexico. American Review of Respiratory Disease 136(2):305-309, 1987. COULTAS, D.B.. HOWARD, G.A., PEAKE. G.T., SKIPPER. B.J.. SAVIET, J.M. Discrepan- cies between self-reported and validated cigarette smoking in a community survey of New Mexico Hispanics. American Review of Respiratory Disease 137:810-814, 1988. CROSSEN, J.R.. DOUGHER. M.J., BELEW, J. Comparison of reactive and non-reactive measures of smoking cessation at follow-up. Addictive Behaviors 9(3 ):295-298, 1984. CUMMINGS, K.M. Changes in the smoking habits of adults in the United States and recent trends in lung cancer mortality. Cancer Detection and Prevention 7:125-134, .1984. CUMMINGS, S.R.. RICHARD. R.J. Optimum cutoff points for biochemical validation of smoking status. American Journal of Puhlic Health 78(5):574-575. May 1988. DELARUE. N.C. A study in smoking withdrawal. Canadian Journal of Public Health 64(Supplement):S5-S 19, March-April 1973. DICLEMENTE. C.C., PROCHASKA. J.O. Self-change and therapy change of smoking be- havior: A comparison of processes of change in cessation and maintenance. Addictive Behaviors 7(2):133-142, 1982. ~ DIGIUSTO, E., ECKHARD.1. Some properties of saliva cotinine measurements in indicating exposure to tobacco smoking. American Journal of Public Health 76(10):1245-1246, Oc- tober 1986. DOLL, R., PETO. R. The causes of cancer: Quantitative estimates of avoidable risks of cancer in the United States today. Journal of the rVational Cancer lnstitrue 66(6): I 191-1308. June 1981. EVANS. R.I., HANSEN, W.B., MITTELMARK. M.B. Increasing the validity of self-reports of smokincr behavior in children. Journal ofApplied Psychology 62(4):521-523, April 1977. FERRIS, B.G. Epidemiology standardization project. American Review of Respiratory Disease 118(Supplement 6):1-120,1978. FLEISS, J.L. Statistical Methods for Rates and Proportions, Second Edition. New York: John Wiley and Sons, 1981. FLETCHER, R.H.. FLETCHER, S.W., WAGNER, E.H. Clinical Epidemiology: The Essen- tials, Second Edition. Baltimore: Williams and Wilkins, 1988. 60 TIMN 438470

x - 1,000 z w ~ 100 ~ 0 0 ~ u U z ~ ~ 10 U z YEARS 1 1 20 x Cigarette smokers by duration of smoking CigaretYe smokers by age Never smokers by age L 30 1 40 I 1 1I 50 60 70 80 FIGURE 3.-Incidence of bronchial carcinoma among continuing cigarette smokers in relation to age and duration of smoking and among never smokers in relation to age, double logarithmic scale SOURCE: Doll ( l971), with correction of printing error in the original figure. and Chu (1987) reexamined data on ex-smokers from the European case-control study of lung cancer (Lubin et al. 1984a) and concluded that smoking had an almost double relative effect on late-stage events compared with first-stage events. Using data from a case-control study in New Mexico, Whittemore (1988) developed a predictive model for lung cancer that showed a twofold stronger effect on late-stage than on early-stage events; the model overpredicted cases among ex-smokers and underpredicted cases among current smokers. Therefore, Whittemore suggested that smoking may have an even stronger effect on late-stage events than was assumed in the model. Alternative models and interpretation of data on former smokers and lung cancer have also been suggested in several recent studies. Freedman and Navidi (1989) tested the 'I'IlVIN 438531 127

FORTMANN, S.P., ROGERS, T., VRANIZAN, K., HASKELL, W.L., SOLOMON, D.S., FARQUHAR, J.W. Indirect measures of cigarette use: Expired-air carbon monoxide versus plasma thiocyanate. Preventive Medicine 13(1):127-135, January 1984. FREDERIKSEN, L.W., MARTIN, J.E. Carbon monoxide and smoking behavior. Addictive Behaviors 4(1):21-30, 1979. FREDERIKSEN, L.W., MARTIN, J.E., WEBSTER, J.S. Assessment of smoking behavior. Journal of Applied Behavioral Analysis 12(4):653-664, Winter 1979. FRIEDMAN, G.D., SIEGELAUB, A.B., DALES, L.G., SELTZER, C.C. Characteristics predictive of coronary heart disease in ex-smokers before they stopped smoking: Comparison with persistent smokers and nonsmokers. Journal of Chronic Diseases 32:175-190, 1979. GARVEY, A.J., BOSSE, R., GLYNN, R.J., ROSNER, B. Smoking cessation in a prospective study of healthy adult males: Effects of age, time period, and amount smoked. American Journal of Public Health 73(4):446-450. April 1983. GARVEY, A.J., HEINOLD. J.W., ROSNER, B. Self-help approaches to smoking cessation: A report from the normative aging study. Addictive Behaviors 14:23-33, 1989. GILLIES, P.A., WILCOX, B., COATES, C., KRISTMUNDSDOTTIR, F., REID, D.J. Use of objective measurement in the validation of self-reported smoking in children aged 10 and 1 1 years: Saliva thiocyanate. Journal of Epidemiology and Communin Health 36(3):205-208, 1982. GLASGOW, R.E.. KLESGES. R.C., GODDING. P.R., VASEY, M.W., 0'NEILL, H.K. Evaluation of a worksite-controlled smoking program. Journal of Consulting and Clinical Psyc•hologr 53( I): ! 37-138, February 1984. ~ GLYNN, S.M.. GRUDER, C.L.. JEGERSKI. J.A. Effects of biochemical validation of self- reported cigarette smoking on treatment success and on misreporting abstinence. Health Psychology 5(2): L25-136, 1986. GORDIS, L. Should dead cases be matched to dead controls'? American Journal of Epidemiol- ogv 1 l5(1):1-5. January 1982. GREENBERG, R.A.. HALEY, N.J., ETZEL. R.A., LODA. F.A. Measuring the exposure of infants to tobacco smoke. Nicotine and cotinine in urine and saliva. New En;land Jinurnal of Medicine 310( l7):1075-1078. April 26. 1984. GREENBERG. R.S.. LIFF. J.M.. GREGORY• H.R., BROCKMAN. J.E. The use of interviews with surrogate respondents in a case-control study of oral cancer. Yale Journal of Biologr and Medicine 59(5):-197-504, September-October 1986. GREENLAND, S. Response and follow-up bias in cohort studies. American Journal of Epidemiology 106(3):184-187, September 1977. GREENLAND. S. The effect of misclassification in the presence of covariates. American Journal of Epidemiology 112(4):564-569, October 1980. GREENLAND, S. Statistical uncertainty due to misclassification: Implications for validation substudies. Journal of Clinical Epidemiology 41(12):1167-1 174, 1988. HADDOW. J.E., KNIGHT. G.J., PALOMAKI, G.E.. KLOZA. E.M., WALD, N.J. Cisarette consumption and serum cotinine in relation to birthweieht. British Journal of Obstetrics and Grnaec•olo,gti 94(7):678-68 1. July 1987. HADDOW, J.E., PALOMAKI. G.E.. KNIGHT. G.J. Use of serum cotinine to assess the accuracy of self reported non-smoking. (Letter.) British Medical Journal 293(6557): 1306. November 15. 1986_ HALEY, N.J.. AXELRAD. C.M.. TILTON, K.A. Validation of self-reported smoking behavior: Biochemical analyses of cotinine and thiocyanate. American Journal of Public Health 73( l0):1204-1207, October 1983. HALEY, N.J., COLOSIMO. S.G.. AXELRAD, C.M., HARRIS, R., SEPKOVIC, D.W. Biochemical validation of self-reported exposure to environmental tobacco smoke. Environ- mental Research 49( I): l 27-135. June 1989. TIMN 438471 61

CONTENTS Introduction ....................................................... 147 Review of Specific Sites ............................................. 147 Oral Cancer ...................................................... 147 Esophageal Cancer ................................................ 152 Pancreatic Cancer .................................................155 Bladder Cancer ...................................................159 Cervical Cancer ................................................... 165 Breast Cancer .................................................... 169 Endometrial Cancer ............................................... 169 Other Cancer Sites ................................................ 172 Multiple Primary Cancers ............................................ 176 Summary ......................................................... 177 Conclusions ....................................................... 178 References ........................................................ 179 TIMN 438546 145

TAIjLE 1.--Stu(lies of oral cancer and smoking cessation ® Risk relative to never smokers Reference Populcuion(yrof data collection) Design (number of subjects) Gender Current smokers Former smokers Yr since quitting Comments Kahn US veterans Prospective Male 3.8 1.9 NP Excludes "doctor's orders" (1966) Cederlof el al. (1954--62) Sweden (248.195) Prospective quitters Cancer mortality (1975) (1 963-72) (27,300) Male 2.7 0.8 NP Cancer incidence (27,700) Female 2.0 0 NP Wynder and Stetlm;tn 6 US cities Casexuntrul Male 8.9 9.0 1-3 (t977) (1969-75) (497:6,534) 3.5 4-6 3.2 7-10 3.4 11-15 1.6 ~16 (270:6.522) Female 4.4 3.8 1-3 2.2 4-b 1.4 7-10 0.6 11-I5 0.8 _ 16 Rogot and Murray lIS veterans Prospective Male 4.2 1.7 NP Excludes "doctor's orders" (Ir)H0) (I95-1-69) (293 ,15H) quitters Cancer mortality Extension of US Veterans Study

TABLE 1.-Histologic changes (%) in bronchial epithelium by smoking status Smoking status Current smokers Ex-smokers Never smokers Sections with I or more epithelial lesions 97.8 66.6 25.7 Cilia present on 3 or more cell rows 92.7 57.3 12.1 Cilia absent 20.5 15.1 14.8 Atypical cells present 93.2 6.0 1.2 Unciliated atypical cells 19.0 0.9 0.1 SOURCE: Auerbach et al. (1962b). showed fewer lesions, suggesting that the number of lesions decreased rather than merely failed to increase after cessation of smoking. Auerbach and colleagues (1964) also reported that among cigarette smokers, there was a high degree of association between all types of histologic changes in the bronchi and in the lung parenchyma. However, the lungs of ex-smokers were more similar to those of never smokers than to those of current smokers with respect to cells with atypical nuclei. In this study of 46 ex-smokers, 32 had few atypical cells in their bronchial epithelium. Auerbach and associates (4964) suggested that with cessation of smoking, cells with atypical nuclei gradually disappeared from the bronchial epithelium and were replaced with normal cells. Other Changes Several reports have described levels of DNA adducts formed by the combination of chemical carcinogens or their metabolites with DNA in the tissues of never, former, and current smokers. Decline of DNA adduct levels in human lungs after smoking cessation has been reported by Phillips and coworkers (1988). These investigators utilized autoradiographs of chromatograms of 32P-postlabeled digests of DNA from lungs of current, former, and never smokers. A linear relationship was observed between number of cigarettes smoked per day and DNA adduct levels (Pearson correlation coefficient, r=0.72, p<0.001). In addition, ex-smokers who had quit smok- ing 1 to 3 months previously had adduct levels typical of the current smokers (12-14 adducts/108 nucleotides), whereas those who had not smoked for 5 years or more had adduct levels similar to those of never smokers (1.7-4.9 adducts/10g nucleotides). These investigators suggested that the reduced risk of lung cancer among ex-smokers may be due to loss of the promutagenic lesions that initiate the process, in addition to late-stage effects. Randerath and colleagues (1989) also used a 32P-postlabeling assay to study DNA damage in relation to cigarette smoking. Adduct profiles and levels were determined in nontumorous surgical specimens taken from patients with lung or taryngeal cancer. TIMN 438513 109

men, the relative risk for former smokers after stopping smoking for 10 years or more was 0.4 for filter cigarette smokers, 0.3 for nonfilter cigarette smokers, and 0.5 for mixed filter and nonfilter cigarette smokers. These data were collected in five western European countries from 1976 to 1980; the tar yields of the products smoked were relatively high in comparison with cigarettes currently smoked in the United States (Lubin et al. 1984b). In most studies, cigar and pipe smokers have lower lung cancer risks compared with cigarette smokers (US DHHS 1982). Former smokers of only pipes or cigars also showed an intermediate risk of lung cancer compared with current smokers and never smokers of these tobacco products (Table 6). In the U.S. Veterans Study, the lung cancer mortality ratio, compared with never smokers, was 1.67 among current smokers who used only pipes or cigars and 1.50 among former smokers (Kahn 1966). In a case-control study of smoking-related cancers conducted in the United States, Higgins, Mahan, and Wynder (1988) reported that ex-smokers of cigars only showed a relative risk of 2.5 compared with 3.1 among current smokers of cigars only. The relative risk was 0.7 amonQ ex-smokers of pipes only compared with 1.9 among current pipe smokers only. ~Analysis of the pattern of risk among ex-smokers of cigars and pipes only by considering the amount and duration smoked prior to smoking cessation revealed similar patterns of risk reduction among light and heavy smokers. Lubin, Richter, and Blot (1984) also examined the pattern of risk reduction by years of smoking abstinence (0, 1-4, _5 years) and types of tobacco smoked (cigars only, mixed cigar and cigarette smokers, pipes only, and mixed pipe and cigarette smokers). No apparent differences were observed in the estimated risks, when analyzed by tobacco products, among those who had stopped smoking for at least 5 years, but the numbers of cases who smoked cigars only and pipes only were quite small. On the other hand, Damber and Larsson (1986) reported that the decrease in relative risk atnong ex-smokers was less pronounced in smokers of pipes compared with cigarette smokers only in a case-control study conducted in Sweden. However, in this population, the risk of lunQ cancer for pipe smokers (RR=6.9) was similar to that of cigarette smokers (RR=7.0).~ In summary, these analyses, limited by the sample sizes within strata of types of products smoked, do not characterize precisely the changing lung cancer risk following cessation for smokers of various tobacco products. y y Effect of Age at Cessation Several researchers have suggested that the reduction in risk after smoking cessation may differ by age at cessation. Wynder and Stellman (1979) reported that the reduction in risk after cessation was appreciably greater for people aged 50 to 69 than for those 70 or older. However, only data for those aged 50 to 69 were presented in this publication. Pathak and associates (1986) also reported a strong interaction between age and duration of ciaarette smoking. Risk of lung cancer among ex-smokers was compared with that of current smokers with adjustment for the amount smoked. For ex-smokers less than 65 years of age, the estimated relative risks compared with current smokers declined to 0.49, 0.24, and 0.06 for 5, 10, and 20 years of smoking abstinence. N125 ,- ~~M ~~MN 438529

MURRAY, D.M., PERRY, C.L. The measurement of substance use among adolescents: When is the "bogus pipeline" method needed? Addictive Behaviors 12(3):225-233, 1987. MYERS, A.H., ROSNER, B., ABBEY, H., WILLET, W., STAMPFER, M.J., BAIN, C., LIPNICK, R., HENNEKENS, C., SPEIZER, F. Smoking behavior among participants in the Nurses' Health Study. American Journal of Public Health 77(5):628-630, May 1987. NANJI, A.A., LAWRENCE, A.H. Skin surface sampling for nicotine: A rapid, noninvasive method for identifying smokers. International Journal of the Addictions 23(11):1207-1210, 1988. NATIONAL INTERAGENCY COUNCIL ON SMOKING AND HEALTH. Guidelines for Research on the Effectiveness of Smoking Cessation Programs: A Committee Report. Chicago: American Dental Association, October 1974, pp. 46. NEATON, J.D., BROSTE, S., COHEN, L., FISHMAN, E.L., KJELSBERG, M.O., SCHOEN- BERGER. J. The Multiple Risk Factor Intervention Trial (MRFIT). VII. A comparison of risk factor changes between the two study groups. Preventive Medicine 10(4):519-543, July 1981. OCKENE, J.K.. HYMOWITZ, N., SEXTON, M., BROSTE, S.K. Comparison of patterns of smoking behavior change among smokers in the Multiple Risk Factor Intervention Trial (MRFIT). Preventive Medicine 11:621-638, 1982. OCKENE, J.K.. KRISTELLE. J.K., GOLDBERG, R., OCKENE. I., BARRETT. S., MER- RIAM, P. A smoking intervention in patients with coronary artery disease: Results of a randomized clinical trial. Presented at Society of Behavioral Medicine, San Francisco, California, April 1989. OCKENE, J.K.. KULLER, L.H., SVENDSEN, K.H., MEILAHN, E. The relationship of smoking cessation to coronary heart disease and lung cancer in the Multiple Risk Factor Intervention Trial (MRFIT). American Journal of Public Health 80(8):954-958, August 1990. OHLIN, P., LUNDH. B.. WESTLING. H. Carbon monoxide blood levels and reported cessation of smoking. Psyc•hopharmacologti 49:263-265. 1976. ORLEANS, C.S.. SHIPLEY, R.H. Assessment in smoking cessation research: Some practical guidelines. In: Keefe. F.J., Blumenthal, J.A. (eds.) Assessment Strategies in Behavioral Medicine. New :',,,K: Grune and Stratton. 1982. PAXTON, R. The effects of a deposit contract as a component in a behavioural programme for stopping smoking. Behaviour Research and Therapy 18:45-50, 1980. PAXTON, R.. BERNACCA, G. Urinary nicotine concentrations as a function of time since last cigarette: Implications for detecting faking in smoking clinics. Behavior Therapy 10:523- 528, 1979. PECHACEK, T.F., FOX, B.H., MURRAY, D.M., LUEPKER, R.V. Review of techniques for measurement of smoking behavior. In: Matarazzo. J.D., Weiss, S.M., Herd, J.A., Miller, N.E. (eds.) Behavioral Health: A Handbook of Health Enhancement and Disease Prevention. New York: John Wiley and Sons, 1984, pp. 729-754. PECHACEK, T.F., LUEPKER. R.. JACOBS, D., FRASER, G., BLACKBURN. H. Effect of diet and smoking on serum and saliva thiocyanates. Cardiovascular Disease Epidemiology Newsletter 27:96, 1979. PECHACEK, T.F., IVIURRAY, D.M.. LUEPKER, R.V.. MITTELMARK, M.B.. JOHNSON, C.A., SHUTZ. J.:bt. Measurement of adolescent smoking behavior: Rationale and methods. Journal of Behavioral Medicine 7(1):123-140, March 1984. PERSHAGEN, G., AXELSON, 0. A validation of questionnaire information on occupational exposure and smoking. Scandinavian Journal of Work, Environment and Health 8(1):24-28, 1982. TIMN 438475 65

References ALEXANDER, C.S., KLASSEN, A.C. Drug use and illnesses among eighth grade students in rural schools. Public Health Reports 103(4):394-399, July-August 1988. AMERICAN CANCER SOCIETY. Unpublished tabulations. ANDERSSON, G., MALMGREN, S. Risk factors and reported sick leave among employees of Saab-Scania, Linkoping, Sweden, between the ages of 50 and 59. Scandinavian Journal of Social Medicine 14(1):25-30, 1986. ASHFORD, J.R. Smoking and the use of the health services. British Journal of Preventive and Social Medicine 27(1):8-17, February 1973. ASTRAND, N.-E., ISACSSON, S.-O. Back pain, back abnormalities, and competing medical, psychological, and social factors as predictors of sick leave, early retirement, unemployment, labour turnover and mortality: A 22 year follow up of male employees in a Swedish pulp and paper company. British Journal of Industrial Medicine 45(6):387-395, June 1988.• BALARAJAN, R., YUEN, P., BEWLEY, B.R. Smoking and state of health. British Medical Journa1291(6510):1682, December 14, 1985. BILLINGS, A.G., MOOS, R.H. Social-environmental factors among light and heavy cigarette smokers: A controlled comparison with nonsmokers. Addictive Behaviors 8(4):38 t-391, 1983. BLAIR, A., BLAIR, S.N., HOWE, H.G., PATE. R.R., ROSENBERG, M., PARKER. G.M., PICKLE, L.W. Physical, psychological, and sociodemographic differences among smokers, ex-smokers, and nonsmokers in a working population. Preventive Medicine 9(6):747-759, November 1980. BROD, M., HALL, S.M. Joiners and non-joiners in smoking treatment: A comparison of psychosocial variables. Addictive Behaviors 9(2):217-221, 1984. CARSTENSEN, J.M., PERSHAGEN, G.. EKLUND, G. Mortality in relation to cigarette and pipe smoking: 16 years' observation of 25,000 Swedish men. Journal of Epidemiologv and Communitv Health 41:166-172. 1987. CHARLTON, A., BLAIR. V. Absence from school related to children's and parental smoking habits. British Medical Journal 298(6666):90-92, January 14, 1989. CHRISTIE, D., ROBINSON, K., GORDON. I., WEBLEY, C., BISBY, J. Current mortality in the Australian petroleum industry: The healthy-worker effect and the influence of life-style factors. Medical Journal of Australia 147(5):222, 224-225, September 7. 1987. COUGHLIN, S.M. Prevalence of smoking at a large sugar cane plantation in Hawaii. Hawaii Medical Journal 46(12):468-473, December 1987. CUTLER, J.A., MACMAHON, S.W., FURBERG, C.D. Controlled clinical trials of dru~ treatment for hypertension: A review. Hypertension 13(5, Part 2):136-144, May 1989. ~ D'AGOSTINO, R.B., KANNEL, W.B., BELANGER, A.J., SYTKOWSKI, P.A. Trends in CHD and risk factors at age 55-64 in the Framingham Study. International Journal of Epidemiology 18(3, Supplement 1):S67-S72, 1989. DOLL, R., HILL, A.B. Mortality in relation to smoking: Ten years' observations of British doctors. British Medical Journal 1(5395):1399-1410, May 30, 1964a. DOLL. R., HILL, A.B. Mortality in relation to smoking: Ten years' observations of British doctors. British Medical Journal 1(5396):1410-1467, June 6. 1964b. DOLL. R., PETO, R. Mortality in relation to smoking: 20 years' observations of male British doctors. British Medical Jour-na12:1525-1536, December 25, 1976. FREEBORN, D.K., MULLOOLY, J.P., POPE, C.R., MCFARLAND, B.H. Smoking and consistently high use of medical care among older HMO members. American Journal of Public Health 80(5):603-605, May 1990. FRIEDMAN, G.D., PETITTI, D.B., BAWOL, R.D., SIEGELAUB, A.B. Mortality in cigarette smokers and quitters. New England Journal of Medicine 304(23):1407-1410, June 4, 1981. 99 TIMN 438505

SEIDELL. J.C., BAKX, K.C., DEURENBERG, P., BUREMA, J., HAUTVAST, J.G., HUYGEN, F.J. The relation between overweight and subjective health according to age, social class, slimming behavior and smoking habits in Dutch adults. American Journal of Public Health 76(12):1410-1415, December 1986. SELTZER, C.C., OECHSLI, F.W. Psychosocial characteristics of adolescent smokers before they started smoking: Evidence of self-selection. A prospective study. Journal of Chronic• Diseases 38(1):17-26, 1985. STELLMAN, S.D., GARFINKEL, L. Smoking habits and tar levels in a new American Cancer Society prospective study of 1.2 million men and women. Journal of the National Canc•et- Institute 76(6):1057-1063, June 1986. STEPHENS, T. Health practices and health status: Evidence from the Canada Health Survey. American Journal of Preventive Medicine 2(4):209-215. July-August 1986. TUOMILEHTO, J., GEBOERS, J., SALONEN, J.T., NISSINEN, A.. KUULASMAA, K., PUSKA, P. Decline in cardiovascular mortality in North Karelia and other parts of Finland. British Medical Journal 293:1068-1071, October 25, 1986. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking for Women. A Report of the Surgeon General. U.S. Department of Health and Human Services. Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health, 1980. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking: Cancer and Chronic Lung Disease in the Workplace. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Office on Smoking and Health. DHHS Publication No. (PHS) 85-50207, 1985. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. Reducing the Health Conse- quences of Smoking: 25 Years of Progress. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service. Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion, Office on Srnoking and Health. DHHS Publication No. (CDC) 89-8411, 1989. U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE. Smoking and Health, A Report of the Surgeon General. U.S. Department of Health, Education, and Welfare, Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health. DHEW Publication No. (PHS) 79-50066, 1979. U.S. PUBLIC HEALTH SERVICE. Smoking and Health. Report of the Advisory Committee to the Surgeon General of the Puhlic- Health Service. U.S. Department of Health. Education. and Welfare, Public Health Service, Center for Disease Control. PHS Publication No. 1103, 1964. U.S. PUBLIC HEALTH SERVICE. The Health Consequences of Smoking. A Public• Health Service Rei•ievv: 1967. U.S. Department of Health, Education, and Welfare, Public Health Service, Health Services and Mental Health Administration. PHS Publication No. 1696, revised 1968. U.S. PUBLIC HEALTH SERVICE. The Health Consequences of Smoking. 1969 Supplement to the 1967 Public Health Sen>ic•e Review. U.S. Department of Health, Education, and Welfare. Public Health Service. DHEW Publication No. 1969-2 (Supplement), 1969. VOGT, T.M., SCHWEITZER, S.O. Medical costs of cigarette smoking in a health maintenance organization. American Journal of Epidemiology 122(6):1060-1066, December 1985. WEINKAM, J.J., ROSENBAUM, W., STERLING, T.D. Smoking and hospital utilization. Social Science and Medicine 24(11):983-986, 1987. WORLD HEALTH ORGANIZATION EUROPEAN COLLABORATIVE GROUP. Multi- factorial trial in the prevention of coronary heart disease. 3. Incidence and mortality results. European Heart Journa14:141-147, 1983. 102 TIMN 438508

TABLE 4.-Conlinued Risk relative to never smokers Yr Population (yr of Design Current Former since Reference data collection) (number of subjects) Gender smokers smokers quiuing Comments Wynder and 6 US cities Case:control Male 2.2 2.6 1-3 Cases are from the same series Goldsmith (1977) (1969-74) (574:568) 2.9 4-6 as reported by Wynder and 1.5 7-9 Stellman (1977) 1.6 10-12 1.2 13-15 1.1 >16 (155:154) Female 2.2 2.5 1-6 1.2 >7 Vineis et al. Italy Case:control Mnle 6.0 3.7 3-9 (1983) (1978-81) (355:276) 3.6 10-14 2.1 >15 Cartwright et al. England Case:control' Male 1.6 1.0 6-15 (1983) (1978-81) (932:1.402) 1.1 16-25 0.9 >26 (327:579) Female 1,4 0.5 6-I5 0.5 >16 Morrison et al. Boston, MA Case:control (1984) (1976-77) (427:391) Mule 3.1" 1.5 >_I (165:142) Female 5.6" 3.4 >I Manchester, UK (398:490) Male 2,6" 1.8 >1 (1976-78) (155:241) Female 2.1" 0.7 >I Nagoya, Japan (224:442) Male 20" I.0 >I (1976-78) (66:146) Female 4.3" NP NP o\

TABLE 4.-Studies of bladder cancer and smoking cessation Risk relative to never smokers Reference Population (yr of data collection) Design (number of suhjects) Gender Current smokers Former smokers Yr since quitting Comments Kahn US veterans Prospective Malc 1.9 1.5 NP Excludes "doctor's orders" (1966) (1954-62) (248.195) quitters Cancer mortality Cederlof et al. Sweden Prospective Male 1.8 2.1 NP Cancer incidence (1975) (I963-72) (27.300) Female 1.0 ll NP (27,700) Wynder and Stellman 6 US cities Case:contrul Male 2.7 2.9 1-3 (1977) (1969-75) (541:6.534) 1.9 4-6 1.4 7-10 1.6 1I-15 1.1 - (6 (150:6.522) Female 2.4 3.1 1-3 1.5 4-6 0 7-10 1.5 11-15 2.4 _ 16 Wigle. Mao. Grace AlbLrta, Canada Caise:cuntrol Male 2.8 2.1 NP Adjusted forcumulutive (1980) (1971-73) ('0a:I.0U2) past dose (51:674) Fenr.de 3.5 3.1 NP Rogot and Murray US veterans Prospective Mide 2.2 1.4 NP Excludes "doctor's order" (19R0) (1954-69) (293,952{) . quitters Cancer mortality Extension of US Veterans Study

MULTIPLE RISK FACTOR INTERVENTION TRIAL RESEARCH GROUP. Multiple Risk Factor Intervention Trial. Risk factor changes and mortality results. Journal of the American Medical Association 248(12):1465-1477, September 24, 1982. MULTIPLE RISK FACTOR INTERVENTION TRIAL RESEARCH GROUP. Mortality rates after 10.5 years forpart_icipants in the Multiple Risk Factor Intervention Trial. Findings related to a priori hypotheses of the trial. Journal of the American Medical Association 263(13):1795-1801, April 4, 1990. MULTIPLE RISK FACTOR INTERVENTION TRIAL RESEARCH GROUP. 10.5 year mortality for participants in the Multiple Risk Factor Intervention Trial. Findings for sub- groups with hypertension at baseline. Submitted for publication. NEWCOMB, M.D., BENTLER, P.M. The impact of late adolescent substance use on young adult health status and utilization of health services: A structural-equation model over four years. Social Science and Medicine 24(1):71-82, 1987. OAKES, T.W., FRIEDMAN, G.D., SELTZER, C.C., SIEGELAUB, A.B., COLLEN, M.F. Health service utilization by smokers and nonsmokers. Medical Care 12(11):958-966, November 1974. OCKENE, J.K., KULLER, L.H., SVENDSEN, K.H., MEILAHN, E. The relationship of smoking cessation to coronary heart disease and lung cancer in the Multiple Risk Factor Intervention Trial (MRFIT). American Journal of Public Health 80(8):954-958, August 1990. OSTER, G., COLDITZ, G.A., KELLY, N.L. The economic costs of smoking and benefits of quitting for individual smokers. Preventive Medicine 13(4):377-389, July 1984. OSTRO, B.D. Estimating the risks of smoking, air pollution, and passive smoke on acute respiratory conditions. Risk Analysis 9(2):189-196, 1989. PEARSON, D.C., GROTHAUS, L.C.. THOMPSON, R.S., WAGNER, E.H. Smokers and drinkers in a health maintenance organization population: Lifestyles and health status. Preventive Medicine l6(6):783-795, November 1987. REED, W.L. Physical health status as a consequence of health practices. Journal of Community Health 8(4):217-228, Summer 1983. RICE, D.P., HODGSON, T.A., SINSHEIMER. P. The economic costs of the health effects of smoking, 1984. Milhank Quarterlr 64(4):489-547, 1986. RODE, A., ROSS, R., SHEPHARD. R.J. Smokina withdrawal programme. Personality and cardiorespiratory fitness. Archives of E'nrironmental Health 24(1):27-36, January 1972. ROGOT, E.. MURRAY, J.L. Smoking and causes of death among U.S. veterans: 16 years of observation. Puhlic• Health Reports 95(3):213-222, May-June 1980. ROSE, G., HAMILTON, P.J. A randomised controlled trial of the effect on middle-ased men of advice to stop smoking. Journal of Epidemiology and Community Health 32(4):275-28 1, December 1978. ROSE, G., HAMILTON, P.J.S., COLWELL. L.. SHIPLEY, M.J. A randomised controlled trial of anti-smoking advice: 10-year results. Journal of Epidemiology and Community Health 36(2):102-]08, June 1982. SALONEN, J.T.. TUOMILEHTO, J., NISSINEN, A., KAPLAN, G.A., PUSKA. P. Contribu- tion of risk factor changes to the decline in coronary incidence during the North Karelia project: A within community analysis. International Journal of Epidemiology 18(3):595-601, Sep- tember 1989. SEGOVIA, J.. BARTLETT, R.F., EDWARDS, A.C. The association between self-assessed health status and individual health practices. Canadian Journal ofPuhlic• Health 80(l):32-37, January-February 1989. TIMN 435507 101

TABLE g.--Continued Reference Population Smoking status Relative risks Wyn<!er and Ste!lman { 1977) 0 US cities Former smokers (yr since stopped) Males Females 1-3 17.9 6.9 4-6 8.5 2.6 7-1 0 4.0 - t1-15 3.4 8.8 ?16 2.5 ---- Current smokers 14.3 11.6 Never smokers t .t} 1.0 Tuyns et al. ( 1988) European countries Former smokers (yr since stopped) Males 1-4 Fnctolarynx Hypopharynx 1.51 1.~} S_t} 0.52 0.28 ?10 0.28 0.32 Current smokers 1.0 1.0 NQTC: ACS C'PS-tt=American Catac4r S4wiety Caeicer 1'rrventiwt Stuity It. "Reference category i:,, never snxokers.

References ALDERSON. M.R., LEE, P.N., WANG. R. Risks of lung cancer, chronic bronchitis, ischaemic heart disease, and stroke in relation to type of cigarette smoked. Journal of Epidemiology and Cammtrnitv Health 39(4):286-293, December 1985. AMERICAN CANCER SOCIETY. Unpublished tabulations. ARMITAGE. P. Discussion on paper by R. Doll. Journal of the Royal Statistical Society A 134:155-156, 1971. ARMITAGE. P., DOLL, R. The age distribution of cancer and a multi-stage theory of carcinogenesis. British Journal of Cancer 8: I-11, 1954. ARMITAGE, P., DOLL. R. A two-stage theory of carcinogenesis in relation to the age distribution of human cancer. British Journal of Cancer 11: 161-169, 1957. AUERBACH, O., GARFINKEL, L., HAMMOND, E.C. Relation of smoking and age to findings in lung parenchyma: A microscopic study. Chest 65(1):29-35, January 1974. AUERBACH, O., GERE. J.B., FORMAN. J.B., PETRICK, T.G., SMOLIN, H.J., MUEHSAM, G.E., KASSOUNY, D.Y., STOUT, A.P. Changes in the bronchial epithelium in relation to smoking and cancer of the lung. A report of progress. Nex, England Journal of Medicine 256(3):97-104, January 17, 1957. AUERBACH. 0.. HAMMOND, E.C.. GARFINKEL, L. Histologic changes in the larynx in relation to smoking habits. Cancer 25(1):92-104. January 1970. AUERBACH. 0.. HAMMOND, E.C., GARFINKEL, L., BENANTE. C. Relation of smoking and age to emphysema. Whole-lung section study. New England Journal of Medicine 286(16):853-857, Apri120, 1972. AUERBACH. 0.. STOUT, A.P.. HAMMOND, E.C.. GARFINKEL. L. Changes in bronchial epithelium in relation to sex, age, residence, smoking and pneumonia. New England Journal of Medicine 267(3 ):111-125, July 19, 1962a. AUERBACH. 0.. STOUT, A.P., HAMMOND, E.C., GARFINKEL. L. Bronchial epithelium in former smokers. New England Journal of Medicine 267(3):119-125. July 19. 1962b. AUERBACH. 0., STOUT. A.P., HAMMOND. E.C., GARFINKEL. L. Smoking habits and age in relation to pulmonary changes. Rupture of alveolar septums, fibrosis and thickening of walls of small arteries an¢ arterioles. New England Journal of Medicine 269(20):1045- 1054. November 14. 1963., AUERBACH, O., STOUT, A.P., HAMMOND, E.C., GARFINKEL, L. Interrelationships among various histologic changes in bronchial tubes and in lung parenchyma. American Review of Respirator:y Disease 90(6):867-876. December 1964. BENHAMOU, S., BENHAMOU, E., TIRMARCHE, lvi., FLAMANT, R. Lung cancer and use of cigarettes: A French case-control study. Journal of the National Cancer Institute 74(6):1169-1175, June 1985. BERGMAN, S., SORENSON, S. Smoking and effect of chemotherapy in small cell lung cancer. European Respiratorv Journal 1:932-937, 1988. BIRRER, M.J., MINNA, J.D. Molecular genetics of lung cancer. Seminars in Oncology 15(3):226-235, June 1988. BROWN, C.C., CHU, K.C. Use of multistage models to infer stage affected by carcinogenic exposure: Example of lung cancer and cigarette smokins. Journal of Chronic Diseases 40 (Supplement 2):171S-179S, 1987. ~ CANADIAN DEPARTMENT OF NATIONAL HEALTH AND WELFARE. A Canadian Study of Smoking and Health. Department of National Health and Welfare, Epidemiology Division, Health Services Branch, Biostatistics Division, Research and Statistics Directorate, 1966, 137 pp. rFIMN 438540 137

CHAPTER 6 SMOKING CESSATION AND CARDIOVASCULAR DISEASE TIMN 438587 187

PETO, J. Early- and late-stage carcinogenesis in mouse skin and in man. In: Borzsonyi, M., Lapis, K., Day, N.E., Yamasaki, H. (eds.) Models, Mechanisms and Etiology of Tumour Promotion. Lyon: IARC, 1984, pp. 359-370. PHILLIPS, D.H., HEWER, A., MARTIN, C.N., GARNER, R.C., KING, M.M. Correlation of DNA adduct levels in human lung with cigarette smoking. Nature 336(6201):790-792, December 22-29, 1988. RANDERATH, E., MILLER, R.H., MITTAL, D., AVITTS, T.A., DUNSFORD, H.A., RANDERATH, K. Covalent DNA damage in tissues of cigarette smokers as determined by 32P-postlabeling assay. Journal of the National Cancer Institute 8 1 (5):341-347, March 1, 1989. ` ~ ROGOT, E.. MURRAY. J.L. Smoking and causes of death among U.S. veterans: 16 years of observation. Public• Health Reports 95(3):213-222, May June 1980. SACCOMANNO, G., ARCHER. V.E., AUERBACH, 0., SAUNDERS. R.P., BRENNAN, L.M. Development of carcinoma of the lung as reflected in exfoliated cells. Cancer 31(1):256-270, January 1974. STENBACK, F., PETO, R., SHUBIK, P. Initiation and promotion at different ages and doses in 2200 mice. I. Methods, and the apparent persistence of initiated cells. British Journal of Canc•er 44(1):1-14, July 1981 a. STENBACK. F.. PETO. R.. SHUBIK. P. Initiation and promotion at different ages and doses in 2200 mice. II. Decrease in promotion by TPA with ageing. British Journal of Cancer 44(l):15-23, July 1981 b. STENBACK. F.. PETO, R.. SHUBIK, P. Initiation and promotion at different ages and doses in 2200 mice. III. Linear extrapolation from high doses may underestimate low-dose tumour risks. British Journal of Cancer 44(1):24-34, July 198 l c. TUYNS, A.J., ESTEVE. J., RAYMOND. L., BERRINO, F., BENHAMOU, E., BLANCHET, F., BOFFETTA, P., CROSIGNANI, P., DEL MORAL, A., LEHMANN, W., ET AL. Cancer of the larynx/hypopharynx, tobacco and alcohol: IARC International Case-Control Study in Turin and Varese (Italy), Zaragoza and Navarra (Spain), Geneva (Switzerland) and Calvados (France). International Journal of Cancer 41(4):483-49 1. April 15, 1988. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking: Canc•er. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Heath Service, Office on Smoking and Health. DHHS Publication No. (PHS) 82-50179, 1982. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking: Chronic Obstructive Lung Disease. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Office on Smoking and Health. DHHS Publication No. (PHS) 84-50205, 1984. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of lrvoluntary Smoking. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control. DHHS Publication No. (CDC) 87-8398, 1986. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. Reducing the Health Conse- quences of Smoking: 25 Years of Progress. A Report of the Sur;eon General. U.S. Department of Health and Human Services, Public Health Service. Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion. Office on Smoking and Health. DHHS Publication No. (CDC) 89-8411, 1989. U.S. DEPARTMENT OF HEALTH. EDUCATION, AND WELFARE. Smoking and Health. A Report of the Surgeon General. U.S. Department of Health, Education. and Welfare, Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health. DHEW Publication No. (PHS) 79-50066, 1979. TIMN 438543 140

. TA13Lr 3.-Studies of cancer of the pancreas and smoking cessation Risk relative to never smokers Itefcrence Pupulutiun(yruf data rullectiun) I)c.ign UI(Iniher ul, subjects) Gender Current smokers Former snwkers Yr since quitting Comments Kahn US velcr.uts Prospective Mole 1.6 1.2 NP Excludes "drxtor's orders" (1966) Cederlof et al. (1')54-fi2) Sweden ('_-It{.I'15) Pru.pective quitters Cancer mortality (1975) (1963-72) (373(IO) Male 2.5 1.7 NP Cancerincidence ('7.7(u)) remale 1.0 3.5 NP Rogot and Murray lIS vetrrans Prospective Male I.}{ 1.2 NP Excludes "da:tor's orders" (I9H()) acMahun cl al. (1954 69) oston. MA (?93,)52{) 'asexuntrol quitters Cancer mortality Extension of US Veterans Study (1981) (I97-1--79) (2 12i:106) Mule 1.3" 1.4 NP Wynder. I lall, 6 US cities (1-19:337) Casexuntrul Female 1.6" 1.3 NP Pulan,ki(19h3) (1977-81) (151:5,-I6-I) Male "" 1.7 _ I ( 1 1=ematle 1.7" 1.4 ?I Gold et al. Baltiuwre, MI) ('asc:runtrul Male and 1.8 1.O NP (1985) ((1978-80) (?UI :?l)1 ) female

TABLE 4.-Continued Risk relative to never smokers Reference Population (yr of data collection) Design (number of subjects) Gender Current smokers Former smokers Yr since quilting Comments Vineis, Esieve, Italy Casc:control Male 8.0° h 3.1" 3-9 Adjusted for number of Terracini (1984) (1978-83) (512:596) 2A° 10-14 cig/day 2.3'' _ 15 VinCi5 et ill. Italy Casexontrol Female 2.3 1.0 23 (1985) (1981-83) (55:202) Jensen et al. Copenhagen, Denmark Case:control Male and 3.4 2.0 NP (19t{7) (1979-81) (388:787) female 13rownson, Chang, Missouri Casexontrol Male 1.9 1.2 NP Adjusted for alcohol Davis (1987) (1 98-1-86) (823:2,-169) l lartge et al. United States Cuse:control Male and 2.9 2.2 1-9 (1987) (1977-78) (2 ,)82:5,7ti2) female 1.6 1()-I 9 1.7 20-29 1.4" ;?!30 Iscovich et al. Argentina Case:control Male and 7.2 4.5 2-4 (1987) (1983-85) (117:234) female 1.8 5-9 1.6 10-19 1.1 ?20 Augustine et al. 9 US cities Case:control Male 2 4` 56 (1988) (1969-84 ) (1,316:3 ,)40) 2.2` 7-12 2.1` >_13 (505:1,5(4) Female 0.ty' 1.7` <_6 1.2` 7-12 I.2` 513

respectively. For those aged 65 or older, the corresponding estimated relative risks were 0.73, 0.54, and 0.29, respectively. These two studies suggest that the risk of lung cancer may decline less steeply with increasing abstinence for older ex-smokers. Multistage Modeling Multistage models provide a conceptual framework for facilitating understanding of the relationship of lung cancer incidence with amount smoked, duration of smoking, and time since cessation. These models, proposing theoretical constructs of fundamen- tal biologic mechanisms, have been useful for evaluating epidemiologic data in a. biologic framework and thereby furthering the understanding of tobacco carcino- genesis. However, fitting these models to epidemiologic data cannot establish the veracity of the underlying biologic theory. Multistage modeling approaches have been used to describe respiratory carcinogenesis and to assess smoking cessation and lung cancer risk. Although a number of different mathematic models of carcinogenesis have been proposed (e.g., two-stage, multicell, multistage), this• discussion primarily ad- dresses the Armitage and Doll (1954, 1957) multistage model, which has been used most extensively in studies of lung cancer. Based on a series of studies examining age-specific mortality rates for various cancers. Armitage and Doll (1954, 1957) proposed a multistage theory of carcino- genesis. Their model assumes that a single cell can generate a malignant tumor only after undergoing a certain number of genetic changes. Animal studies also support the multistage model. Multistage theories also predict the age pattern of occurrence of many tumors induced in experimental animals by continuous exposure to chemical carcinogens. Experimental regimens involving initiation and promotion provide direct evidence of the effect of early- and late-stage events in the carcinogenic process (Stenback, Peto. Shubik 1981 a.b,c). Using data from the British Physicians Study, Doll (1971) showed that when the incidence of lung cancer in cigarette smokers was plotted against duration of smoking, incidence increased approximately in proportion to the fourth power of duration, similar to the slope of the regression line when incidence in never smokers is plotted against age (Figure 3). Thus, a first-stage effect was implicated because the excess lung cancer risk among smokers increased with the same power of duration of smoking as the risk with age among never smokers. Moreover, the lung cancer mortality rates among ex-smokers decreased somewhat initially and then increased slowly in keeping with the increase in risk among never smokers with aQe (Doll 1971). Armitage (1971) noted that the stabilization of excess lunQ cancer risk at the level when smoking stopped suggested that smoking also affected a late stage, namely, the penultimate stage in the carcinogenic process. Day and Brown (1980) conducted a detailed analysis of the pattern of change in cancer risk after cessation of an exposure. The results supported the Armitage-Doll model. In addition, Day and Brown proposed that the stage affected by the agent and the relative magnitude of the effect of the agent on early and late stages of the carcinogenic process are critical in the determination of risk subsequent to cessation of an exposure. To quantify the magnitude of smoking effects on the two stages, Brown 126 rFgMN 438530

TA13LE 2.-Continned Risk relative to never smokers Rel•crence Pupulatiun(yrof dL•ua cullecttun) Design (nuntlxr of subjects) Gender Current smukers Former smokers Yr since quitting Comments La Vecchia. Liati Nurthern Italy C ase:cunlrul Male und 4.3 3.4 <5 Adjusted for SES, diet, et al. ( 1986) (19K-1--85) (129:426) female 2.5 ?5 and alcohol La Vecchia and Northern Italy Case:cuntrul Male and 1.6° 1.1 NP Analysis limited to only Negri (1999) (19t{a-t{8) (3ll:lt{9) female nondrinker.r• NOTL: NP=nul provided: SES=socioccunumic a:uus. "Cumputed as a wcighted average frum cig:u'ctte du.c-Specific rcl:uive risks prescnted in Ihe paper. Weights are the numher ul' cuuuols within each +Irutum of unoking.

risk of pancreatic cancer than current smokers. This diminution of risk with abstinence serves to strengthen the hypothesis that smoking is a contributing cause of pancreatic cancer. Although alcohol does not appear to be a confounder in the assessment of the benefits of smoking cessation, the possibility of confounding by other factors, such as diet or amount of prior cigarette consumption, has not been adequately studied. Bladder Cancer As with pancreatic cancer, the relationship between bladder cancer risk and smoking has been noted for many years. However, because relative risks have not been greatly elevated and because of uncertainty about the effects of unidentified confounding factors in this disease, the causality of this association has been considered less certain compared with other diseases in earlier reports of the Surgeon General (US DHHS 1982). Smoking has nonetheless been regarded as a contributing factor in bladder cancer; in 1985, it was estimated that in the United States 47 percent of bladder cancer deaths in males and 37 percent in females are attributable to smoking (US DHHS 1989). A particular problem with causal inference in smoking and bladder cancer arises because of the inconsistent finding of clear exposure-response relationships in all studies, as has been observed between cigarette smoking and respiratory cancers. However, the usual measures of exposure to tobacco smoke may not accurately index the bladder's dose of tobacco-related carcinogens. The International Agency for Research on Cancer (IARC) concluded, based on evidence available through 1985, that smoking of different forms of tobacco is causally related to cancers of the bladder and renal pelvis (IARC 1986). In addition to the studies reviewed in the 1982 Surgeon General's Report (US DHHS 1982) and in the 1986 report of IARC (1986), more recent data document a consistent association between cigarette smoking and bladder cancer. In an extended followup of a cohort of 25,000 Swedish males, mortality rates for bladder cancer were increased fourfold amonQ ever smokers compared with never smokers (Carstensen, Pershagen, Eklund 1987).+ In current smokers, the risk of death from bladder cancer was approximately three times greater at all levels of consumption. The excess mortality from bladder cancer among current smokers was comparable in the American Cancer Society (ACS) Cancer Prevention Study II (CPS-II) (Table 4). An extension of a large hospital-based case-control study, originally reported in 1977 (Wynder and Goldsmith 1977), showed similar increases in risk among male and female smokers (Augustine et al. 1988). The study included 1,316 male and 505 female cases and 3,940 male and 1,504 female controls interviewed in 9 U.S. cities between 1969 and 1984. For current smokers, odds ratios increased to approximately 3.5 for male and female smokers of 21 to 30 cigarettes per day. Odds ratios were lower among former smokers, although the risk did not decline as the duration of abstinence lengthened (Table 4). The findings of a recent population-based case-control study documented similar levels of bladder cancer risk associated with cigarette smoking (Slattery et al. 1988). Slattery and coworkers (1988) assessed cigarette smoking and bladder cancer in 332 white male cases and 686 controls in Utah. The overall crude odds ratio for current TIMN 438559 159

TABLE 7.-Studies of breast cancer and smoking cessation Risk relative to never smokers Reference Location (yr of data collection) Design (number of subjects) Menopausal status Current smokers Former smokers Yr since quitting Comments Cederlof et al. Sweden Prospective Pre 0.6 0.4 NR Cancer incidence (1975) (1963-72) (27,700) and post Schechter, Miller, Howe (1985) Canada (1980-82) Case:control (49:134) Pre 4.6 1.8 21 Adjusted for several breast cancer (71:219) Post 1.1 0.8 2.1 risk factors tliatt and Fireman Northern California Prospectivt! Pre 1.2 1.2 NR Cancer incidence (1986) (1964-80) (84,172) Post 1.1 1.3 NR Brinton, Schairer, Stanford et al. United States (1973-75) Case:control (447:503) Pre 1.1 1.4 NR (1986) (614:818) Post 1.1 1.0 NR Stockwell and Florida Case:control Pre 1.3" 0.9 NR Lyman (1987) (1981) (4,011:2,952) Post 1.2" 0.9 NR Brownson et al. (1988) Missouri (1979-86) Case:control (114:208) Pre 2.3 1.2 NR (206:872) Post 1.2 0.7 NR ~ Adami et al. Sweden and Norway Case:control Pre 1.0 0.8 Relative risk calculated from Z (1988) (1984-85) (422:527) and post crude data .~ Rohan and Baron Australia Case:control ~ (1989) (1982-84) (146:132) Pre 1.3 2.4 ?1 Adjusted for several breast cancer (280:288) Post 1.5 0.9 _1 risk factors ~ d ®

TABLE 8.-Continued Risk relative to never smokers Reference Population (yr.of data collection) Design (number of subjects) Cancer site Current smokers Former smokers Yr since quitting Comments Hellberg et al. Sweden Case:control Penis 1.6 1.7 NP (1987) (NP) (244:232) Cederlof et al. Sweden Prospective Liver 2.4 1.0 NP Cancer incidence in males (1975) (1963-72) (27,300) Rogot and Murray US veterans Prospective Liver 2.3 1.8 NP Cancer mortality (19811) (1954-69) (248,(H)(1) Yu et al. Los Angeles, CA Case:control Liver 1.8" 1.1 NP Abstainers for Z10 yr were (1983) (1975-79) (76:76) considered never smokers Kahn United States Prospective Stomach 1.4 1.1 NP • Excludes "doctor's orders" quitters (1966) (1954-62) (248,195) Cancer mortality Cederlof et al. Sweden Prospective Stomach 1.3 0.7 NP Cancer incidence in males (1975) (1963-72) (27,300) Rogot and Murray US veterans Prospective Stomach 1.5 1.1 NP Extension of US Veterans Study (1980) (1954-69) (293,958) Nomura et al. Japanese men in I lawaii Prospective Stomach 2.7 1.0 NP Cohort identified 1965-68 and (199O) (1965-68) (7,990) followed through October 1986 Kahn US veterans Prospective Leukemia 1.4 1.5 NP Excludes "doctor's orders" quitters (1966) (1954-62) (248,195) Cancer mortality

TABLE 1.-Continued Risk relative to never smokers Reference Population (yr of datacollection) Design (numtx:rof subjects) Gentler Current smokers Former smokers Yr since quitting Comments Wigle, Mao, Grace Alberta, Canada Case:control (1980) (1971-73) (84:1,(X)2) Mate 8.7 3.5 NP (41:674) Female 4.3 0.8 NP Spitz et al. I louston, TX Case:control Male 4.5" 6.1 <5 (1988) (1985-87) (121:127) 2.2 5-14 1.0 _15 (50:49) Femate 5.5" 9.8 <5 4.5 5-14 1.5 ?I5 Blot et al. 4 areas in United States Case:control Male 3.4 1.1 1-9 Adjusted for alcohol (1988) (1984-85) (762:837) 1.1 IO-19 consumption 0.7 z20 (352:431) Female 4.7 1.8 1-9 0.8 10-19 0.4 ~20 Franco et al. Brazil Case:control Male 9.3 2.9 <10 Data for commercially (1989) (1986-88) (232:464) and I'emate 0.6 ?I() produced cigarettes only

TABLE 8.-Continued Risk relative to never smokers Reference Population (yr of duta collection) Design (number of subjects) Cancer site Current smokers Former smokers Yr since quitting Comments Cederlof et al. (1975) Sweden (1963-72) Prospective (27,300) Leukemia (Males) 1.1 0.8 NP Cancer incidence (27,700) (Females) 0.4 1.0 NP Rogot and Murray US veterans Prospective Leukemia 1.6 1.5 NP Extension of US Veterans Study (198(1) (1954-69) (248,(x)l1) Trichopoulos et al. (1987) Greece (1976-84) Case:control (104:454) Liver fi[3,Ag 3.3'' 2.8 NP (89:454) HQ,Ag+ 1.62 1.3 NP ACS CPS-11 (unpublished United States (1982-86) Prospective (421,623) Kidney (Males) 2.7 2.1 NP Cancermortality tabulations) (605,758) (Females) 1.5 1.1 NP NOTE: NP=not provided; It13,Ag=hepatitis l3 surface antigen; ACS CPS-I1=American Cancer Society Cancer Prevention Study 11. 'ComputeJ as a weighted average from cigarette dose-specific relative risks presented in the palxr. Weights are the numtHrrof comrolx within each xtratwn of smoking.

CARSTENSEN. J.M.. PERSHAGEN, G., EKLUND, G. Mortality in relation to cigarette and pipe smoking: 16 years' observation of 25,000 Swedish men. Journal of Epidemiology and Community Health 41:166-172, 1987. CEDERLOF, R., FRIBERG, L., HRUBEC, Z.. LORICH, U. The Relationship of Smoking and Some Social Covariables to Mortality and Cancer Morbidity. A Ten Year Follow- Up in a Probability Sample of55,000 Swedish Stthjec•tsAge 18-69, Part 1/2. Stockholm, Sweden: The Karolinska Institute. Department of Environmental Hygiene, 1975. CORREA, P.. PICKLE, L.W., FONTHAM. E., DALAGER. N., LIN, Y., HAENSZEL, W., JOHNSON, W.D. The causes of lung cancer in Louisiana. In: Mizell. Ivl., Correa, P. (eds.) Lung Cancer: Causes and Prevention. Proceedings of the International Lung Cancer Update Conference. New Orleans: Verlag Chemie International, Inc.. 1984, p. 73. DAMBER, L.A., LARSSON, L.G. Smoking and lung cancer with special regard to type of smoking and type of cancer. A case-control study in north Sweden. British Journal of Canc•er 53(5):673-681, May 1986. DAY, N.E. Epidemiological data and multistage carcinogenesis. In: Borzsonyi. M.. Lapis, K., Day, N.E., Yamasaki, H. (eds.) Models. Mechanisms and Etiology of Tumor Promotion. Lyon: IARC, 1984, pp. 339-357. DAY, N.E., BROWN, C.C. Multistage models and primary prevention of cancer. Journal of the National Cancer Institute 64(4):977-989. April 1980. DEVESA, S.S.. BLOT, W.J.. FRAUMENI. J.F. JR. Declining lung cancer rates among young men and women in the United States: A cohort analysis. Journal of the Nutiona.l Cancer Institute 81:1568-1571, 1989. DOLL, R. The age distribution of cancer: Implications for models of carcinogenesis. Journal of the Royal Statistical Society A 134: l 33-166, 1971. DOLL, R.. GRAY, R., HAFNER, B.. PETO. R. Mortality in relation to smoking: 22 years' observations on female British doctors. British Medical Journa1280(6219):967-971, April 5, 1980. DOLL. R.. HILL, A.B. Mortality in relation to smoking: Ten years' observations of British doctors. British Medical Journal l:1399-1410. May 30, 1964.. DOLL, R.. PETO. R. Mortality in relation to smoking: 20 years' observations on male British doctors. Brr'tish Medical Journal 2:1525-1536, December 25. 1976. DOLL, R.. PETO. R. Cigarette smoking and bronchial carcinoma: Dose and time relationships among re,ular smokers and Iifelong non-smokers. Journal of Epidemiology and Community Health 32(4):303-313. December 1978. ELWOOD, J.M., PEARSON, J.C.G.. SKIPPEN. D.H., JACKSON, S.M. Alcohol, smoking, social and occupational factors in the aetiology of cancerof the oral cavity, pharynx and larynx. International Journal of Cancer 34:603-612. 1984. FALK, R.T., PICKLE, L.W., BROWN, L.M.. :vtASON, T.J., BUFFLER. P.A., FRAUMENI. J.F. JR. Effect of smoking and alcohol consumption on laryngeal cancer risk in coastal Texas. Cancer Research 49(14):4024=1029, July 15. 1989. FARBER, E. The multistep nature of cancer development. Cancer Research 44:4217-4223, October 1984. FLANDERS, W.D., ROTHMAN, K.J. Interaction of alcohol and tobacco in laryngeal cancer. American Journal of Epidemiology 115(3):371-379, March 1982. FREEDMAN, D:A.. NAVIDI, W.C. Multistage models for carcinogenesis. Environmental Health Perspectives 81:169-188, May 1989. GAFFNEY, M., ALTSHULER, B. Examination of the role of cigarette smoke in luna carcinogenesis using multistage models. Journal of the National Cancer Institute 80(12):925-931, August l7, 1988. GAO, Y.T., BLOT, W.J., ZHENG, W.. FRAUMENI. J.F., HSU, C.W. Lung cancer and smoking in Shanghai. International Journal of EpidemioloQy 17(2)277-280, June 1988. 138 TIMN 438541

INTRODUCTION Lung cancer, the first neoplasm causally linked to cigarette smoking, has been the cancer most thoroughly studied with respect to exposure-response relationships and benefits of cessation (US DHHS 1982). Subsequently, cigarette smoking has been established as a cause of cancer at diverse other sites. For some sites (e.g., oral cavity), the target cells are exposed directly to the various constituents of tobacco smoke. For other sites (e.g., urinary bladder), absorption, transport, and metabolic activation of carcinogens in tobacco smoke result in exposure of target tissues. This Chapter reviews the evidence on smoking cessation and cancer risk at various nonrespiratory sites. The sites selected for review are those for which cigarette smoking has been determined to be a cause of cancer, or contributing cause, or those for which evidence indicates a possible association. Methodologic issues encountered in inferring causality on the effects of smoking cessation have been discussed in Chapter 2 and will not be reviewed in detail in this Chapter. Potential confounding by differences in prior tobacco exposure at the time of quitting, and by differences between former smokers and continuing smokers in other cancer-related risk factors may pose a greater obstacle to causal inference for the nonrespiratory cancers than for cancers of the lung or larynx: the smoking effects are generally smaller for nonrespiratory cancers, and the potential confounding factors are more numerous. REVIEW OF SPECIFIC SITES Oral Cancer Tobacco use is a major cause of oral cancer (US PHS 1964; US DHHS 1982, 1989). An exposure-response relationship has been identified between the amount of tobacco consumed and the risk of cancer of the oral cavity after considering the effects of alcohol consumption. The proportion of 1985 oral cancer deaths attributable to cigarette smoking in the United States has been estimated to be 92 percent for men and 61 percent for women (US DHHS 1989). The oral cavity, like the lung, receives direct exposure to cigarette smoke. Presumably, the causal association of cigarette smoking with cancer of the oral cavity reflects this contact and the same initiating and promoting agents that are considered to determine the development of lung cancer. Table I summarizes studies that have examined the relationship between smoking cessation and oral cancer risk. In these studies, the risk of oral cancer among current smokers ranges from 2.0 to 18.1 times (median of approximately 4) the risk among never smokers. Oral cancer risks for women who are currently smoking seem lower than those for men in studies conducted prior to the mid- 1970s, but little difference by gender has been noted in more recent research. This gender pattern may be because of the initiation of smoking at an older age among earlier birth cohorts of women (US DHHS 1989) born during this century and the resultant low cumulative lifetime exposure of such women. TIMN 438547 147

Table 6 summarizes findings from studies that have examined the relationship between cervical cancer risk and cigarette smoking cessation. In these studies, the risk among current smokers ranges from 1.0 to 5.0 times the risk among never smokers (median of approximately 2). Smoking-associated risks for invasive cancer and for carcinoma in situ are generally similar. After the first year of abstinence, former smokers have lower cervical cancer risk than current smokers in most studies. Exceptions include the study conducted in Milan (La Vecchia, Franceschi et al. 1986), which showed risk reduction for invasive cancer but not for carcinoma in situ among former smokers, and the study conducted in Central America (Herrero et al. 1989) in which no association with smoking was observed at all, even for current smokers. The effect of time since stopping has not yet been well studied for cervical cancer, but observations from a large multicenter study conducted by NCI (Brinton. Schairer, Haenszel et al. 1986) suggested that risk reduction may occur fairly rapidly after cessation. One study found that smokers tended to have a poorer prognosis for survival after radiation treatment for invasive cervical cancer, but no data • were presented regarding smoking cessation (Kucera et al. 1987). A major concern in studies of smoking and cervical cancer has been the potential for confounding by factors that would predispose a woman to become infected with a sexually transmitted agent that might be causally related to the disease, such as human papilloma virus (Stellman. Austin, Wynder 1980; Winkelstein et al. 1984; IARC 1986). Therefore, it is important to note that those studies that controlled for risk factors for sexually transmitted disease (Trevathan et ai. 1983; Greenberg et al. 1985; Herrero et al. 1989; Slattery et al. 1989) produced relative risk estimates for current and former smokers that were quite similar to those from studies that made no such adjustments. The association of smoking and cervical cancer has been considered by some to be a result of residual confounding by inadequately measured indicators of exposure to a sexually transmitted aQent. Although factors such as the number of past sexual partners are only surrogates for a hypothetical etiologic infectious agent, they are the very same social correlates of tobacco smokinQ that would suggest this type of confounding. Therefore, even though such factors as age at first intercourse and the number of sexual partners are imperfect indicators of infection by a possible etiologic agent, their inclusion as covariates in multivariate analyses may be sufficient to control confound- ing to some extent in the analysis of the effects of smoking on cervical cancer risk. This review of the evidence on cervical cancer and ciQarette smoking cessation indicates that there is a consistently observed association between cervical cancer risk and cigarette smoking and that former smokers experience a lower risk of cervical cancer than current smokers, even after adjusting for the social correlates of smoking and risk of sexually acquired infections. This observed diminution of risk after cessation lends support to the hypothesis that smoking is a contributing cause of cervical cancer. Based on a recent comprehensive review of epidemioloaic studies providing data on smoking and cervical cancer. Winkelstein (1990) concluded that smoking is causally associated with cervical cancer. 166 TIMN 438566

TABLE 8.--Studies of cancer at selected sites that have examined the effect of smoking cessation Risk relative to never smokers Reference Popul.uion (yr of data collection) Design (number of subjects) Cancer site Current smokers Former smokers Yr since quitling Comments Cederlof et al. Sweden Prospective Endometrium 0.5 1.6 NP Cancer incidence (1975) (1963-72) (27,700) Lesko et al. 8 North American cities Case:controt Endometrium 0.8" 0.9 ? 1 Adjusted for obesity and (1985) (1976-83) (508:706) exogenous estrogens Stockwell and Florida Case:control Endometrium 0.8° 0.6 NP Lyman(1987) (1981) (990:2,952) Cederlof et al. Sweden Prospective Ovary 0.5 1.6 NP Cancer incidence (1975) (1963-72) (27,7O0) Stockwell and Florida Case:control Ovary 0.9 NP Lyman (1987) (1981) (696:2')52) Franks et at. United States Case:control Ovary I.I 0.9 21 Adjusted for age, parity, and use of ~ (1987) (1980-82) oral contraceptives ~ Kahn US veterans Prospect ive Kidney 1.4 1.5 NP Excludes "doctor's orders" quitters (1966) (1954-62) (248,195) Cancer mortality ~ Rogot and Murray US veterans I'rospective Kidney 1.4 1.2 NP Extension of US Veterans Study W (1980) (1954--69) (293 ,I58) ®0 (h Jensen et al. Copenhagen Case}control Renal pelvis 3.7 1.9 NP Crude relative risks computed -4 (1988) (1979-82) (96:288) and ureter from data presented W J w

cervix. This lower risk appears to be neither an artifact of a lower exposure to cigarettes in former smokers prior to quitting nor a result of confounding by other known risk factors for these cancers. This observation of a diminution in risk further supports the hypothesis that cigarette smoking is a causal factor for cancers of many sites other than the respiratory system. Although smoking is not as strong a risk factor for non- respiratory cancers as it is for cancers of the lung and larynx, substantial numbers of cases of many nonrespiratory cancers can be attributed to tobacco use (US DHHS 1989). The patterns of diminution in risk with increasing duration of abstinence indicate that smoking cessation provides a substantial reduction in the risk of nonrespiratory cancer. CONCLUSIONS 1. Smoking cessation halves the risks for cancers of the oral cavity and esophagus, compared with continued smoking, as soon as 5 years after cessation, with further reduction over a longer period of abstinence. 2. Smoking cessation reduces the risk of pancreatic cancer, compared with continued smoking, although this reduction in risk may only be measurable after 10 years of abstinence. 3. Smoking is a cause of bladder cancer; cessation reduces risk by about 50 percent after only a few years, in comparison with continued smoking. 4. The risk of cervical cancer is substantially lower among former smokers in com- parison with continuing smokers, even in the first few years after cessation. This finding supports the hypothesis that cigarette smoking is a contributing cause of cervical cancer. 5. Neither smoking nor smoking cessation are associated with the risk of cancer of the breast. 178 TIMN 438578

stood, but may be due to smoking effects on estrogen production and metabolism, including increased 2-hydroxylation of estradiol in smokers (Michnovicz et al. 1986), an earlier age at menopause in smokers (Baron 1984), and indirect effects of the body weight differences between smokers and nonsmokers, such as the production of estrogens from precursors within adipose tissue (MacDonald et aI. 1978; Chapters 8 and 10). Table 8 includes a summary of findings from studies of endometrial cancer that have examined cigarette smoking cessation. Although the risk of endometrial cancer among current smokers in these studies is approximately 30 percent lower than that among never smokers, the risk among ex-smokers is similar to, or slightly greater than, that among current smokers. This review of past research on endometrial cancer risk and cigarette smoking cessation suggests that current smokers are at lower risk of endometrial cancer than never smokers, but it is not clear whether this protective effect of smoking on endo- metrial cancer risk might be reversed soon after cessation of cigarette smokina. AlthouQh further investigation of the mechanisms for the protective effect of smoking on endometrial cancer is of scientific interest to better understand the effects of smokina on hormones and of hormones on endometrial cancer risk, this inverse association with smoking has no public health relevance, as the well-substantiated risks to other organ systems from continued smoking far outweigh any potential benefits to the endo- metrium. Other Cancer Sites The metabolic products of tobacco smoke can be found in-ovarian follicular fluid (Hellberg and Nilsson 1988). However, there is little evidence that smoking is as- sociated with cancer of the ovary (Byers et al. 1983; Baron 1984; Baron et al. 1986; Stockwell and Lyman 1987; Whittemore et al. 1988; Mori et al. 1988). The risk of ovarian cancer differs little for either current or former smokers, as indicated in the only two studies that have examined the effect of cigarette smokinQ cessation on ovarian cancer risk (Table 8). ~ Tobacco has been regarded as a contributing factor for cancer of the kidney (US DHHS 1982, 1989). The U.S. Veterans Study (Kahn 1966; Rogot and Murray 1980) and ACS CPS-II (ACS, unpublished tabulations) suggest only small differences in mortality from renal cancer between current and former smokers (Table 8). A study of renal pelvis and ureteral cancers in Copenhagen (Jensen et al. 1988), however, showed a pattern of risk diminution with abstinence similar to that observed in bladder cancer, a site with the same histologic type of transitional-cell tumors. Cancers of the anus and penis are considered possibly to result from infection by a sexually transmitted agent in a way analogous to cancer of the uterine cervix (Daniell 1985; Daling et al. 1987; Heilberg et al. 1987). Smokers have been found to be at increased risk both for cancer of the penis (Hellberg et al. 1987) and anus (Daling et al. 1987; Holmes et al. 1988) in recent studies. Only one study has examined the effect of cessation on the risk of these cancers (Hellberg et ai. 1987). This study found that 172 TIMN 438572

smoking, compared with never smoking, was 3.69 (95-percent confidence interval (CI), 2.58-5.26). However, an exposure-response relationship was not evident with reported average number of cigarettes. smoked daily. The odds ratios for former smokers declined only after 8 years or more of abstinence. Table 4 summarizes findings from studies that have examined the relationship between cigarette smoking cessation and risk of bladder cancer. Of all the non- respiratory cancer sites, the relationship between bladder cancer risk and cigarette smoking cessation has been most extensively studied. In these studies, the risk among current smokers ranges from 1.0 to 7.2 times the risk among never smokers (median of approximately 3); risks are similar among males and females. More recent studies conducted since the mid-1970s tend to show higher risks for current smokers than do the earlier studies. The higher risks in more recent studies may reflect the earlier age of starting to smoke of more recent cohorts of smokers (US DHHS 1989) or the presence of a long latency period for the smoking effect to become fully manifest after initiation in susceptible persons. Beyond the first few years of abstinence, former smokers generally have lower risks than current smokers. The study conducted in six U.S. cities (Wynder and Stellman 1977: Wynder and Goldsmith 1977) indicated an approximate 50-percent reduction in risk after 6 years of abstinence, with risk returninQ to that of nonsmokers among men after 15 years. A similar return to nonsmoker risk was also observed after 6 years of abstinence in an English study (Cartwright et al. 1983) and in an Argentine study after 20 years (Iscovich et al. 1987). However, results from other studies (Howe et al. 1980; Vineis, Esteve, Terracini 1984; Hartge et al. 1987; Burch et al. 1989) indicated that the reduction in risk in the first few years after cessation is followed by little subsequent additional reduction', even beyond 10 or 15 years of abstinence. These observations are in contrast to those for the other cancer sites reviewed in this Chapter. In some studies. the analyses controlled for the possible confounding effects of lower cigarette consumption among former smokers prior to cessation. The U.S. Veterans Study (Kahn 1966) showed no reduction in risk for former smokers, compared with current smokers, at levels of past cigarette consumption of 1 pack or less per day. There was an approximate 50-percent reduction in risk, however, for those former smokers who had previously smoked more than I pack per day. Most studies that included past ' cigarette smoking exposure as a covariate in multiple logistic regression analyses (Wigle, Mao, Grace 1980; Howe et al. 1980; Vineis, Esteve, Terracini 1984: Claude, Frentzel-Beyme, Kunze 1988; Slattery et al. 1988; Burch et al. 1989) showed relative risks that were similar to those observed in studies in which no such adjustment was made. A large muiticenterstudy conducted by NCI (Hartge et al. 1987) contained sufficient numbers of subjects for detailed subgroup analyses. Table 5 displays the findings of this study when both average cigarette dose per day and duration of smoking are cross-classified for current and former smokers. In each of these nine categories, bladder cancer risk was lower among former smokers than amona current smokers. As reviewed above, the amount of evidence supporting cigarette smoking as a cause of bladder cancer has become increasingly compelling since the 1982 Report of the Surgeon General (US DHHS 1982), which focused on cancer. Multiple studies of 164 TIMN 438564

TABLE l.-Continued Risk relative to never ' smokers Reference Population (yr of data collection) Design (number of suhjects) Gender Current smokers Former smokers Yr since quiuing Comments Kabat and Wynder (1989) 18 US cities (1976-83) Case:cnntrol (511:1.057) Mitle 5.S' 2.1 _I Adjusted for alcohol (226:453) Female 4.1'' 1.5 ?1 Kabat, I Iehert, 7 US cities Case:contrul Female 2.0 1.0 NP Adjusted for alcohol and Wynder (1989) (1981-87) ( I25:107) previous number of ACS CPS-II (unpublished United States (1982-86) Prospective (421,623) Male 18.1 6.4 NP cig/day Cancer mortality tabulations) (605,758) Female 5.8 2.5 NOTE: NP=noi proviJed: ACS CPS-11=American Cancer Society Cancer Prevemion StuJy II. "Computed ax a weighieJ average from cigareue do.e-,pecitic relvive risks presemeJ in the paper. Weights are the number of controls within each strutum of smoking.

Daily Cigarette Consumption Previous smoking intensity or number of cigarettes smoked per day also affects the pattern of risk reduction after smoking cessation. In the U.S. Veterans Study, the mortality ratios for lung cancer were 1.41. 3.47, 8.34, and 10.05 for ex-smokers who smoked I to 9, 10 to 20, 21 to 39, and 40 cigarettes or more per day, respectively (Kahn 1966). The pattern of lung cancer risk reduction by years of smoking abstinence and number of cigarettes smoked has been reported for several studies. In ACS CPS-I and ACS CPS-II (Hammond 1966; Garfinkel and Stellman 1988), the decline in risk with stopping smoking showed a comparable proportional reduction in risk among those who had smoked less (Table 3). In the European case-control study (Lubin et al. 1984a), men who had stopped smoking for 10 years or more, but had previously smoked 30 cigarettes or more per day, had a 40-percent risk of developing lung cancer compared with corresponding current smokers, whereas men who had smoked 1 to 9 cigarettes per day had a 67-percent risk compared with corresponding current smokers. Similar results were observed for female ex-smokers (Lubin et al. 1984a). As pre- viously discussed, duration of smoking was considered in these analyses. Thus, heavier smokers have less reduction of lung cancer risk following cessation than smokers of fewer cigarettes per day. Inhalation Practices The pattern of lung cancer risk by years of smoking abstinence-and by inhalation practices (i.e., frequency and depth of inhalation) was examined by Lubin and col- leagues (1984a). Their analysis indicated a somewhat greater reduction in risk for those ex-smokers who had inhaled less often or less deeply. Among men who had stopped smoking for 10 years or more, relative risk by reported frequency of inhalation compared with current smokers was lowest for those who had rarely or never inhaled (relative risk (RR)=0.39) and for those whose depth of inhalation was reported as only slijht or not at all (RR=0.37). In comparison, the relative risk after 10 years or more of abstinence was highest for those who had inhaled all the time (RR=0.50) and for those who had inhaled deeply (RR=0.47). The same pattern was observed among women. Different Tobacco Products Differences in the reduction in risk following cessation also have been investigated by types of cigarettes smoked. A lower risk of lung cancer has been observed for smokers of filter cigarettes compared with smokers of nonfilter cigarettes (US DHHS 1982, 1989; Wynder and Kabat 1988), a pattern suggesting that the reduction in risk among former smokers may be more apparent for filter cigarette smokers. However, no significant differences in the trend of risk reduction by years of smoking abstinence (0, 1-4, 5-9, and ->10) and by type of cigarettes smoked (filter, mixed, nonfilter) were observed by Lubin and coworkers (1984b) in the European case-control study. Among 124 TII~~N 438528

TABLE 7.-Continued Risk relative to never smokers Yr Location (yr of Dcsign Menopausal Current Fonner since Reference dutacollection) (nwnherofsubjects) status smokers smokers quitting Comments London et al. United Stutes Prospective Pre 1.0" 1.1 NR (1989) (1976-80) (117,557) Post I.1° 1.1 NR NOTE: NR=not reported. aComputed as a weighted average from cigarette dose-specific relative risks presented in the paper. Weights are the number of controls within each stratum of smoking. J

TABLE 2.-Studies of esophageal cancer that have examined the effect of smoking cessation Risk relative to never smoker. Reference Population (yr of data collection) t)esien (number of subjects) Gender Current smoker, Former smokers Yr since qtiitting Comments Kahn US veterans Prospective Male 5.3 1.6 NP Excludes "dtx:tor's orders" (1966) (1954-62) (248.195) . quitters Cancer mortality Cederlof et al. Sweden Prospective Male 1.7 1.7 NP Cancer incidence (1975) (1963-72) (27,300) Wynder and 6 US cities Case:contrul Male 3.6 4.}{ 1-3 Stellman (1969-75) (159;6,5i-1) 1.5 4-6 (1977) 1.4 7-t0 1.3 I I-15 1.() ?16 (76:6.522) Fetrrtle 5.3 3.0 1-3 3.1 4-6 l) 7-10 2.2 1 I-15 1.8 -16 Wigle, Mao, Grace Allxxta, Canada Casc:contrul Male 5.1 1.1 NP (198O) (1971-73) (45:1,(N)3) Rogot and Murray US veterans Prospective Mule 6.4 2.4 NP Excludes "doctor's orders" (1980) ((1954-69) ('_93')5t{) yuitters Cancer mortality Extension of US Veterans Study

the initial malignancy, or simply the consequence of chance (Schottenfeld 1982). Thus, multiple primary cancers have been investigated with the goals of examining environ- mental and host factors increasing cancer risk and of identifying adverse consequences of cancer treatment. Tobacco use, including cigarette smoking, has been examined as a risk factor for the development of a second primary cancer, after diagnosis of a first malignancy at cigarette-associated and non-cigarette-associated sites; the effect of smoking cessation on the occurrence of second cancers has also been addressed in several investigations. Descriptive studies have shown that an initial malignancy at a smoking-associated site is followed by an increased risk for cancer at the same or another cigarette- associated site (Wynder et al. 1969; Schottenfeld 1982). In an early study of multiple primary cancers. Berg, Schottenfeld, and Ritter (1970) examined the risks of second primary cancers in persons evaluated at Memorial Hospital for squamous cell cancers of the respiratory or upper digestive tract or other histologic types of lung cancer. In comparison with expected numbers of cases based on incidence rates for New York State, significant excesses were observed for cancers of the lip, oral cavity or pharynx, esophagus, larynx, and lung. Only limited evidence is available on the effects of smoking cessation on the occurrence of multiple primary cancers. Moore reported two studies (1965, 1971) of second primary cancers in persons with an index malignancy of the mouth, pharynx, or larynx; both showed reduced risk for a second primary cancer in persons who stopped smoking after diagnosis of the first cancer. For I to 15 years, Silverman. Gorsky, and Greenspan (1983) observed 117 smokers who had a primary cancer of the head and neck region. Thirty percent of continuing smokers developed a second oral primary cancer compared with 15 percent of those reducing smoking and 13 percent of those completely stopping. In contrast, an effect of cessation was not found in two other studies (Castigliano 1968; Schottenfeld, Gantt, Wynder 1974). Castigliano's 1968 study included 88 subjects with mouth or throat cancer who survived for at least 3 years without evidence of recurrence. During a minimum followup period of 3 years, the occurrence of a second primary cancer was not related to smoking status. Schottenfeld, Gantt, and Wynder (1974) examined multiple primary cancers in 733 patients admitted to Memorial Sloan-Kettering Cancer Center with a primary epidermoid carcinoma of the oral cavity, pharynx, or larynx. During the 5-year followup period, the smoking status of those developing and not developing a second primary did not differ significantly. Interpretation of these studies is limited by the small numbers of subjects and the limited duration of followup. Furthermore, the interactions of tobacco smoking with other risk factors of cancers of the head and neck, particularly alcohol consumption, complicate interpretation of these data. SUMMARY This review of the relationship between cigarette smoking cessation and the risk of nonrespiratory cancers has shown that former smokers tend to have lower risk than current smokers for cancers of the oral cavity, esophagus, pancreas, bladder, and uterine TIMN 438577 177

current smokers had a penile cancer risk 1.6 times that of never smokers, but the risk among former smokers was similar to that among current smokers (Table 8). Primary hepatocellular cancer has been associated with smoking in a number of recent studies (Trichopoulous et al. 1980; Lam et al. 1982; Yu et al. 1983; Oshima et al. 1984; Trichopoulos et al. 1987; Hirayama 1989). This association is of potentially great public health importance because of the high incidence of primary -liver cancer and the epidemic of cigarette smoking worldwide, which is increasingly involving countries in which liver cancer is the leading cause of cancermortality. The mechanism whereby smoking might affect liver cancer risk is unknown. Although potential confounding by alcohol consumption is of concern in interpreting this association, the association of smoking with hepatocellular cancer has remained significant in several studies after controlling for alcohol intake (Trichopoulos et al. 1980; Yu et al. 1983; Oshima et al. 1984; Trichopoulos et al. 1987). One case-control study (Yu et al. 1983) and two cohort studies (Cederlof et al. 1975; Carstensen, Pershagen, Eklund 1987; Rogot and Murray 1980) have examined the effects of smoking cessation on liver cancer risk. In all three studies, current smokers were found to have higher risks than either never smokers or former smokers. In the case-control study, potential confounding by different alcohol consumption of current and former smokers was controlled (Yu et al. 1983). Many of the earlier studies (including the prospective studies reviewed in this Chapter) did not exclude the possibility that cancer of the liver may have been primary in another (smoking-related) organ. The possible role of hepatitis B as a modifier of the effect of smoking on the risk of liver cancer is not clear (IARC 1986). Tobacco has been associated with stomach cancer, but whether this association is causal remains unclear (IARC 1986: US DHHS 1982, 1989). Few studies have considered the effect of cessation on the' risk of stomach cancer. The U.S. Veterans Study (Kahn 1966: Rogot and Murray 1980) and the Swedish study (Cederlof et al. 1975) indicate a reduction in stomach cancer risk after cessation, although the relative risks among current smokers were small in these studies (Table 8). Leukemia has recently been implicated as a smoking-related disease (Austin and Cole 1986; Severson 1987; Kinlen and Rogot 1988), but this observation has not been consistent (for review, see Kinlen and Rogot 1988). The U.S. Veterans Study showed only a slight dose-response relationship for myelogenous leukemias, but there was little difference in risk between current and former smokers (Kahn 1966: Rogot and Murray 1980; Kinlen and RoQot 1988). In the earlier presentation of these data, there was no difference in risk among ex-smokers, compared with current smokers, at any of four levels of prior cigarette smoking (Kahn 1966). The most recent analysis of these data indicated there was little difference in risk among former smokers compared with current smokers for any of the subtypes of leukemia. One study demonstrated a poorer prognosis for patients with myelogenous leukemia who were cigarette smokers (Ar- chimbaud et al. 1989). MULTIPLE PRIMARY CANCERS The occurrence of multiple primary cancers may reflect the effects of the same risk factors in the pathogenesis of the multiple cancers, the.effects of agents used in treating 176 TIMN 438576

WILLETT, W.C., BROWNE, M.L., BAIN, C., LIPNICK, R.J., STAMPFER, M.J., ROSNER, B., COLDITZ, G.A.. HENNEKENS, C.H., SPEIZER. F.E. Relative weight and risk of breast cancer among premenopausal women. American Journal of Epidemiology 122:731-740, 1985. WILLIAMS, R.R., HORM, J.W. Association of cancer sites with tobacco and alcohol consump- tion and socioeconomic status of patients: Interview study from the Third National Cancer Survey. Journal of the National Cancer Institute 58(3):525-547, March 1977. WINKELSTEIN, W. JR. Smoking and cervical cancer-Current status: A review. American Journal of Epidemiology 131(6): 945-957, June 1990. WINKELSTEIN, W. JR., SHILLITOE, E.J., BRAND, R., JOHNSON, K.K. Further comments on cancer of the uterine cervix, smoking, and herpesvirus infection. American Journal of Epidemiology 119(1):1-8, January 1984. WINN, D.M., BLOT, W.J., SHY, C.M., PICKLE, L.W., TOLEDO, A., FRAUMENI. J.F. Snuff dipping and oral cancer among women in the southern United States. New England Journal of Medicine 304(13):745-749, March 26, 1981. WYNDER, E.L., DODO, H., BLOCH, D.A., GANTT, R.C., MOORE, O.S. Epidemiologic investigation of multiple primary cancer of the upper alimentary and respiratory tracts. I. A retrospective study. Cancer 24:730-739, 1969. WYNDER, E.L.. GOLDSMITH, R. The epidemiology of bladder cancer. A second look. Cancer 40(3):1246-1268, September 1977. WYNDER, E.L.. HALL, N.E.L, POLANSKY, M. Epidemiology of coffee and pancreatic ~ cancer. Cancer Research 43(8):3900-3906, August 1983. WYNDER. E.L., STELLMAN, S.D. Comparative epidemiology of tobacco related cancers. Cancer Research 37(12):4608-4622, December 1977. YU. M.C.. MACK. T.. HANISCH. R., PETERS, R.L., HENDERSON, B.E., PIKE, M.C. Hepatitis, alcohol consumption, cigarette smoking, and hepatocellular carcinoma in Los Angeles. Cancer Research 43( l•2, Part 1):6077-6079, December 1983. 186 TIMN 438586

In each study summarized in Table 1, the risk of oral cancer was lower among former smokers after the first few years of abstinence than for current smokers. After 3 to 5 years of smoking abstinence, oral cancer risk decreased by 50 percent. In a study in Argentina (Iscovich et al. 1987) and in the large multicenter study conducted by the U.S. National Cancer Institute (NCI) (Blot et al. 1988), the risk of oral cancer among former smokers after 10 years of abstinence was comparable with that among never smokers. This observation has been interpreted as an indication that the greatest effect of smoking on oral cancer risk may be in the later (postinitiation) stages of carcino- genesis (Blot et al. 1988). Although it is well known that smokeless tobacco (ST) increases the risk of oral cancer (Winn et al. 198 1; US DHHS 1986) and that stopping the use of ST reduces the prevalence of premalignant tissue changes in the mouth (Gupta et al. 1986), there is little information on the risk of oral cancer in former users of ST. Compared with current smokers, former smokersmay have different alcohol drinking habits before and after smoking cessation, and thus comparisons of risk between current and former smokers may be confounded by alcohol consumption (Chapter 11). In three investigations, the effect of smoking cessation was examined and past alcohol con- sumption was controlled by multiple logistic regression (Blot et al. 1988: Kabat and Wynder 1989: Kabat. Hebert, Wynder 1989). In the three studies, estimates of relative risks for both current and former smokers were similar to those observed in studies in which alcohol was not included as an adjustment factor. The stability of the relative risk estimates for smoking with adjustment for alcohol intake suggests that alcohol does not substantially confound the relationship between oral cancer risk and cigarette smoking status and that the lower risk of former smokers cannot be explained by lower levels of alcohol consumption (Chapter 11). One study was sufficiently large to permit detailed stratified analysis of the modification of the smoking effect by alcohol consumption (Blot et al. 1988). In this study, former smokers were observed to have a lower risk than current smokers for both men and women at each of five levels of alcohol consumption. The U.S. Veterans Study (Kahn 1966) demonstrated that at each of three levels of past cigarette smoking exposure, former smokers had lower risk of oral cancer than did current smokers. Kabat, Hebert. and Wynder (1989) controlled for past cigarette exposure by multiple logistic regression and found that relative risk estimates, which were adjusted for past alcohol and ciQarette consumption, did not differ from the crude estimates for former smokers (1.0 vs. 1.0 relative to never smokers). Second primary cancers of the mouth and pharynx occur commonly in persons with an initial primary cancer in the mouth, pharynx, or larynx. Several studies have addressed the incidence of second primaries of the mouth. pharynx, or larynx in relation to smokinQ status after diagnosis and treatment of the first primary. The findings of these studies are inconclusive, with some indicating reduced risk of a second primary after cessation (Moore 1965; Moore 1971: Wynder et al. 1969; Silverman, Gorsky, Greenspan 1983) and others showing no clear benefit of cessation (Castigliano 1968; Schottenfeld, Gantt, Wynder 1974; Chapter 5, see section on Multiple Primary Cancers). TIMN 438551 151

TABLE 1.-Continued Number of Relative risk as comp ared with never smokers Reference Population Numher of cases Number of controls Source of controls cases among former smokers Former smokers Current smokers Rosenberg. Palmer, Eastern US women aged <65 910 2,375 tlospital-based 149 Overall Shapiro (1990) 1.2 (1.(3-1.7) 3.6 (3.0-4.4) Quit <24 mo 2.6 (1.8-3.8) Quit 24-35 mo 1.3 Quit ?36 mo 0.8-1.1 NOTE: CI ID=coronry heart disease. '95%conlidence imcrval shown in parentheses when available.

TABLE 3.-Continued Risk relative to never snwkct:s ReferencC Population (yr of d:na collection) Desien (number o('.ubjects) Gender Current snwl:ers Former smokers Yr since yuitting Comments Cuzick and Babiker (1989) Gtglund ((1983-86) Case:control (123:150) Male 2.1'" 3.6 <1O 3.6 10-20 1.3 >20 (93:129) Female I.3" 0.8 <10 1.0 10-20 1.1 >20 Olsen et al. Minneapolis-St. Paul. MN Casexonirol Mule 2.5" 0.8 NP (1989) ((1980-83) (212:220) ACS CPS-11 (un'publitihed llnited States (1982-86) Prospective (421.663) Male 2.0 1.2 NP Cancer mortality tubulations) (605,75R) Female 2.7 1.6 Farrow and Davis Seaule, WA Catse:conu•ol Male 3.2 1.03 NP Adjusted for age, race, and lin press) (I9t{3-}{6) ( la2{:18}{) education N(YI'1:: NP=uut pruvidrJ: A('S ('I'S-I1=American ('amcer Society Cnncer Prrvcution Study II. . C'unipwrd wa a wcighted :rverage 1'rom Cig:u'ette Juu-.Ixxilic retalive ria.% presented io the paper. Weights are the numher ot'comrols within each stratum ol'smoking.

TABLE 5.-Bladder cancer risk according to smoking dose, duration of smoking, and smoking status Duration of Risk relative to never smokers Smoking dose (cig/day) smoking (yr) Current smokers Former smokers <20 <20 1.7 1.3 20-39 1.6 1.5 _40 2.7 1.9 20-39 <20 2.2 1.4 20-39 3.8 1.8 _40 3.1 2.5 >_40 <20 2.4 1.0 20-39 4.0 2.1 >40 3.8 2.8 SOURCE: HartRe et A (1987). varying design conducted throughout the world have shown statistically significant increases in risk of bladder cancer amona smokers. Cigarette smoking, determined to be a contributory factor in bladder cancer in past reports of the Surgeon General (US DHHS 1982. 1989). can now be identified as causally associated with bladder cancer. The evidence adequately meets the criteria for causality established in the 1964 Report (US PHS 1964). The decline in risk of bladder cancer with cessation further supports the conclusion that cigarette smoking causes bladder cancer. This diminution in risk cannot be explained by confounding from lower cumulative consumption among former smokers compared with continuing smokers. Cervical Cancer Recently, an association has been noted between cancer of the uterine cervix and cigarette smoking (Williams and Hotm 1977; Stellman, Austin, Wynder 1980; Lyon et al. 1983; Hellberg, Valentin, Nilsson 1983; BerQgren and Sjostedt 1983; Peters et al. 1986; Brock et al: 1988; Nischan, Ebeling, Schindler 1988). However, because of the possibility of confounding by unidentified factors (in particular, a sexually transmitted etiologic agent), this association has not been identified as causal (US DHHS 1982. 1989; IARC 1986). Components of tobacco smoke can be identified in the cervical mucus of smokers (Sasson et al. 1985; Schiffman et al. 1987). These compounds have been found not only to display mutagenic activity in this environment (Holly et al. 1986), but also to have the ability to impair local immunity by reducing the populations of Langerhans' cells within the cervical epithelium (Barton et al. 1988). The reduction in circulating levels of j3-carotene caused by cigarette smoking is yet another mechanism whereby cigarettes may increase the risk of cervical cancer (Harris et al. 1986; Brock et al. 1988; Stryker et al. 1988). Thus, the association of cigarette smoking with cervical cancer is biologically plausible. TIMN 438565 165

'fA13LE 6.-Continued Risk relative to never smokers Reference Location (yr of data collection) Design (number of subjects) Current smokers Former smokers Yr since quitting Comments Brisson el al. Quebec Case:control 3.5 1.9 NR Carcinoma in situ (1988) (1982-85) (247:137) I lerrero et al. 4 Central American cities Case:control 1.0 LO NR Adjusted for sexual partners (1989) (1986-87) (666:1,427) Slattery el al. Utah Case:control 3.4 1.4 NR Adjusted for sexual partners and education (1989) (1984-87) (266:408) ACS CPS-11 United States Prospective 2.1 1.9 NR Cancer mortality (unpubl ished (1 982-86) (605,758) tabulations) NOTE: NR=not repurtcd; SGS=,ocioeconomic slatus: ACS CPS-II=Amcrican Cancer Sox iely Cancer Prevention Study II. "Coniputed as a weighted average from cigarette dose-specilic rel:uive risks presented in the paper. Weights are the number of incident cases within each stratum of smoking. hComputed as a weighted average frum cigarette duse-.pecitic relative risks presented in Ihe paper. Weights are the number of controls within each stratum of smoking. 4b. U4 00 ~ ~ ~

TAI3LE 3.-Continued Risk rclative tu never stuuRers Rel'erence Population(yraf (Iatncollectiun) 1)csign (nwnik:ruf suhjectti) Ciender Current smul.ers Former smokers Yr since quilting Comments M;ick ct al. Los Angeles. CA C;1se:conlrul Male ;tnd 3.3" 3.3 <5 (1996) (1976-tt 1) (490:490) female 2.3 5-9 1.0" >10 Norell et al. Sweden Case:cuntrul M;tle and 1.6" 1.1 NP Datu for population controls (1986) ((1982-84) (98:134) lemale La Vecchia et al. (1987) Nurtlicrn Ituly (1983-86) Case:contrul (99:471) Male 1.6 1.4 NP Crude relative risk computed (51:134) Fenwle 1.1 0.9 NI' from d;ita presented Mills et al. Culi)iornia Prospective M;tle und 5.4 1.5 NP Cancer mortality study (1998) (1976--83) ( 3-1.(X)0) Irmale Felk et ;II. LUUISI;UYt C;ISC:L'OnIrOI Male and 1.1{" 1. 0 >_3 Adjusted liir diet und;tlcohol (1988) ((1979-83) (363:1234) Iemale Clavel et al. (1989) Paris, f=rance ( I9t{2-t{5 ) C;ttic:cuntrul (9ti:161) M;de 1.(i'' 1.1) NP Adjusted for alcohol und (63:1/)7) Fenrtle 1.5" 0.9 NI' col'Ice

(per 100,000 p.rson - y.ars) 700 600 600 400 300 200 100 0 40 45 6o aa Age ao 65 FIGURE 1.--Hypotheticai effects of smoking cessation on risk of CHD if mechanisms are predominantly rapidly reversible (A), irreversible (B), or slowly reversible (C). (CHD mortality rates shown in solid lines are for men in ACS CPS-II, 1982-86.) NOTE: CHD=coronary heart disease; ACS CPS-I1=American Cancer Society Cancer Prevention Study 11. SOURCE: Unpublished tabulations. American Cancer Society. modifiable risk factors for CHD" (US DHHS 1983, p.6). Overall, the Report noted that smokers have about a 70-percent excess death rate from CHD, and heavier smokers have an even areater excess risk. 198 TIMN 438596 .

INTRODUCTION Cigarette smoking is firmly established as an important cause of coronary heart disease (CHD), arteriosclerotic peripheral vascular disease, and stroke (US DHHS 1983, 1989). Eliminating smoking presents an opportunity for bringing about a major reduction in the occurrence of CHD, the leading cause of death in the United States. Before examining the epidemiologic evidence relating smoking cessation and risk of CHD and other forms of cardiovascular disease (CVD), the mechanisms by which smoking leads to these diseases are briefly reviewed. The objectives in considering these mechanisms are to address the plausibility that smoking cessation reduces risk of CVD, to estimate the expected magnitude in risk reduction, and to assess the rapidity with which any risk reduction might occur. Whether these mechanisms are immedi- ately reversible, irreversible, or slowly reversible is of particular relevance to the rapidity with which smoking cessation will reduce risk. The role of smoking in the pathogenesis of CHD is discussed at length. The etiologies of peripheral vascular disease and stroke share several common features with CHD: thus. discussion focuses on distinguishing features. PATHOPHYSIOLOGIC FRAMEWORK Smoking and Development of CHD Pathogenesis of CHD, which includes the clinical manifestations of myocardial infarction (MI), angina pectoris, and sudden death, is extremely complex and mediated by multiple mechanisms and etiologic factors (Munro and Cotran 1988). At least five interrelated processes are likely to contribute to the clinical manifestations of MI- atherosclerosis, thrombosis, coronary artery spasm, cardiac arrhythmia, and reduced capacity of the blood to deliver oxygen. Smokins appears to influence many steps in the development of CHD. Although not all of these effects are proven fully, the evidence for an influence on several mechanisms is convincing. The exact components of cigarette smoke that are responsible are not known in each instance, but experimental data have implicated nicotine and carbon monoxide (CO) in several processes. Other products of cigarette smoking, such as cadmium, nitric oxide, hydrogen cyanide, and carbon disulfide, have been hypothesized to play a role, but their quantitative contribu- tions remain unknown (US DHHS 1983). Atherosclerosis Atherosclerosis is the mechanical narrowing of medium-sized arteries by the proliferation of smooth muscle cells, lipid accumulation, and ultimately. plaque forma- tion and calcification (Munro and Cotran 1988). These lesions develop over decades and are not immediately reversible; whether they are substantially reversible at all in humans is a matter of current interest. Reversibility has been demonstrated in non- human primates (Clarkson et al. 1984; Malinow and Blaton 1984) and suggested in studies of humans using repeated arteriography (Blankenhorn et al. 1987). Smoking is TIMN 438589 ' 191

Breast Cancer In general, prior research has shown little relation between cigarette smoking and the risk of breast cancer (Baron 1984; Rosenberg et al. 1984; Baron et al. 1986); however, in recent years, several reports have raised the possibility that there might be a weak positive association (Table 7). Because there has been considerable discussion about the possible role of smoking in breast cancer in recent literature, the relationships among cigarette smoking, smoking cessation, and breast cancer risk are reviewed. Cigarette smoking creates a set of physiologic conditions that result in various antiestrogenic effects (Baron 1984; Jensen, Christiansen, Rodbro 1985; Michnovicz et al. 1986), as well as affecting body mass (Carney and Goldberg 1984; Hofstetter et al. 1986; Chapters 9, 10, 11). The relationship between cigarette smoking and body mass is a particularly important consideration in studies of breast cancer, because body mass has a complex age-dependent association with breast cancer risk, with obesity being protective in premenopausal ages but slightly risk-enhancing later in life (Willett et al. 1985). Table 7 summarizes findings from studies that have examined the relationship between breast cancer risk and the cessation of cigarette smoking. The risk of breast cancer among current smokers ranQes from less than 1.0 to 4.6 times greater than among never smokers (median approximately 1). The relative risks of smoking do not consistently differ in premenopausal and postmenopausal age groups. In addition, there is little consistency regarding the change in risk observed after smoking cessation. Former smokers have lower risks in some studies, but higher risks in others. Adjustment for other breast cancer risk factors does not appear to completely remove the weak association observed in some studies (Schechter, Miller, Howe 1985; Rohan and Baron 1989). In one study it was found that smokers tended to have a greater prevalence of tumor-positive axillary lymph nodes at the time of diagnosis than did never smokers and former smokers, a finding that could not be explained by patient delay (Daniell 1988). This association was not confirmed, however, in a recent report based on 10-year followup of the Nurses Health Study cohort that included 1,373 cases with information on extent of disease at diagnosis (London et al. 1989). This review of breast cancer and cigarette smoking suggests that cigarette smoking is not associated with breast cancer. Consistent changes in risk are not observed with smoking cessation. Endometrial Cancer The relationship between cigarette smoking and cancer of the endometrium is unique among the associations of smoking with cancers at various sites; of the sites for which smoking has been associated with a change in risk, endometrial cancer is the only cancer for which there is fairly consistent evidence of an inverse (protective) relationship (Baron 1984; Lesko et al. 1985; Stockwell and Lyman 1987), an effect that may be limited to postmenopausal women (Smith, Sowers, Burns 1984; Koumantaki et al. 1989). The reasons for the lower risk among women who smoke are not well under- TIMN 438569 169

similar to those derived from other studies, supporting an earlier observation of elevated risk for esophageal cancer in nondrinking smokers (Tuyns 1983). This review of past research on esophageal cancer and cigarette smoking cessation indicates that former smokers experience a lower risk of esophageal cancer than do current smokers, and that this lower risk is not because of confounding by lower alcohol intake among former smokers. Pancreatic Cancer The association, noted for many years, between smoking and cancer of the pancreas is considerably weaker than that between smoking and oral or esophageal cancer (US DHHS 1982). Although the causal mechanisms underlying this association are unclear, smoking has nonetheless been regarded as a contributing factor in cancer of the pancreas (US DHHS 1982, 1989). In the United States in 1985, the proportion of pancreatic cancer deaths attributable to smoking has been estimated to be 29 percent in men and 34 percent in women (US DHHS 1989). Table 3 summarizes studies of the relationship between pancreatic cancer and smoking cessation. In these studies. current smokers had risks ranging from 1.0 to 5.4 times (median of approximately 2) the risk among never smokers. Risks for pancreatic cancer associated with smoking were similar for males and females. Former smokers generally had lower risk than current smokers for pancreatic cancer, but the available data do not characterize adequately the change in risk with duration of abstinence. The larae case-control study conducted in Los Angeles, CA, (Mack et al. 1986) would suggest that risk is not substantially reduced until after 10 years of abstinence,.whereas the smaller English study (Cuzick and Babiker 1989) suggests that substantial risk reduction is moreyimmediate amona women than among men: risk reduction may take as lona as 20 years among men. This difference in the time course of risk after cessation according to gender has no clear biologic explanation and may be only a chance finding. ~ ~ y The question of potential confounding by differences in cigarette smoking exposure prior to quitting was addressed in the analysis of the U.S. Veterans Study (Kahn 1966). In each of four levels of past cigarette consumption, the risk amonQ former smokers was found to be lower than that amon2 current smokers. In the study conducted by Falk and colleagues (1988), former smokers had a lower risk of pancreatic cancer than current smokers at each of three levels of numbers of cigarettes consumed per day and also at each of four levels of numbers of years smoked. Because alcohol can cause insult to the pancreas and has been thought to be a possible pancreatic carcinogen (Cubilla and Fitzgerald 1979), two investigators adjusted for lifetime alcohol consumption in multiple logistic regression analyses (Falk et al. 1988: Clavel et al. 1989). These analyses produced relative risk estimates similar to those derived from other studies that did not adjust for alcohol and thus suggested that alcohol consumption is not a confounding factor in the smoking-pancreatic cancer association. The results of epidemiologic investigations on pancreatic cancer and cigarette smok- ing cessation indicate that there is a weak, but consistently observed, association between smoking and pancreatic cancer and that former smokers experience a lower TIMN 438555 155

TABLE 4.-Continued Risk relative to never smokers Reference Population (yr of datacollection) Design (numberof subjects) Gender Current smokers Former smokers Yr since yuitting Comments Slattery et al. Utah Case:controi Male 3.7 3.7 0.5-7 (1988) (1977-83) (332:686) 2.7 8-15 1.9 16-29 1.8 z3q Claude, Frenzel- Germany Case:control Male 3.5 1.8 NP Adjusted for number of Beyme, Kunze (1988) (1977-84) (531:531) cig/day ACS CPS-II (unpublished United States (1982-86) Prospective (421,663) Male 2.9 2.0 NP Cancer mortality tabulations) (6(15,758) Female 2.8 2.0 NP Burch et .d. (1989) Canada (1979-82) Case:control (627:601) Male 2.7 1.7 NP (19'):19O) Female 2.6 1.2 NP NO'PE: NP=not providcd; ACS CPS-11=American C'aneer Society Cancer Prevention Suidy I1. "CompuieJ as a weighted average from cigarette dose-specific relative risks presented in the paper. Weights ure the numlx'r ol'controls within each str:uum of smol.ing. hlncludes current and frxmer smokers who yuit in the past 2 yr. `Crude (unadjusted) odds ratio calculated front tables prescnled in the paper.

The results of two studies indicated that continued smokine after diagnosis of oral cancer may reduce survival, particularly in combination with alcohol consumption (Johnston and Ballantyne 1977; Stevens et al. 1983). These analyses, however, did not adjust for the more advanced stage of cancer among users of alcohol and tobacco at presentation (Johnston and Ballantyne 1977). The results of studies of oral cancer and cigarette smoking cessation indicate that former smokers experience a lower risk of oral cancer than current smokers and that this lower risk does not appear to be a result of confounding by alcohol or level of cigarette consumption prior to cessation. The risk of oral cancer has been shown to drop substantially within 3 to 5 years of cessation. Esophageal Cancer Smoking is a major cause of esophageal cancer (US DHHS 1982. 1989). In the United States, the proportion of esophageal cancer deaths attributable to tobacco has been estimated to be 78 percent for men and 75 percent for women (US DHHS 1989). As for cancer of the oral cavity, cigarette smoking is an independent risk factor for esophageal cancer but can also act in conjunction with alcohol to increase cancer risk. Table 2 summarizes the studies that have examined the relationship between smoking cessation and esophageal cancer risk. In these studies, the risk of esophaaeal cancer for current smokers rances from 1.7 to 6.4 times the risk among never smokers (median of approximately 5). These findings are similar to those for oral cancer as shown in Table 1. The risks for smoking and esophageal cancer were similar among males and females. Three years after.cessation, former smokers showed lower risks than current smokers in each study summarized in Table ?, with the exception of the Swedish prospective study (Cederlofet al. 1975) in which smokina-associated risks were considerably lower than in any other study. However, in followup of this cohort. more dramatic elevations in male mortality from esophageal cancer were observed in current smokers relative to never smokers: standardized mortality ratios were 1.1 for I to 7 Q tobacco per day. 4.5 for 8 to 15 a tobacco per day, and 5.4 for more than 15 Q of tobacco per day (Carstensen, Pershasen.-Eklund 1987). For former smokers, the standardized mortality ratio was 1.3. Approximately 3 to 5 years after cessation, risk of esophaszeal cancer was reduced by approximately 50 percent in the two studies providing information by duration of abstinence (Table 2). Data are very scant about the effects of cessation on the risk of esophageal cancer over long periods of abstinence. The U.S. Veterans Study showed that the risk among former smokers was lower at each of four levels of past numbers of cigarettes smoked per day. A multivariate analysis in which lifetime alcohol consumption was included as an adjustment factor (La Vecchia. Liati et al. 1986) produced relative risks for current and former smokers that were similar to those observed in other studies. In this study, the crude relative risk for ex-smokers was nearly identical to one that was adjusted for alcohol consumption (2.7 vs. 3.0). suagestinQ that alcohol was not a confounder in the estimates of the benefits of cessation. A study that was limited to nondrinkers (La Vecchia and Neari 1989) also produced risk estimates for smokinQ that were very 152 TIMN 438552

TABLE 6.-Studies of cervical cancer and smoking cessation Risk relative to never smokers Reference Location (yr of data collection) Design (number of subjects) Current smokers Former smokers Yr since quitting Comments Cederlof et al. Sweden Prospective 5.0 3.0 NR Cancer incidence (1975) (1963-72) (27,7(X)) Clarke, Morgan, Toronto, Ontario Case:control 2.3 1.7 NR Invasive cancer Newman (1982) (1973-76) (178:855) Marshall et al. f3uffalo, N Y Case:control 1.6 0.8 NR (1983) (1957-65) (513:490) Trevathan et al. Atlanta, GA Case:control 4.2 2.1 NR Carcinoma in situ (1983) (1980-81) (99:288) Adjusted for sexual partners, birth control pills, SES Greenberg et al. England Prospective 30 0.7 NR Invasive cancer incidence (1985) (1968-83) (17,032) - Adjusted for age at marriage, birth control pills. SES Brinton, Schairer, 5 US cities Case:control 1.5 3.2 1 Adjusted for sexual partners, age at first tlaenszel et al. (1986) (1982-84) (480:797) 1.1 2-4 intercourse, SES 1.0 5-9 1.1 _l0 La Vecchia, Milan, Italy Case:control 1.4h 2.5 NR Carcinoma in situ Franceschi et al. (1981-84) (183:183) 1.7 0.8 NR Invasive cancer (1986) (230:230)

TABLE 2.-Continued Reference I lanmiond and Garfinkcl (1969) (continued) ACS (unpublished tabulations) Relative risks compared with Number of cases never smokers" among Former Current Population Followup former smokers smokers smokers Comments 62 154 135 133 90 ACS CPS-11: 1.2 million men and women 4 yr for CHD deaths 14 48 47 88 90 359 19 33 36 67 71 182 Previously ?20 cig/day 2.55 Quit <I yr 1.61 (2.822 cases) 111yr 1.51 5-9 yr 1.16 10-14 yr 1.25 -15 yr 1.05 Men <21 cig/day Quit <1 yr 1.43 1-2 yr 1.61 3-5 yr 1.49 6-10 yr 1.28 11-15 yr 0.99 >_I6yr 0.88 1.93 Persons with cancer, heart disease, and stroke excluded at baseline Men _21 cig/day Quit <1 yr 2.56 I-2 yr 1.57 3-5 yr 1.41 6-10 yr 1.63 1 I-15 yr 1.16 ? 16 yr 1.09 2.02

LESKO, S.M., ROSENBERG, L., KAUFMAN, D.W., HELMRICH, S.P., MILLER, D.R., STROM, B., SCHOTTENFELD, D., ROSENSHEIN, N.B., KNAPP, R.C., LEWIS, J., SHAPIRO, S. Cigarette smoking and the risk of endometrial cancer. New England Journal of Medicine 313(10):593-596, 1985. LONDON, S.J., COLDITZ, G.A., STAMPFER, M.J., WILLETT, W.C., ROSNER, B.A., SPEIZER. F.E. Prospective study of smoking and the risk of breast cancer. Journal of the National Cancer Institute 81(21):1625-I631, November 1, 1989. LYON, J.L., GARDNER, J.W., WEST, D.W., STANISH, W.M., HEBERTSON, R.M. Smok- ing and carcinoma in situ of the uterine cervix. American Journal of Puhlic Health 73(5):558- 562. May 1983. MACDONALD, P.C.. EDMAN, C.D., HEMSELL, D.L., PORTER, J.C.. SIITERI, P.K. Effect of obesity on conversion of plasma androstenedione to estrone in postmenopausal women with and without endometrial cancer. American Journal of Obstetrics and Gynecology 130:448-455, 1978. MACK, T.M., YU, M.C., HANISCH, R., HENDERSON, B.E. Pancreas cancer and smoking, beverage consumption, and past medical history. Journal of the National Cancer Institute 76(1):49-60, January 1986. ' MACMAHON. B., YEN. S.. TRICHOPOULOS, D., WARREN, K., NARDI, G. Coffee and cancer of the pancreas. New EnglandJournal ofMedicine 304(1 l):630-633. March 12, 198 I. MARSHALL. J.R., GRAHAM. S., BYERS. T., SWANSOIV, M., BRASURE. J. Diet and smoking in the epidemiology of cancer of the cerv ix. Journal of the National Cancer Institute 70(5):847-851. May 1983. MICHNOVICZ, J.J., HERSHCOPF, R.J., NAGANUMA, H., BRADLOW, H.L.. FISHMAN. J. Increased 2-hydroxylation of estradiol as a possible mechanism for the anti-estrogenic effect of cisarette smoking. New En,~lan_d Journal of Medicine 315(21):1305-1309, Novem- ber 1986. ~ MILLS, P.K., BEESON, W.L., ABBEY, D.E., FRASER, G.E., PHILLIPS, R.L. Dietary habits and past medical history as related to fatal pancreas cancer risk among Adventists. Cancer 61(12):2578-2585, June 15, 1988. MOORE. C. Smoking and cancer of the mouth. pharynx, and larynx. Journal of the American Medic•al Association 191(4):107-110, January 25, 1965. MOORE, C. Cigarette smoking and cancer of the mouth, pharynx, and larynx. A continuing study. Journal of the American Medical Association 218(4):553-558, October 25, 1971. MORI. M., HARABUCHI, I., MIYAKE, H., CASAGRANDE, J.T., HENDERSON, B.E., ROSS, R.K. Reproductive, genetic, and dietary risk factors for ovarian cancer. American Journal of Epidemiology 128(4):771-777, October 1988. MORRISON, A.S., BURING, J.E., VERHOEK, W.G., AOKI. K.. LECK, I., OHNO, Y., OBATA. K. An international study of smoking and bladder cancer. Journal of Urology 13 l (4):650-654, April 1984. NISCHAN, P., EBELING. K., SCHINDLER, C. Smoking and invasive cervical cancer risk. Results from a case-control study. American Journal of Epidemiology 128(1):74-77, July 1988. NOMURA, A., GROVE. J.S., STEMMERMANN, G.N., SEVERSON, R.K. A prospective study of stomach cancer and its relation to diet, cigarettes, and alcohol consumption. Cancer Research 50:627-631, February 1. 1990. NORELL, S.E., AHLBOM, A., ERWALD, R., JACOBSON, G., LINDBERG-NAVIER. I., OLIN, R., TORNBERG. B., WIECHEL, K.L. Diet and pancreatic cancer: A case-control study. American Journal of Epidemiology 124(6):894-902, December 1986. OLSEN, G.W., MANDEL, J.S., GIBSON, R.W.. WATTENBERG, L.W., SCHUMAN. L.M. A case-control study of pancreatic cancer and cigarettes, alcohol, coffee, and diet. American Journal of Public Health 79(8):1016-1019, 1989. TIMN 438583 183

TABLE 2.-Continued Relative risks compared with never smokers° Number of cases Reference Popu lat ion among Fol lowup former smokers Former smokers Current smokers Comments . Doll et al. (1980) British physicians: 6,194 women 22 yr for Cl ID deaths 26 0.91 1.(1-2.2 Smoking assessed depending on at baseline and 9 u ft Previously <t ppd yr nt a er amo smoked Hammond and I lorn 187.783 men aged 50-60 44 mo for CI ID deaths 1.75 (1958a,b) 23 Quit < I yr 2.09 (143 cases) 80 1-10 yr 1.54 40 >10yr 1.09 Previously_1 ppd 2.20 18 Quit < I yr 3.00 (122 cases) 64 1-10 yr 2.06 40 >10 yr 1.60 Previously 1-19 cig/day Hammond and ACS CPS-l: 358,534 men free of 6 yr for CHD mortality 1.90 Garfinkel (1969) diagnosed CI ID 29 Quit <I yr 1.62 (1,063 cases) 57 1-4 yr 1.22 55 5-9 yr 1.26 52 10-14 yr 0.96 70 ?20 yr 1.O8

Anticipated Effects of Smoking Cessation on Risk of Cardiovascular Diseases Based on Knowledge of Mechanisms The possible effects of smokinQ cessation on the risk of CHD are illustrated in Figure 1. The incidence of CHD increases sharply with age among both smokers and never smokers; similar patterns are seen with other smoking-related cardiovascular diseases. At each age, the rates are higher for smokers. and the increase with age is more rapid among smokers (US DHHS 1983; ACS, unpublished tabulations), probably because of the ongoing, cumulative damage caused by smoking. Thus, the absolute excess incidence or mortality (attributable risk) of CHD due to smoking, represented by the vertical difference between the lines for smokers and never smokers in Figure 1, increases with age. However, the relative risk, represented by the ratio of incidence or mortality rates, tends to decrease with age. Theoretically possible outcomes of smoking cessation are depicted by lines A. B, and C(FiQure 1). Line A represents an immediate and complete reversal of the effect of smoking, so that the quitter almost instantly assumes the rate of the never smoker. Line B represents the worst-case scenario: although the stimulus for progressive damage is removed, no reversibility exists so that the former smoker assumes a constant absolute excess risk above that of the never smoker. In this case, it is apparent that quitting would still provide a substantial benefit compared with not quitting and that the relative risk for a former smoker compared with a never smoker would decline over time. An intermediate effect of smoking cessation is depicted by line C; the effects of smoking are slowly reversed, and the rate for the quitter gradually approaches that of the never smoker. The effects of smoking on CHD are probably mediated by multiple mechanisms. several of which are well established. Some of the effects of smokinQ appear to be reversible within days or weeks, including the increase in platelet activation, clottinQ factors. COHb, coronary artery spasm. and increased susceptibility to ventricular arrhythmias. Other effects may be irreversible or only slowly reversible, such as the development of atherosclerosis as a result of smooth muscle proliferation and lipid deposition in the arterial intima resulting from lower HDL-C levels. Thus, persons who stop smoking are likely to experience a component of rapid decline in risk compared with those who continue to smoke and another component that more slowly approaches the risk of never smokers. Because the effects of smoking are multiple and complex, the rapidity and magnitude of risk reduction achieved by smoking cessation can best be estimated by empirical data based on epidemioloaic studies in humans. Available data are examined in detail in the remaining sections of this Chapter. SMOKING CESSATION AND CHD Epidemiologic evidence on smoking and CHD has been reviewed in detail in previous reports of the U.S. Surgeon General (US PHS 1964; US DHEW 1971, 1979; US DHHS 1983, 1989). After an exhaustive review of the data, the 1983 Report of the Surgeon General concluded that "cigarette smoking is a major cause of CHD in the United States for both men and women" and "should be considered the most important of the known TIMN 438595 197

TABLE 2.-Continued Number of cases Relative risks compared with never smokers' Reference Population Followup aamong former smokers Former smokers Current smokers Comments ACS (unpublished Women <20 cig/day 1.76 tabulations) 3 Quit <I yr 2 13 (continued) 7 I-2yr . 0.87 11 3-5yr 1.31 12 6-IO yr 0.74 17 1 I-I 5 yr 1.20 82 >I6yr 1.17 Women _20 cig/day 2 27 9 Quit <1 yr 1.41 . 10 1-2 yr 1.16 16 3-5 yr 0.96 24 6-10 yr 1.88 12 11-I 5 yr 1.37 32 -16 yr 1.12 Dorn (1959); Kahn US veterans: 248,046 men 16 yr for 9,027 Stopped (overall) 1.15 1.58 Those who quit (1966); Rogot and b cardiovascular <5 yr 1.40 on doctor's orders Murray (1980) deaths 5-9 yr 1.40 were excluded 10-14 yr 1.30 15-I9 yr 1.20 ?20 yr 1.0O

clearly associated with the presence of atherosclerosis of the coronary arteries, small arteries of the myocardium, the aorta, and other vessels as demonstrated in many autopsy and angiographic studies (US DHHS 1983). The development of athero- sclerosis is complex, and several processes are likely to be important. Endothelial damage is thought to play a primary role in the development of atherosclerosis by exposing the arterial intima to blood lipids and white cells and by stimulating platelet adhesion. The endothelial damage can be an actual physical denudation, but toxic functional damage may have similar consequences. In animal studies, serum nicotine at levels similar to those of human smokers caused endothelial damage (Krupski et al. 1987; Zimmerman and McGeachie 1987). Evidence that smoking has a direct toxic effect on human endothelium is provided by the observation that smoking 2 tobacco cigarettes approximately doubled the number of nuclear- damaged endothelial cells in circulating blood (Davis et al. 1985, 1986); smoking non-tobacco cigarettes had little effect. In addition, Asmussen and Kjeldsen (1975) found pronounced degenerative changes of the umbilical artery endothelium at the time of delivery among mothers who smoked; these changes were not present in the arteries of nonsmoking mothers. Smooth muscle cell proliferation is a primary feature of atherosclerotic lesions and may result from several stimuli; the most clearly demonstrated is platelet-derived arowth factor from adherent platelets. Smoking appears to increase the adherence of platelets to arterial endothelium; blood drawn from persons after smoking 2 cigarettes results in a more-than-hundredfold adhesion of platelets to rabbit endothelium than does blood drawn from persons before smoking or from never smokers (Pittilo et al. 1984). Platelets from chronic smokers have a greater tendency to aggregate on an artificial surface than do those from nonsmokers (Rival. Riddle, Stein 1987). In minipicrs, both cigarette smoke and CO increase the adhesion of platelets to arterial endothelium (Marshall 1986). The influence of smoking on platelet activity is discussed further in the following section. Lipid infiltration of the arterial intima. largely cholesterol, is another primary feature of atherosclerosis and is directly related to higher blood levels of low-density lipoprotein cholesterol (LDL-C) and reduced blood levels of high-density lipoprotein cholesterol (HDL-C). Smoking reduces the level of HDL-C. A strong inverse associa- tion between daily cigarette consumption and HDL-C has been observed in many cross-sectional studies in the United States (Freedman et al. 1987; Gordon and Doyle 1986; Reichley, Mueller, Hanis et al. 1987; Willett et al. 1983) and in other countries (Assmann, Schulte, Schriewer 1984; Goldbourt et al. 1986: Gomo 1986: Jacobsen and Thelle 1987; Pelletier and Baker 1987; Robinson et al. 1987; Tuomilehto et al. 1986). In a loncitudinal, community-based study, HDL-C decreased among persons starting to smoke and increased amonQ those who stopped smoking (Fortmann. Haskell, Williams 1986). In other prospective studies, smoking abstinence has been associated with substantial increases in HDL-C levels in both men and women (Hulley, Cohen, Widdowson 1977; Hubert et al. 1987; Rabkin 1984). In a study among younQ adults in Louisiana, those who began smoking experienced substantial reductions in HDL-C compared with those who did not start (Freedman et al. 1986). HDL-C increased among 13 adult women who successfully stopped smoking for 48 days, but decreased to its 192 TIMN 438590

GARFINKEL. L.. STELLMAN, S.D. Smoking and lung cancer in women: Findings in a prospective study. Cancer Research 48(23):6951-6955, December 1, 1988. GRAHAM, S., LEVIN, M.L. Smoking withdrawal in the reduction of risk of lung cancer. Cancer 27(4):865-87 1, April 1971. HAENSZEL. W., LOVELAND, D.B., SIRKEN, M.G. Lung-cancer mortality as related to residence and smoking histories. I. White males. Journal of the National Cancer Institute 28:947-1001, April 1962. HAMMOND. E.C. Smoking in relation to the death rates of one million men and women. In: Haenszel, W. (ed.) Epidemiological Approaches to the Study of Cancer and Other Chronic Diseases. NCI Monograph 19. U.S. Department of Health, Education, and Welfare, Public Health Service, National Cancer Institute, January 1966, pp. 127-204. HIGGINS,_I.T., WYNDER, E.L. Reduction in risk of lung cancer among ex-smokers with particular reference to histologic type. Cancer 62(1 l):2397-2401, December l, 1988. HIGGINS, I.T.T.. MAHAN, C.M., WYNDER, E.L. Lung cancer among cigar and pipe smokers. Preventive Medicine I 7(1):116-128, January 1988. JOHNSTON-EARLY, A., COHEN, M.H., MINNA, J.D., PAXTON, L.M., FOSSIECK, B.E. JR., IHDE. D.C., BUNN, P.A. JR.. MATTHEWS, M.J., MAKUCH, R. Smoking abstinence and small cell lung cancer survival. Journal of the American Medical Association 244(19):2175-2179, November 14. 1980. JOLY. O.G.. LUBIN, J.H., CARABALLOSO. M. Dark tobacco and lung cancer in Cuba. Journal of the National Cancer Institute 70(6):1033-1039, June 1983. ~ KAHN, H.A. The Dom study of smoking and mortality among U.S. veterans: Report on eight and one-half years of observation. In: Haenszel, W. (ed.) Epidemiological Approaches to the Study of Cancer and Other Chronic- Diseases. NCI Monograph 19. U.S. Department of Health, Education, and Welfare, Public Health Service, National Cancer Institute. January 1966, pp. 1-125. . - LUBIN, J.H., BLOT, W.J. Assessment of lung cancer risk factors by histologic category. Journal of the National Cancer Institute 73(2):383-389, August 1984. LUBIN, J.H.. BLOT. W.J., BERRINO. F., FLAMANT, R., GILLIS, C.R., KUNZE, M., SCHMAHL. D., VISCO, G. Modifying risk of developing lung cancer by changing habits of cigarette smoking. British Medical Journal 288(6435):1953-I956. June 30. 1984a. LUBIN, J.H.. BLOT. W.J.. BERRINO, F.. FLAMANT, R.. GILLIS, C.R., KUNZE, M., SCHMAHL. D., VISCO. G. Patterns of lung cancer risk according to type of cigarette smoked. International Journal of Cancer 3:569-576, 1984b. ~ LUBIN, J.H.. RICHTER, B.S., BLOT, W.J. Lung cancer risk with cigar and pipe use. Journal of the National Cancer Institute 73(2):377-38 1, August 1984. MCDOWELL. E.M., HARRIS, C.C.. TRUMP, B.F. Histogenesis and morphogenesis of bronchial neoplasm. In: Shimosato. Y.. Melamed, M., Nettesheim, P. (eds.) Morphogenesis of Lung Cancer, Volume I. Boca Raton. Florida: CRC Press, 1982, pp. 1-36. MOOLGAVKAR. S.H.. DEWANJI. A., LUEBECK, G. Cigarette smoking and lung cancer: Reanalysis of the British doctors' data. Journal of the National Cancer Institute 81(6):=115- 420, March 15, 1989. MULLER, K.M.. KROHN, B.R. Smoking habits and their relationship ro precancerous lesions of the larynx. Journal of Cancer Research and Clinical Oncology 96(2):21 1-217, 1980. OLSEN. J., SABROE, S., FASTING, U. Interaction of alcohol and tobacco as risk factors in cancer of the laryneeal region. Journal of Epidemiology and Community Health 39(2):165- 168, June 1985. PATHAK, D.R., SAMET, J.M., HUMBLE, C.G.. SKIPPER, B.J. Determinants of lung cancer risk in cigarette smokers in New Mexico. Jocu-nal of the National Cancer Institute 76(4):597- 604, April 1986. TIMN 438542 139

ISCOVICH, J., CASTELLETTO, R.. ESTEVE, J., MUNOZ, N., COLANZI. R.. CORONEL, A., DEAMEZOLA, I., TASSI, V., ARSLAN, A. Tobacco smoking, occupational exposure and bladder cancer in Argentina. International Journal of Cancer 40(6):734-740, December 15, 1987. JENSEN, J., CHRISTIANSEN, C., RODBR!^b, P. Cigarette smoking, serum estrogens, and bone loss during hormone-replacement therapy early after menopause. New England Journal of Medicine 313:973-975, 1985. JENSEN, O.M., KNUDSEN, J.B., MCLAUGHLIN, J.K., SORENSON, B.L. The Copenhagen case-control study of renal pelvis and ureter cancer: Role of smoking and occupational exposures. International Journal of Cancer 41(4):557-561, April 15, 1988. JENSEN, O.M., WAHRENDORF, J., BLETTNER, M., KNUDSEN, J.B., SORENSEN, B.L. The Copenhagen case-control study of bladder cancer: Role of smoking in invasive and non-invasive bladder tumours. Journal of Epidemiology and Community Health 41(1):30- 36, March 1987. JOHNSTON, W.D., BALLANTYNE,.A.J. Prognostic effect of tobacco and alcohol use in patients with oral tongue cancer. American Journal of Surger;v 134:444-447, 1977. KABAT, G.C., HEBERT, J.R., WYNDER, E.L. Risk factors for oral cancer in women. Cancer Research 49(10):2803-2806, May 15, 1989. KABAT. G.C.. WYNDER, E.L. Type of alcoholic beverage and oral cancer. International Journal of Cancer 43(2):190-194, February 15, 1989. KAHN, H.A. The Dom study of smoking and mortality among US veterans: Report on eight and one-half years of observation. In: Haenzel. W. (ed.) Epidemiological Approaches to the Study of Cancer and Other Chronic Diseases. NCI Monograph 19. U.S. Department of Health, Education, and Welfare, U.S. Public Health Service, National Cancer Institute, January 1966. pp. 1-125. KINLEN, L.J.. ROGOT, E. Leukemia and smoking habits among United States veterans. British Medic•u/ Journal 297(6649):657-659, September 10, 1988. KOUMANTAKI.. Y., TZONOU, A., KOUMANTAKIS, E., KAKLAMANI. E., ARAVAN- TINOS, D., TRICHOPOULOS, D. A case-control study of cancer of the endometrium in Athens. lnternationalJournalofCancer43(5):795-799. May 15, 1989. KUCERA, H., ENZELSBERGER. H., EPPEL, W.. WEGHAUPT, K. The influence of nicotine abuse and diabetes mellitus on the results of primary irradiation in the treatment of carcinoma of the cervix. Cancer 60(l):1-4, July 1, 1987. LA VECCHIA, C., FRANCESCHI. S., DECARLI, A., FASOLI. M., GENTILE. A.. TOGNONI. G. Cigarette smoking and the risk of cervical neoplasia. American Journal of Epidemiology 123(1):22-29, January 1986. LA VECCHIA, C., LIATI, P., DECARLI. A.. NEGRELLO. I., FRANCESCHI. S. Tar yields of cigarettes and the risk of oesophageal cancer. International Jotu•nal of Cancer 38:381 --')85, 1986. LA VECCHIA, C.. LIATI, P.. DECARLI. A.. NEGRI, E.. FRANCESCHI. S. Coffee consump- tion and risk of pancreatic cancer. lnternatronal Jow•nal of Cancer 40:309-313, September 15,1987. LA VECCHIA. C., NEGRI. E. The role of alcohol in oesophageal cancer in non-smokers, and the role of tobacco in non-drinkers. Internutionul Journal of Cancer 43(5):784-785, May 15, t989. LAM, K.C.. YLJ, M.C., LEUNG. J.W.C.. HENDERSON, B.E. Hepatitis B virus and cigarette smoking: Risk factors for hepatocellular carcinoma in Hong Kong. Cuncer Resern•ch 42(12):5246-5248, December 1982. 182 TIMN 438582

STOCKWELL, H.G., LYMAN, G.H. Cigarette smoking and the risk of female reproductive cancer. American Journal of Obstetrics and Gynecology l57(1):35-40, July 1987. STRYKER, W.S., KAPLAN, L.A., STEIN, E.A., STAMPFER, M.J.. SOBER, A., WILLETT, W.C. The relation of diet, cigarette smoking, and alcohol consumption to plasma beta- carotene and alpha-tocopherol levels. American Journal of Epidemiology 127(2):283-296, 1988. TREVATHAN, E., LAYDE, P., WEBSTER, L.A., ADAMS, J.B., BENIGNO, B.B., ORY, H. Cigarette smoking and dysplasia and carcinoma in situ of the uterine cervix. Journal of the American Medical Association 250(4):499-502, July 22-29, 1983. TRICHOPOULOS. D., DAY. N.E., KAKLAMANI, E., TZONOU, A., MUiVOZ, N., ZAVITSANOS, X., KOUMANTAKI, Y., TRICHOPOULOU, A. Hepatitis B virus, tobacco smoking and ethanol consumption in the etiology of hepatocellular carcinoma. International Journal of Cancer 39(1):45-49, January 1987. TRICHOPOULOS, D., MACMAHON, B., SPARROS, L., MERIKAS, G. Smoking and hepatitis B-negative primary hepatocellular carcinoma. Journal of the National Cancer Institute 65(1):1 l 1-114, July 1980. TUYNS, A.J. Oesophageal cancer in non-smoking drinkers and non-drinking smokers. Inter- national Journal of Cancer 32(4):443-444, October 15, 1983. ~ U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking: Cancer. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Office on Smoking and Health. DHHS Publication No. (PHS) 82-50179, 1982. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Using Smokeless Tobacco. A Report of the Advisory Committee to the Surgeon General. U.S. Department of Health and Human Services. Public Health Service, National Institutes of Health. NIH Publication No. 86-2874, April 1986. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. Reducing the Health Conse- quences of Smoking: 25 Years of Progress. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion. Office on Smoking and Health. DHHS Publication No. (CDC) 89-8411, 1989. Smoking U.S. PUBLIC HEALTH SERVICE. Smoking and Health. Report of the Advison• Committee to the Sur geon General of the Public Health Ser vice. U.S. Department of Health, Education, and Welfare, Public Health Service, Center for Disease Control. PHS Publication No. 1103, 1964. VINEIS, P., CICCONE, G., GHISETTI, V., TERRACINI, B. Cigarette smoking and bladder cancer in females. Cancer Letters 26(1):61-b6, February 1985. VINEIS, P., ESTEVE, J., TERRACINI, B. Bladder cancer and smoking in males: Types of cigarettes, age at start, effect of stopping and interaction with occupation. International Journal of Cancer 34(2):165-170, August 15, 1984. VINEIS, P., FREA, B., UBERTI, E., GHISETTI, V., TERRACINI, B. Bladder cancer and cigarette smoking in males: A case-control study. Tumori 69(l):17-22, February 28, 1983. WHI'I'TEMORE, A.S., WU, M.L., PAFFENBARGER. R.S. JR., SARLES, D.L., KAMPERT, J.B., GROSSER, S., JUNG, D.L., BALLON, S., HENDRICKSON, M. Personal and environ- mental characteristics related to epithelia] ovarian cancer. II. Exposures to talcum powder, tobacco, alcohol, and coffee. AmericanJoeu•nal ofEpidemiology 128(6):1228-1240, Decem- ber 1988. WIGLE, D.T., MAO, Y., GRACE, M. Relative importance of smoking as a risk factor for selected cancers. Canadian Journal of Public Health 71(4):269-275, July-August 1980. TIMN 438585 185

EX-SMOKERS 00 40 r 1 5 10 15 20 lnterva/ Since Quitting Smoking (yr) FIGURE 2.-Estimated relative risk of MI after quitting smoking among men under age 55, adjusted for age; 95% CIs are indicated by vertical line; relative risk for men who never smoked is 1.0 NOTE: M1=myocardial infarction: Cl=confidence interval. SOURCE: Rosenberg. Kaufman. Helmrich. Shapiro (1985). with first infarction; their smoking histories were compared with those of 2,375 hospitalized controls. Amona former smokers overall, the relative risk of MI was 1.2 (95-percent CI) compared with never smokers; for current smokers the relative risk was 3.6. When former smokers were subdivided according to duration of abstinence, women who had stopped smoking within the previous 24 months had a relative risk of 2.6 (95-percent CI, 1.8-3.8). The relative risk was 1.3 for those who stopped smoking 24 to 35 months earlier. After 3 years of abstinence, relative risks ranged from 0.8 to 1.1 and were indistinguishable from that of women who had never smoked. 204 TIMN 438602

TABLE 2.--Continued N Relative risks compared with Number of cases never smokers" Reference Po ulation Follo,vu among Former Current Comments p p former smokers smokers smokers Cederlof et at. (1975) Saniple of 5 t, 911 Swedish l0 yr 97 Quit 1-9 yr 1.5 total 1.7 Only baseline men aged 18-69 Smoked <20 cig/day 0.9 smoking data used Smoked ?20 cig/day 1.6 86 Quit _ 10 yr 1.0 total Smoked <20 cig/day 0.9 Smoked ?20 cig/day 1.1 Fuller et al. (1983) Whitehall civil servants: 1(l yr for 208 171 normo-glycemic 1.3 18,403 men aged 40-64 Cl iD deaths 23 glucose intolerant 0.7 14 diabetics 3.8 Friedman et al. (1981) 25,917 Kaiser-Pennanente 4 yr for 31 0.9 subscribers in the San Francisco C11D deaths area, aged 20-79 2.5 Prevalent cases of 1.5 CHD not excluded 2.9 1.6 Prevalent cases of CND not omitted; exclusion of those cases increased the apparent benefit of quitting

TABLE 3.- Estimated probability of dying from ischemic heart disease in the next 16.5-year interval (95% CI) for quitting at various ages compared with never smoking and continuing to smoke, by amount smoked and sex Age at quitting or at start of interval MEN 40-44 45-49 50-54 55-59 60-64 65-69 70-74b Continuing smokers Former smokers Never smokers <21a ?21' <21a » la 0.01 0.03 0.03 0.01 0.02 (.01-.01) (.02-.03) (.03-.04) (.00-.02) (.01-.02) 0.02 0.04 0.04 0.02 0.02 (.01-.02) (.04-.05) (.04-.05) (.01-.03) (.01-.03) 0.04 0.07 0.06 0.04 0.04 (.03-.03) (.06-.07) (.06-.07) (.03-.05) (.02-.05) 0.05 0.10 0.09 0.05 0.08 (.05-.06) (.08-.1 I ) (.07-.10) (.04-.07) (.06-.10 0.10 0.14 0.16 0.12 0.10 (.09-.11) (.12-.16) (.10-.21) (.09-.15) ' (.06-.15) 0.15 0.20 0.13 0.14 0.12 (.13-.17) .(.16-.25) (.08-.19) (.07-.)1) (.00-.24) 0.13 0.17 0.10 0.19 0.11 (.11-.14) (.13-Z2) (.05-.16) (.10-.29) (.02-.20) sick at interview or giving a history of heart disease, cancer, or stroke. For both women and men, during the next decade-and-a-half cumulative CHD mortality for those who stopped smoking before age 60 was about half that of those who continued to smoke. This same pattern of reduced risk extended to those who stopped smoking between ages 60 and 64. After age 65, few persons stopped smoking, as indicated by wide confidence intervals, so that no clear patterns could be determined. Because the methods used in CPS-I and CPS-II are similar, it is appropriate to compare the results of the two studies. In CPS-II, the relative risks of CHD for current smoking among men and women are substantially higher at every age than those observed in CPS-I. The higher relative risks for CHD and other smoking-related diseases among women in CPS-II are possibly due to the earlier age of, smoking 216 TIMN 438614

TABLE 1.-Case-control studies of CHD risk among former smokers Number of Relative risk as comp ~red with never smokers Reference Population Number of cases Number of controls Source of controls cases among former smokers Former smokers Current smokers Willett et al. Nurses f leulth Study: women 263 -~,260 Nested in cohort 29 Overall (1981) aged 30-55 , 1.0 (0.7-1.6) 3.0 (2.3-J.0) Quit 1-4 yr 1.5 (0.7-3.1) Quit 5-9 yr 1.5 (0.8-3.0) Quit _10 yr 0.6 (0.3-1.3) Rosenberg, Eastern US men aged <55 1,873 2.775 1lospital-based 348 1.1 (0.9-1.4) 2.9 (2.4-3.4) Kaufman, l-lelmrich, Shapiro (1985) Rosenberg, Eastern US women aged <50 555 4,864 I lospital-based 35 1.0 (0.7-1.6) 1.4-7.0 depending on Kaufman, Iielmrich, , cig/day Miller et al. (1985) ' LaVecchia et al. Italian women aged <55 168 251 liospital-based 3 0.8 (0.2-3.8) 3.6-13.1 depending (1987) on cig/day

TABLE 2.-Cohort studies of CHD risk among former smokers Relative risks compared with never smokers" Number of cases among Former Current Reference Population • Followup former smokers smokers smokers Comments DoII and IiiII (1964) British physicians: 34,445 men IO yr for CIID deaths 28 Quit 1-4 yr 1.05 1.41 61 5-9 yr 1.25 59 10-14 yr 1.16 40 _15yr 1.12 DoII and Peto (1976) British physicians: 34,440 men 20 yr for CHD dealhs Aged 3(}-54 3.5 7 Quit 1-4 yr 1.9 10 5-9 yr 1.3 10 10-14 yr 1.4 7 zl5 yr 1.3 Aged 55-64 . 1 7 19 Quit 1-4 yr 1.9 . 34 5-9 yr 1.4 38 10-I4 yr 1.7 45 _15yr 1.3 Aged ?65 1.3 24 Quit 1-4 yr 1.0 76 5-9 yr 1.3 62 10-14 yr 1.2 148 ->15 yr 1.1 Smoking ascertained 195 1, updated 1958 Smoking data assessed at batieline and after 7 yr

OSHIMA. A.. TSUKUMA, H., HIYAMA. T.. FUJIMOTO, L. YAMANO. H., TANAKA. M. Follow-up study of HBs Ag-positive blood donors with special reference to effect of drinking and smoking on development of liver cancer. Internatiorrcrl Jotrrnal of Canc•er 34:775-779, 1984. ~ PETERS. R.K., THOMAS. D.. HAGAN, D.G.. MACK, T.M.. HENDERSON, B.E. Risk factors for invasive cervical cancer among Latinas and non-Latinas in Los Angeles County. Jottrrral of the National Cancer Institute 77(5):1063-1077, November 1986. ROGOT, E.. MURRAY. J.L. Smoking and causes of death among U.S. veterans: 16 years' of observation. Pnhlic• Health Repar•ts 95(3):213-222, May/June 1980. ROHAN. T.E.. BARON. J.A. Cigarette smoking and breast cancer. Anrerican Jotu•nal of Epidenticrltr~qy 129(1):36-42, January 1989. ROSENBERG. L.. SCHWINGL. P.J.. KAUFMAN. D.W.. MILLER, D.R., HELMRICH, S.P., STOLLEY. P.D.. SCHOTTENFELD, D.. SHAPIRO, S. Breast cancer and cigarette smoking. New Etrglctnd Jottrnal of Medicine 3 l0(2):92-94, January 12, 1984. SASSON, I.M., HALEY. N.J., HOFFMANN, D., WYNDER, E.L., HELLBERG. D.. NILSSON. S. Cigarette smoking and neoplasia of the uterine cervix: Smoke constituents in cervical mucus.• New England Journal of Medicine 312(5):315-316, January 31. 1985. SCHECHTER, M.T.. MILLER, A.B., HOWE, G.R. Cigarette smoking and breast cancer: A case-control study of screening participants. AmericanJournal crf Epidemiolcr,gv 121(4):479- 487, April 1985. SCHIFFMAN. M,H., HALEY. N.J., FELTON, J.S., ANDREWS, A.W., KASLOW, R.A., LANCASTER. W.D., KURMAN, R.J.. BRINTON. L.A.. LANNOM, L.B.. HOFFMANN, D. Biochemical epidemiology of cervical neoplasia: Measuring cigarette smoke constituents in the cervix. Caru•er Research 47(14):3886-3888, July 15, 1987. SCHOTTENFELD. D. Multiple primary cancers. In: Schottenfeld, D.. Fraumeni, J.F. Jr. (eds.) Cancer Epidentiolo,q}' and -Pt'erention. Philadelphia: W.B. Saunders, Co., 1982, p. 1025. SCHOTTENFELD, D., GANTT, R.C., WYNDER. E.L. The role of alcohol and tobacco in multiple primary cancers of the upper digestive system, larynx and lung: A prospective study. Prerentire Medicine 3(2):277-293, June 1974. SEVERSON, R.K. Cigarette smoking and leukemia. Cancer 60(2):141-144, July 15, 1987. SILVERMAN. S. JR.. GORSKY, M., GREENSPAN, D. Tobacco usage in patients with head and neck carcinomas: A follow-up study on habit changes and second primary oral/oropharyngeal cancers. Journal of the American Dental Association 106(1):33-35, January 1983. SLATTERY, M.L., ROBISON, L.M.. SCHUMAN, K.L., FRENCH, T.K.. ABBOTT, T.M., OVERALL. J.C. JR.. GARDNER, J.W. Cigarette smokins and exposure to passive smoke are risk factors for cervical cancer. Journal of the American Medical Association 261(1 1):1593-1598,1Vtarch 17, 1989. SLATTERY, M.L., SCHUMACHER, M.C., WEST, D.W., ROBISON, L.M. Smoking and bladder cancer. The modifying effect of cigarettes on other factors. Cancer 61(2):402-408, January 15, 1988. SMITH, E.M.. SOWERS. M.F., BURNS, T.L. Effects of smoking on the development of female reproductive cancers. Journal of the National Cancerlnstitrrte 73(2):371-376, August 1984. SPITZ, M.R., FUEGER, J.J., GOEPFERT, H., HONG, W.K., NEWELL, G.R. Squamous cell carcinoma of the upper aerodigestive tract. A case comparison analysis. Cancer 61:203-208, 1988. STELLMAN, S.D., AUSTIN, H., WYNDER, E.L. Cervix cancer and cicarette smokin~: A case-control study. American Journal of Epidemiology 1 11(4):383-388, April 1980. STEVENS, M.H., GARDNER, J.W., PARKIN, J.L., JOHNSON, L.P. Head and neck cancer survival and life-style change. Archives of Otolmyngology 109(1 1):746-749, November 1983. 184 TIMN 438584

CONTENTS Introduction ....................................................... 191 Pathophysiologic Framework ......................................... 191 Smoking and Development of CHD .................................. 191 Atherosclerosis ................................................. 191 Thrombosis ..................................................... 193 Spasm ........................................................ 195 Arrhythmias ................................................... 195 Reduced Blood Oxygen Delivery ..................... .......... . . . 195 Smoking and Development of Peripheral Arterial Disease ................. 196 Smokin- and Development of Cerebrovascular Disease .................. 196 Anticipated Effects of SmokinQ Cessation on Risk of Cardiovascular Diseases Based on Knowled~e of Mechanisms ............................... 197 Smokina Cessation and CHD ......................................... 197 Cross-Sectional Studies ............................................ 199 Studies of Smokinc, Cessation and Risk of MI Amona Healthy Persons ......?00 Case-Control Studies ............................................. '200 Cohort Studies ................................................. 205 Intervention Trials ..............................................2?4 Smokine Cessation and CHD Risk Amona Persons With Diaanosed CHD ...??9 Summary of Smoking Cessation and CHD Risk ......................... 239 Smoking Cessation and Aortic Aneurysm ................................ 241 Studies of Smoking Cessation and Risk of Aortic Aneurysm ............... 241 Smokincy Cessation and Peripheral Arterial Occlusive Disease ............... 241 Smoking Cessation and Development of Peripheral Artery Disease .......... 43 Smokina Cessation and Pro-nosis of Peripheral Artery Disease ............ ?43 Summary ........................................................ 244 Smoking Cessation and Cerebrovascular Disease .......................... 245 Studies of Smokin2 Cessation arnd Risk of Cerebrovascular Disease .........?46 Cross-Sectional Studies .......................................... 246 Case-Control Studies ............................................. 246 Prospective Cohort Studies ........................................ 249 Summary of Observational Studies ................................. 251 Intervention Studies ............................................. 251 Influence of Prior Levels of SmokinQ ............................... ?51 Effect of Duration of Abstinence ................................... - 252 Oral Contraceptives and Smoking Cessation ......................... 258 Effect of Smoking Cessation After Stroke ........................... 260 Summary ........................................................ 260 Conclusions ........................................................ 260 References ........................................................ ?61 1TIMN 438588 s9

TABLE 3.-Continued Age at quitting Continuing smokers Former smokers or at start of Never interval. smokers <20a 220a <20" >-20a WOMEN 40-44 45-49 50-54 55-59 60-64 65-69 70-74b 0.00 (.00-.00) 0.00 (.00-.01) 0.01 (.01-.01) 0.02 (.02-.02) 0.04 (.03-.04) 0.07 (.07-.08) 0.07 (.06-.07) 0.01 (.00-.01) 0.01 (.01-.01) 0.02 (.02-.03) 0.04 (.03-.05) 0.06 (.04-.07) 0.11 (.07-.15) 0.09 (.05-.13) 0.01 (.01-.01) 0.01 (.01-.02) 0.03 (.02-.03) 0.05 (.04-.06) 0.08 (.06-.10) 0.12 (.07-.18) 0.11 (.05-.16) 0.00 (.00-.01) 0.00 (.00-.00) 0.01 (.00-.02) 0.01 (.00-.02) 0.02 (.00-.05) 0.12 (.03-.21) 0.03 (.00-.08) 0.00 (.00-•01) 0.01 (.00-.01) 0.02 (.01-.02) 0.02 (.0 t-.04) 0.04 (.01-.06) 0.09 (.01-.17) 0.02 (.00-.05) NOTE: Based on subjects not sick at enrollment or giving a history of cancer, heart disease, or stroke: 95~7c confidence interval (Cl) shown in parentheses. ' Cig/day. ° Estimates for quitting at this age are estimates of the probability of dying in the next 12.5-yr interval. SOURCE: Unpublished tabulations. American Cancer Society. initiation in the more recent cohort (US DHHS 1989). The higher relative risks among men are more difficult to explain because the age of initiation has not changed substantially among men over time (US DHHS 1989). The large size and careful methodology of the three ACS cohorts provide consider- able evidence for the benefit of quitting in reducing risk of CHD. These studies also provide strong evidence that there is some residual risk of CHD attributable to past smoking that persists for a considerable duration after cessation. The U.S. Veterans Study (Dorn 1959; Kahn 1966; Rogot 1974; Rogot and Murray 1980) has also provided useful information on the health effects of smokinQ. The population was drawn from 293,958 U.S. veterans who held Government life insurance policies in December 1953. In 1954, a total of 198,820 individuals returned mailed 217 TIMN 438615

FRANKS. A.L., LEE. N.C., KENDRICK. J.S., RUBIN, G.L., LAYDE, P.M., THE CANCER AND STEROID HORMONE STUDY GROUP. Cigarette smoking and the risk of epithelial ovarian cancer. American Journal of Epidemiology 126(1):112-117, 1987. GOLD, E.B.. GORDIS, L., DIENER. M.D.. SELTSER, R., BOITNOTT,I.K., BYNUM. T.E.. HUTCHEON, D.F. Diet and other risk factors for cancer of the pancreas. Cancer 55(2):460- 467, January 15. 1985. GREENBERG, E.R., VESSEY, M., MCPHERSON, K., YEATES, D. Cigarette smoking and cancer of the uterine cervix. British Journal of Cancer 51:139-141, January 1985. ~ GUPTA. P.C.. PINDBORG, J.J., BHONSLE. R.B.. MURTI. P.R.. MEHTA, F.S., AGHI. M.B., DAFTARY, D.K.. SHAH. H.T.. SINOR. P.N. Intervention study for primary prevention of oral cancer amons 36.000 Indian tobacco users. Lancet 1(8492):1235-1239. May 31, 1986. HARRIS. R.W.. FORMAN, D., DOLL. R.. VESSEY. M.P.. WALD, N.J. Cancer of the cervix uteri and vitamin A. British Journal of Cancer 53(5):653-659. May 1986. HARTGE. P.. SILVERMAN. D.. li-IOOVER, R.. SCHAIRER. C.. ALTMAN. R.. AUSTIN. D., CANTOR. K.. CHILD. M.. KEY. C.. MARRETT, L.D., MASON. T.J.. MEIGS. J.W.. MYERS. M.H.. NARAYANA. A.. SULLIVAN. J.W.. SWANSON. G.M.. THOMAS. D.. WEST, D. Changing cigarette habits and bladder cancer risk: A case-control study. Journal uf'tlre Natronul Cancer Institute 78(6):1119-1 125. June 1987. HELLBERG. D.. NILSSON, S. Smoking and cancer of the ovary. (Letter.) New England Jcrurncrl ol',Yleclicine 318(12 ):782-783, March 24, 1988. HELLBERG. D.. VALENTIN. J.. EKLUND. T.. STAFFAN. N. Penile cancer: Is there an epidemiological role for smoking and sexual behaviour'? British Medic•ul .Iotu•rurl 295(6609):1 i06-1308, November 21. 1987. HELLBERG. D.. VALENTIN. J.. NILSSON. S. Smoking as a risk factor forcervical neoplasia. ( Letter. ) Lurrcet 2( 8365/8366):1-I97, December 24-3 l, 1983. HERRERO. R.. BRINTON. L.A.. REEVES. W.C.. BRENES. M.M.. TENORIO. F.. DEBRIT- TON, R.C.. GAITAN. E.. GARCIA, M.. RAWLS, W.E. Invasive cervical cancer and smokina in Latin America. Journcrl nftleNcrtiuncrlCuncerlnstrtute 81(3):205-21 l, February I, 1989- H1ATT. R.A.. FIREMAN. B.H. Smoking. menopause, and breast cancer. Journal of the Nutiortul Cuncerlnstitute 76:833-838. 1986. HIRAYAMA. T. A larse-scalecohort study on risk factors for primary livercancerr with special reference to the role of cigarette smokin_. Curlcer Chc~motherupr urrcl Plrurntucolo~ ti• 23(Sup- plement ):S I 1-~-S 117. 1989. ~ HOFSTETTER. A.. SCHUTZ. Y.. JEQUIER. E., WAHREN. J. Increased 24-hour energy expenditure in cigarette smokers. (Letter.) New Errglurrd.Ituu•nal of Medicine 31-I(25):164 1, June 19. 1986. HOLLY, E.A.. PETRAKIS. N.L., FRIEND, N.F.. SARLES. D.L.. LEE. R.E.. FLANDER. L.B. Mutagenic mucus in the cervix of smokers. Journal of the Nuticrnul Cuncer Institute 76(6):983-986, June 1986. HOLMES. F.. BOREK, D.. OWEN-KUMMER. M., HASSANEIN. R.. FISHBACK. J., BEHBEHANI. A.. BAKER. A.. HOLMES. G. Anal cancer in women. Gusn•oenterolog}• 95( I):107-1 11. July 1988. HOWE. G.R.. BURCH. J.D.. MILLER. A.B.. COOK. G.M.. ESTEVE. J.. MORRISON. B.. GORDON. P., CHAMBERS. L.W., FODOR. G.. WINSOR. G.M. Tobacco use. occupation. coffee, various nutrients, and bladder cancer. Journal of the National Cancer Institute 64(4):701-713. April 1980. INTERNATIONAL AGENCY FOR RESEARCH ON CANCER. Tobacco Smol;ln,q, IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. Volume 38. Lyon: International Agency for Research on Cancer, 1986. TIMN 438581 181

N N TABLE 2.-Continued Number of cases Relative risks compared with never smokers' Reference Population among Followup former smokers Former smokers Current smokers Keys (1980) 7-Countries Study of 12,096 men free of CI ID 10 yr for About 13` Cl ID deaths (Northern Europe) About 9 (Italy, Greece, Yugoslavia) About 7` (US) 2.3 0.8 0.7 2.4-4.5 depending on amount 0.7-I .8 depending on amount 1.6-3.0 depending on amount Shapiro et al. (1969) I IIP cohort about 39,000 men aged 35-64 3 yr for MI NR 1.0 1.8 Jajich, Ostfeld, Freeman ( 1984) 2,674 poor persons in Cook County, II_ aged 64-75 4.5 yr lior . 20 C I ID deaths 1.11 1.94 Willett et al..(1987) Nurses I lcalth Study: 121,70(1 US women abed 30-55 6 yr li>r nonfatal 55 MI and CI1D deaths 1.5 (1.0-2.1) 2.1-10.8 depending on amount smoked Comments Relative risk based on only about 5 cases in never smokers, very small numbers Numbers extrapolated from figures Stroke excluded but prevalent CI ID not excluded at baseline 0

with 10 years or more of abstinence, the risk was nearly the same as that for never smokers. A similar pattern was observed among those smoking lI pack or more per day. Current smokers at that level had a relative risk of 2.55. Quitters of less than 1 year had a relative risk of 1.61, and those with between 10 and 20 years of abstinence had only a slightly elevated relative risk of 1.25. Because of the very large number of deaths and the careful followup, the estimates of effect are relatively precise. In this period, cigarette smoking declined substantially, especially in the predominantly white, mid- dle- to upperciass a oups represented by the study population. Hence, some misclas- sification of the current smoking group may have occurred, but the relative risks among former smokers, apart from the most recent quitters (some of whom inevitably resumed smoking), are likely to be accurate. In 1982, a third ACS cohort, CPS-lI, was initiated in 50 States. The methods for recruitment and the population enrolled were similar to CPS-I, but the cohort was larger, with more than 1.2 million participants (Chapter 3). Preliminary data based on 4 years of followup were published in the 1989 Surgeon General's Report (US DHHS 1989). Among men, former smokers aged 35 or younger had relative risks of CHD of 1.41, those aged 36 to 64 had 1.75, and those 65 or older had 1.29; the relative risks among current smokers were 1.94, 2.81, and 1.62, respectively. A generally similar pattern was seen among women. When the data are examined by amount of previous smoking and time since quitting, the pattern of changing risk is influenced by the presence of disease at enrollment. When those who reported themselves as sick or as having previously diagnosed cancer, heart disease, or stroke at baseline were not excluded from the analysis, men who previously smoked fewer than 21 cigarettes per day and who had quit smoking within the previous 3 years experienced a CHD mortality rate that was about 6 percent higher than that among current smokers. However, with increasing duration of abstinence, the risk among former smokers came very close to that of never smokers; after 16 years or more, the relative risk was 1.01 (US DHHS 1989). It is likely that the early peak in mortality among recent quitters partly reflects the effect of having included those who quit because of smoking-related illness. After excluding those with cancer, heart disease, and stroke at baseline, this early excess mortality is less apparent (Table 2). In all categories, those who quit I to 2 years earlier had relative risks substantially lower than those of current smokers. Findincrs are less consistent for those who quit within the past year, presumably because of a high incidence of smoking resumption in that group and the possible inclusion of persons who stopped smoking as a result of symptoms due to undiagnosed illness. A very similar pattern was observed among, men who smoked 21 cigarettes or more per day, except that the relative risks were higher for all but those with the shorter period of abstinence. The absolute rates were lower for women, as expected, and the relative risks are thus statistically unstable. Neverthe- less, the overall patterns among female smokers were aenerally similar to those among male smokers. y y To examine the effects of smoking cessation at different ages, CPS-II data on cumulative mortality rates due to CHD were tabulated for 5-year categories of age at cessation. (See Table 3 and Chapter 3 for a description of the methods used to calculate these rates.) The mortality rates used for these calculations were based on subjects not TIMN 438613 215

tJ O TABLE 2.--Cilntintted Reference Dorn (1959); Kahn (1966); Rogot gnd Murray(1980) (continued) Doyle et al. (1962) Doyle et al. (1964) Gordon, Kannel, McGee (1974) Population Frainingham and Albany cohorts 4,1201tealthy men aged 30-62 Fruntingham and Albany cohorts of 4.120 tte:tlthy men aged 30-62 2,336 men in Framingham I iutrt Study, aged 29-62 ~ Rosenman et al. (1975) 3, 154 healthy California 00 men nged 39-59 CZ'1 ~ 00 Number of cases Relative risks compared with never smokers" among Followup former smokers Former smokers Current smokers For CI ID deatSts Stopped (overall) 1.16 1.58 <5 yr 1.40 5-9 yr 1.40 10-14 yr 1.30 15-19 yr 1.20 ?20 yr 1.10 6-8 yr for fatal wui nonfatal M! 10 0.9 2.3 10 yr (Framingltam) 8 yr (Albany) MI and Cl ID deaths 13 1.1(0.5-2.2) 2.0-3.0 depending on amount smoked 18 yr for CI IE) excluding angina 24 0.7 1.3 8-9 yr li>r fatal and 16 Aged 39-40 1.9 2.5 nonfatal CI11) Aged 50-59 1.1 Comments No update of smoking informat ion Only baseline' smoking data used No data on duration Smoking information updated biennially Only baseline smoking data used

Since 1983, additional evidence has accumulated to further support these con- clusions. Some of these data were presented or summarized in the 1989 Report of the Surgeon General (US DHHS 1989). For 1985, cigarette smoking was estimated to be responsible for 21 percent of all CHD deaths in the United States among men aged 65 years or older and for 45 percent of CHD deaths among younger men. Twelve percent of the CHD deaths among women aged 65 or older and 41 percent of those in younger women were attributed to cigarette smoking. In 1985, 115,000 deaths from CHD were attributed to cigarette smoking. A large amount of data supports the view that active cigarette smoking substantially increases risk of CHD. Data also indicate that former s'mokers have a lower risk of CHD than do current smokers. Despite methodologic and geographic differences, the studies are remarkably consistent in demonstrating a reduced risk of CHD among former smokers. Much of this literature has been reviewed in earlier reports of the Surgeon General (US DHEW 1979: US DHHS 1983) as well as by Kuller and colleagues (1982). This Section reviews the epidemiologic evidence of the effects of cigarette smoking cessation on CHD risk, specifically MI and CHD death. The relevant studies may be divided into those that examine the effect among apparently healthy individuals (primary prevention) and the effect among individuals already diagnosed with CHD for risk of recurrence or CHD death (secondary prevention). Cross-sectional studies of the extent of coronary atherosclerosis also provide relevant information. Cross-Sectional Studies In a detailed study of coronary atherosclerosis, Auerbach and coworkers (1976) examined 1,056 autopsied hearts from patients at the East Orange Veterans Administra- tion Hospital and found that smokers had more severe disease than never smokers, with past smokers having intermediate levels. Those who died from CHD or diabetes or those who had hearts weighing more than 500 a were excluded. After adjustment for age, current cigarette smokers had a prevalence of advanced CHD that ranged from 11.7 to 23.4 percent, depending on the number of cigarettes smoked per day. The prevalence among never smokers was 5.3 percent compared with 11.0 percent among former smokers. The prevalence odds ratio of advanced versus no disease or minimal disease was 2.4, when former smokers were compared with never smokers. In contrast, among current smokers of I to 2 packs per day, the ratio was 6.7. A similar pattern was observed for different pathologic manifestations of CHD. The effect of duration of abstinence amonc, former smokers was not analyzed. Ramsdale and coworkers (1985) used arteriography to assess the extent of coronary atherosclerosis before surgery for valve replacement among 387 patients. All patients provided a smoking history, including age at initiation of smokina and cessation of smoking and average number of cigarettes smoked per week. Among never smokers, 87 percent had no stenosis greater than 50 percent; only 60 percent of past smokers and 60 percent of current smokers were without this degree of stenosis. Of never smokers, only 2.6 percent had three or more arteries affected compared with 10.6 percent of former smokers and 12.2 percent of current smokers. Both current and past smokers TIMN 438597 199

BYERS. T.. MARSHALL. J.. GRAHAM. S.. METTLIN, C.. SWANSON, M. A case-control study of dietary and nondietary factors in ovarian cancer. laur-ncrl of the National Cancer Institute 71(4):681-686. October 1983. CARNEY. R.M.. GOLDBERG. A.P. Weight gain after cessation of cigarette smoking. A possible role for adipose-tissue lipoprotein lipase. New Englund Jvru'turl of Medicine 310(10):61-3-616. March 8. 1984. CARSTENSEN. J.M.. PERSHAGEN. G.. EKLUND. G. Mortality in relation to cigarette and pipe smoking: 16 years' observation of 25.000 Swedish men. Jow-nal of Epidemiolo,qti• and Cummwtin• Health 41:166-172, 1987. CARTWRIGHT. R.A.. AD[B. R.. APPLEYARD. I.. GLASHAN. R.W.. GRAY. B., HAMILTON-STEWART. P.A.. ROBINSON, M.. BARHAM-HALL. D. Cigarette smoking and bladder cancer: An epidemiolo,ic inquiry in West Yorkshire. Jow-nal of Epideminlc,,y unclCnnrnrturirr Health 37(4):?56-363, December 1983. CASTIGLIANO. S.G. Intluence of continued smoking on the incidence of second primary cancers involving mouth, pharynx, and larynx. Jow-nal of the Americarr Dental Association 77:580-585. 1968. CEDERLOF. R.. FRIBERG. L.. HRUBEC. Z.. LORICH. U. The Relatiatship uf Smnkin,; and SN/NC' Social CUl'u17u171L's to Ml)rtUlrtC and CGnCPP Mnrhiditl•. A Ten Yeur Fnllow-uh in a Prnbuhllin• Sunrhle nf .5j.0U0 Swedish Subjects Aae /8-69. Purt 11?. Stockholm. Sweden: Karolinska Institute. Department of Environmental Hygiene. 1975. CLARKE. E.A.. MORGAN. R.W.. NEWMAN. A.M. Smoking as a risk factor in cancer of the cervix: Additional evidence from a case-control study. Amc•riccrrr.lournul r#'Epideminlc~, ti- 1 15(1):5y-66, January 1982. CLAUDE. J.C.. FRENTZEL-BEYME. R.R.. KUNZE. E. Occupation and risk of cancer of the lower urinary tract among men. A case-tontrol study. Intc•rncttiorrul Jow-nal qf Ccrncer -i l( 3):371-379, March 15, 1988. CLAVEL. F.. BENHA,b1OU. E.. AUQUIER, A.. TARAYRE, M.. FLANIANT. R. Coffee, alcohol. smoking and cancer of the pancreas: A case-control study. lnternurivnul.lournul u}' Culrc•c•i• 43(1):17-? l, January 15. 1989. CUBILLA. A.L.. FITZGERALD, P.J. Classification of pancreatic cancers (nonendocrine). Mctvo Clinic Prorec•dirr,qs 54:4-19-i58. 1979. , CUZICK. J.. BABIKER. A.G. Pancreatic cancer. alcohol, diabetes me[litus. and Qall-bladder disease. liuernutinnul Jow-nal of Cuncer 43( 3):=)15--1? 1. March 15. 1989. DALING. J.R.. WEISS. N.S.. HISLOP. T.G., MADEN. C.. COATES. R.J.. SHERMAN. K.J.. ASHLEY. R.L.. BEAGRIE. M.. RYAN, J.A., COREY. L. Sexual practices, sexually trans- mitted diseases, and the incidence of anal cancer. New England Jow-nal of Medicine 317(16):973-977, October 15. 1987. DANIELL. H.W. Re: Causes of anal carcinoma. (Letter.) Jucu•nul of tlre Americun Medical Assvc•iutinn 354(3):358. July 19. 1985. DANIELL. H.W. Increased lymph node metatastases at mastectomy for breast cancer associated with host obesity, ci`arette smokinQ. a_e, and larQe tumor size. Cuncer 63(?):429---135. July 15. 1988. ~ ~ FALK. R.T.. PICKLE. L.W., FONTHAM. E.T.. CORREA. P.. FRAUMENI. J.F. JR. Life-style risk factors for pancreatic cancer in Louisiana: A case-control study. Amerirun Journal of Epideminlngr 1?8(2): 3?4-336. August 1988. FARROW, D.C.. DAVIS. S. Risk of pancreatic cancer in relation to medical history and the use of tobacco. alcohol and coffee. lnternutinnul Jnurnul of Cuncer, in press. FRANCO. E.L.. KOWALSKI. L.P.. OLIVEIRA. B.V., CURADO, M.P.. PEREIRA. R.N.. SILVA, M.E.. FAVA, A.S., TORLONI. H. Risk factors for oral cancer in Brazil: A case-control study. Inrernutionul Jow-nal uf Cuncer 43(6):992-1000, June 15, 1989. 180 TIMN 438580

of thrombi, are lower in smokers (Wilhelmsen et al. 1984; Belch et al. 1984), but the levels increase after smoking cessation (Harenberg et al. 1985). Spasm Coronary artery spasm can cause acute ischemia manifested as angina pectoris and may promote thrombus formation at the site of repeated arterial constriction (Folts and Bonebrake 1982). Both chronic and acute cigarette smoking have a demonstrable vasoconstrictor effect on the coronary vasculature (Klein 1984). Compared with never smokers, current smokers have an approximately twentyfold risk of vasospastic angina pectoris (Scholl et al. 1986). Coronary artery spasm has also been identified by angiography after smoking a single cigarette (Maouad et al. 1984). Smoking-induced vasoconstriction has been demonstrated in patients with atherosclerotic coronary artery disease (Martin et al. 1984) that is mediated by an a-adrenergic increase in coronary artery tone (Winniford et al. 1986). In addition, smoking acutely increases platelet and plasma vasopressin (Nussey et al. 1986) as well as the carrier protein of vasopressin and oxytocin (de Lorgeril et al. 1985). In addition to causing acute arterial spasm, cigarette smoking appears to be associated with a reduction in long-term coronary artery diameter independent of atherosclerotic plaque (Fried, Moore, Pearson 1986), although the mechanism for this relationship is unclear. Arrhythmias In some instances, arrhythmias can precipitate MI by reducing cardiac output or increasing myocardial demand. More importantly, arrhythmias are a major complica- tion of infarction. Thus, reducing the threshold forserious arrhythmias tends to inorease the case-fatality rate of MI. Cigarette smoking was found to lower the threshold for ventricular fibrillation in a study of animals (Downey et al. 1977) and was found to be associated with a 21-percent increased prevalence of ventricular premature beats on two-minute electrocardiographic rhythm strips obtained from 10,1 19 men (Hennekens et al. 1980). Smoking-related ventricular arrhythmias may contribute to the occurrence of sudden death and to increased case-fatality ratios during the course of MI. Reduced Blood Oxygen Delivery Cigarette smoking acutely increases myocardial oxygen demand by raising peripheral resistance, blood pressure. and heart rate (Martin et al. 1984; Klein 1984). Concurrently, the capacity of the blood to deliver oxygen is reduced by increased COHb, greater viscosity (Galea and Davidson 1985), and higher coronary vascular resistance. Imbalance between oxygen requirement and delivery as a result of these factors is not likely to be a cause of MI but may contribute to infarction in the presence of significant atherosclerotic narrowing of vessels. Consistent with these mechanisms, low levels of COHb exacerbate myocardial ischemia during graded exercise (Allred et al. 1989), and smoking is associated with more frequent and longer ischemic episodes detected by ambulatory electrocardiographic monitoring among patients with chronic TIMN 438593 195

previous levels among those who returned to smoking (Stamford et al. 1986). Thus, data indicate that smoking reduces the level of HDL-C, a potent protective factor against CHD. In a number of studies, smokers have been found to have higher levels of triglycerides (Freedman et al. 1986; Jacobsen and Thelle 1987; Gomo 1986; Willett et al. 1983); however, the independent relation of trialyceride level with risk of CHD is not clear. Smoking appears to have little, if any, relation with LDL-C level. However, smokers have approximately twice the level of serum malondialdehyde of nonsmokers (Nadiger, Mathew, Sadasivudu 1987); malondialdehyde can alter LDL-C and may promote its incorporation into arterial wall macrophages (Steinberg et al. 1989). In a metabolic study among young men, smokers had a decreased cholesterol net transport from cell membranes into plasma, which could partially explain the accumulation of cholesterol in arterial walls (de Parscau and Fielding 1986). Thrombosis Coronary artery thrombosis, resulting from platelet-fibrin thrombi. is a key element in most cases of MI. Thrombi are visualized in a high percentage of coronary arteries studied angiographically within hours of the onset of infarction (DeWood et al. 1980), and agents that lyse thrombi are effective treatments for MI (Stampfer et al. 1982: Loscalzo and Braunwald 1988). The efficacy of aspirin, an antiplatelet aQent, in preventing MI further supports the role of thrombus formation (Steering Committee of the Physicians' Health Study Research Group 1989). The finding that smoking is associated with history of MI even after controlling for atherosclerosis (Hartz et al. 1981) emphasizes the importance of mechanisms in addition to those that promote atherosclerosis. Platelets play a central role in thrombus formation in addition to releasing growth factors that stimulate the proliferation of smooth muscle cells in arterial intima (Pack- ham and Mustard 1986). Platelets can form microthrombi that become incorporated into the arterial wall, thus contributing to plaque formation and participating in generation of larger platelet-fibrin thrombi that may acutely occlude a coronary artery. Smokin2 ciearettes acutely increases spontaneous platelet aggregation in humans (Davis et al. 1985) and in dogs with coronary artery stenosis (Folts and Bonebrake 1982). Madsen and Dyerberg (1984) observed that smoking 2 high-nicotine ciQarettes substantially reduced bleeding time among healthy young men, although ex vivo tests of platelet agareaability were only minimally inhibited. In this study, smoking low- nicotine cigarettes and inhalation of CO had little effect on bleeding time. Shortened platelet survival, an indirect indicator of activation, was observed in smokers and reverted to normal after 4 weeks of smoking abstinence (Fuster et al. 1981). Studies of smoking and platelet aggregation ex vivo in response to the typical stimuli used in the laboratory, such as adenosine diphosphate (ADP) or thrombin, are incon- sistent. Increased aggregation has been seen with platelets from chronic smokers (Belch et al. 1984) and in blood drawn 10 minutes after smoking I cigarette (Renaud et al. 1985; Renaud et al. 1984); in the latter study, aggregation was associated with blood nicotine levels but not with carboxyhemoglobin (COHb) levels. However, in TIMN 438591 193

stable CHD (Barry et al. 1989). Blood and plasma viscosities among former smokers are lower than those among current smokers and similar to those among never smokers (Ernst and Matrai 1987). In the same study, both blood and plasma viscosity decreased after smoking cessation and were similar to levels of never smokers after 8 weeks. Reduced oxygen delivery to the myocardium may play a role in lowerin; the threshold for ventricular arrhythmias. In addition to influencing the development of CHD, smoking has been hypothesized to have direct toxic effects on the myocardium. Hartz and coworkers (1984) found a nearly threefold increased prevalence of diffuse ventricular hypokinesis among heavy smokers compared with never smokers within a population of patients underaoing diagnostic coronary angiography and ventriculography. ~ Smoking and Development of Peripheral Arterial Disease The extremely strong association between smoking and peripheral artery disease is likely to be mediated largely through the mechanisms that promote atherosclerosis (Criqui et al. 1989). The peripheral vasoconstrictive effects of smoking, mediated by nicotine-stimulated release of catecholamines (US DHHS 1983), are likely to play a further important role (Lusby et al. 1981). Smoking and Development of Cerebrovascular Disease Cerebrovascular disease represents a heterogeneous group of pathologic processes that include infarction due to stenosis and thrombosis (referred to here as ischemic stroke), embolism from the heart, and hemorrhage from medium-sized vessels in the subarachnoid space (subarachnoid hemorrhage)yand from microaneurysms of small penetrating vessels (intracerebral hemorrhage). The association of smoking with ischemic stroke is likely to be mediated larQely throush the mechanisms that promote atherosclerosis and thrombus formation. Associations between smokinQ and extent of cerebral artery atherosclerosis have been observed at autopsy among persons who have died of causes unrelated to CVD (Reed et al. 1988) and among volunteers in a cross-sectional study evaluated by a noninvasive method (Rogers et al. 1983). Smoking was also a strong predictor of the extent and severity of cerebral vessel atherosclerosis in an Italian multicenter study of reversible cerebral ischemic attacks (Passero et al. 1987) and in an investigation of 28 pairs of Finnish twins (Haapanen et al. 1989). The mechanistic basis is unknown for the strons relation between smokinQ and subarachnoid hemorrhage (US DHHS 1989; Shinton and Beevers 1989), which is thought to result most commonly from the rupture of a saccular aneurysm. Although hypertension is associated with this occurrence. chronic smoking is unrelated to sustained elevation in blood pressure. A weak and clinically unimportant inverse relation with hypertension has been seen in several studies (SchoenenberQer 1982: US DHHS 1983), although the association between cigarette smoking and risk of hyper- tension was observed in a large prospective investigation (Witteman et al. 1990). 196 TIMN 438594

TABLE 2.--Continued Relative risks compared with never smokers' Reference Population Followup Ntunber of cases among lirrmer srtwkers Furmcr sntokers Current sntokecs Floderus, Cederlof, Friberg (1989) 10,495 Swedish twins aged 36-75 21 yr for Ci ID tleaths 188 men 10 women I.(1((1.8-t. l) t).6 ((1.4-L.(1) 1.4-1.8 depending on amount smoked Lannerstad, Isacsson, Lindel4 (1979) 703 MaImb men, age 55 5 yr 0 CIID deadts 2.0 I lotme et al. (1980) 14,816 healthy Oslo men, 4.7 NR aged 44--19 Neuerstrom and 2,465 Danislrbus drivers 7.75 yr for Ml and 9 3.2 (0.4-25.6) 5.0 (0.7-36.0) 1uel (1988) CI II) death Comments No reassessment of smoking during folfowup; no data on duration No cases among former smokers; only 2 in never smokers Never and ex-smokers had about 40% of the risk of cigarette smokers NO'PE: CIiD=eoronary he:art disease; pp4=packshlay; ACS CPS-1 and -II=American C'ancer Sixiety Cancer Preveutiun StuJies I and 11; I IIP=heulth insurance plan; MI=tnyoeardial infaretion; NR=not reported. '95% confidence interval shown in parentheses when available. 1, l3reakdowns of relative risk derived frout flgure presmited in paper cited. `Exlrapolated from /igure presemed in paper ciied.

Many studies of large cohorts examined the effects of smoking primarily among men. However, the Nurses Health Study investigators reported on smoking and CHD in a cohort of 121,700 women monitored through biennial questionnaires from 1976 to 1982 (Willett et al. 1987). Women with previously diagnosed CHD were excluded from the analysis. Compared with never smokers, former smokers had a relative risk of 1.5 (95-percent CI, 1.0-2.1). In contrast, current smokers had a substantially elevated relative risk, ranging from 2.1 for smokers of 5 to 14 cigarettes per day to 10.8 for those who smoked 45 cigarettes or more per day. There was no further analysis for the effect of duration of abstinence. The authors suggested that the slight elevation in risk of ex-smokers was due, in part, to resumption of smoking by some fraction of the former smokers. Adjustment for age; obesity; menopausal status; estrogen use; family history of MI; and personal history of diabetes, hypertension, and high cholesterol in a multivariate analysis led to an identical relative risk of 1.5, demonstrating the absence of confounding by these coronary risk factors in this population. In another cohort study, Floderus, Cederlof, and Friberg (1988) monitored 10,945 twins born in Sweden between 1886 and 1925. Smoking behavior was ascertained at baseline in 1961, and the cohort was studied for mortality for 21 years using matched- pair analysis. Among the males, former smokers compared with never smokers had a risk of coronary mortality of 1.0 (95-percent CI, 0.8-1.1). In contrast, current smokers had relative risks ranging from 1.4 to 1.8 depending on amount smoked. There were no data on duration of abstinence at baseline, and there may have been changes in smoking prevalence during the long foilowup that would tend to attenuate the relative risk. ~ In a unique cohort design, Raichlen and coworkers (1986) examined progression of atherosclerosis among 32 men who underwent coronary angiographies at least 2 years apart. Among current smokers, progression of disease was statistically significant and was correlated with pack-years smoked during the interval. Among past smokers, the degree of progression of atherosclerosis was far less than among current smokers: it was not statistically different from lack of progression. ~ Several other cohort studies have reported on the relation of smoking cessation with risk of CHD; however, the number of subjects was generally too small to contribute substantially to knowledge in this area (Table 2). Intervention Trials In several clinical trials, an attempt has been made to evaluate the effect of altering risk factors for CHD, including smoking (Chapter 3). Most of the trials including smoking cessation have also incorporated interventions for other CHD risk factors makin- it difficult to assess the independent effect of quitting. Nonetheless, these data have extended the understanding of the effects of smokine cessation on CHD risk. Assessing self-report of smoking cessation or decrease in~cisarette consumption is another potential difficulty. There may be a tendency for subjects in a trial to seek approval and avoid negative feedback by reporting less cigarette use than is actually the case (Chapter 2). Such a tendency would have the effect of misclassification and would yield an underestimate of the benefits of cessation (Table 4). 224 TIMN 438622

<10/day 10-201day <10/day 10-20/day CIGARETI'ES SMOKED ]/ Ex-Smokers 21-39/day 21-39/day ?40/day _40/day El Current Smokers FIGURE 4.-Mortalitv ratios for all cardiovascular diseases and CHD. b_v daily cigarette consumption, US Veterans Study, 195-4-69 NOTE: Ex-smokers includes only former cisarette smokers who stopped smoking for reasons other than physician's orders. ~ SOURCE: Rogot and Murray (1980). TIl'!IN 438617 219

was 1.1 (95-percent CI, 0.5-2.2). Current smokers had significantly elevated relative risks ranging from 2.0 to 3.0, depending on the amount smoked. In a later report from the Framingham Study based on 18 years of followup biennial examinations, Gordon, Kannel, and McGee (1974) assessed the effects of smoking cessation. In this analysis, anyone who smoked for 1 year or more during the most recent 2-year interval between examinations was considered a current smoker. Ap- proximately 20 percent of men who reported that they had quit smoking at entry into the study resumed smoking; about half of those smoked very little or only intermittently after resumption. Compared with current smokers, former smokers had a 30-percent reduction in fatal and nonfatal CHD (excluding angina); the relative risk among current smokers compared with that among never smokers was 1.3. Other coronary risk factors were examined in detail; there were no significant differences between persistent smokers and those who quit, but those who quit were more likely to be ill. Hence, it would be expected that adjustment for confounding would have revealed even greater benefit from cessation. The benefit of quitting seemed more marked in younger men. However, there were only 24 cases of CHD among the quitters so that a detailed analysis could not be performed. ~ The Western Collaborative Group Study monitored a cohort of 3.524 men for an average of 8.5 years for CHD incidence (Rosenman et al. 1975). Information collected at baseline amon= men ased 39 to 49 indicated that former smokers had a relative risk of 1.9 compared with that of never smokers, 20 percent lower than among current smokers. For men aged 50 to 59. former smokers had a relative risk of 1.1 compared with never smokers, 40 percent less than among current smokers. This effect of cessation was slightly greater than that observed after 4.5 years of followup (Jenkins. Rosenman, Zyzanski 1968). The difference between the age groups could be a true effect or may reflect different levels of misclassification: it is possible that a greater proportion of the quitters in the younger group than in the older group resumed smoking. In 1963. a prospective study of smoking and mortality was conducted in Sweden by sending questionnaires to a probability sample of men acred•18 to 69 (Cederlof et al. 1975). A total of 51.911 respondents provided some information; a subsample of 11,739 were sent followup questionnaires in 1969. In that interval. 12 percent of the former smokers had resumed ciaarette smokina, and an additional 8 percent initiated pipe or cigar smoking. The men were monitored for 10 years for mortality and cancer morbidity. Men who quit within the past 9 years had a significantly elevated relative risk (RR=1.5) that was nearly as high as the relative risk for current smokers (RR= 1.7). In contrast. those with a lonser duration of abstinence had a relative risk of 1.0. Men with diseases at baseline were not excluded, so it is likely that the benefits of recent cessation are obscured by the inclusion of men with disease-induced quitting. The Whitehall Civil Servants Study (Rose et al. 1977; Fuller et al. 1983) is another important source of data on risk factors for CHD. Between 1967 and 1969. a total of 18,403 male civil servants aged 40 to 64 were examined. In the 19-year followup. the age-adjusted CHD mortality rate among 17,051 persons with normall blood sugar was 50 percent lower for quitters than for current smokers. When compared with never smokers, the relative risk for former smokers among normoglycemics was 1.3. Among the 999 men with slucose intolerance (but not diabetes), the risk for former smokers TIMN 438619 221

TABLE 4.-Intervention trials of smoking cessation and CIiD risk Reference Population Intervention Outcome l lughes et al. ( I981); MRFIT: 12,866 healthy US Diet, reduction in weight, CI ID deaths MRFIT Research men aged 35-57 at high CI ID hypertension, and smoking Group(1982,f986); risk Grimm (1986); Oc):ene et al. MRFIT: 7,663 participant Diet, reduction in weight, CI ID deaths (1990) snwkers at entry hypertension, and smoking MRFIT: 6,943 participant smokers at entry Diet, reduction in weight. hypertension, and smoking Iljermann et al. Oslo study: 1,232 healthy Diet and smoking (1981) Oslo men aged 40-49 at high CI11) risk Kornitzer et al. 19,409 male Belgian factory Antismoking, (1983) workers, aged 40-59 hypertension control ClID deaths Fatal and nonfatal MI Fatal and nonfatal MI Cases among former smokers 15 33 12 16 169 Overall effect of intervention 7% decline in intervention group 47% decline in intervention group 24.5% reduction in intervention group Effect of smoking cessation (nonrandomized) 44% reduction compared with persistent smokers Quitters had 42% reduction (16-60%) comparing quitters at first annual exam to smokers at that time Quitters had 65% reduction (37-80%) comparing 3-yr persistent quitters with persistent smokers Smoking cessation accounted for about 25% of the difference between the groups No specific analysis conducted for effect of smoking cessation

questionnaires about their smoking behavior, and in 1957, an additional 49,226 responded. Those who stopped smoking on a physician's orders were excluded from the analysis. Mortality in this cohort was monitored, and death certificates were obtained to assess cause of death. Smoking status after the baseline questionnaire was not ascertained. After 16 years of followup, quitters at enrollment when compared with never smokers had relative risks of 1.15 for all cardiovascular mortality and 1.16 for CHD death specifically (Rogot and Murray 1980). In contrast, men who were current smokers at baseline had relative risks of 1.58 for these two categories. Among past smokers, risk of death due to CVD increased with higher previous usual daily cigarette consumption. The relative risks among past smokers, compared with never smokers, ranged from 1.02 for less than 10 cigarettes per day to 1.34 for 40 cigarettes or more per day. This gradient was more pronounced among current smokers (Figure 4). A gradient was also apparent for decreasing risk with increasing duration of smoking abstinence. For both cardiovascular and coronary mortality, there was a moderate decrease in risk with short duration of abstinence and a smaller, but consistent decline in risk with longer periods of abstinence (Figure 5). After 20 years or more of abstinence, the relative risk of CVD was !.04, and for coronary death, the risk was 1.05. The major strenQth of the U.S. Veterans Study is the large numbers, with 21.413 deaths from CVD amonQ smokers and 9,027 among former smokers. The long followup period without reclassification of smokina status is a limitation, which will tend to lead to an underestimate of the effect of sustained smokinQ and an underestimate of the benefits of quitting (Chapter 2). This source of potential bias may not have markedly distorted the estimates in this study: in the followup of this cohort (Rogot and Murray 1980). the relative risk for cardiovascular mortality associated with current smokinc, at enrollment was 1.62 at 8.5 years and 1.58 at 16 ye-ars; forcoronary disease, the relative risk was 1.61 at 8.5 years and 1.58 at 16 years. Thus. the impact of misclassification of current smokers who quit (and therefore lowered their risk) as persistent smokers appears to be slight. A similar comparison of the relative risks among former smokers is less informative in assessing the impact of misclassification. Most quitters who resume smoking do so within 2 years after cessation. Therefore, misclassitication of ex-smokers between 8.5 and 16/years of cessation is likely to be small. For both cardiovascular mortality and coronarv mortality, the relative risks among ex-smokers declined slightly from 1.21 at 8.5 years of followup to 1.15 and 1.16 at 16 years of followup. This is consistent with the inverse relation between duration of smokincr cessation and mortality ratio. AmonQ current smokers in the U.S. Veterans Study, the relative risks of coronary disease were slightly hiQher after 8.5 years of followup (relative risk (RR)= 1.95 for >20 cig/day) than after 2.5 years of foilowup (RR=1.75) (Dorn 1959). As expected. those who stopped smoking on a physician's orders were at hi;her risk of death regardless of their smoking status. ~ An early report of combined data from the Framingham and Albany Heart Studies (Doyle et al. 1962) included 4.120 men free from coronary disease at entry into the study. The Framingham Study data were based on 6 years of followup and the Albany Heart Study data on 8 years of followup. Among the 411 former smokers in the combined cohort, the relative risk of MI (age-adjusted) was 0.9 compared with never 218 TIMN 438616

TABLE 5.-Continued Reduction in risk Cases among compared with i Reference 1'opulation Followup ti>rmer snwkers persistent smokers" Comments Johansson et al. 156 Giiteborg women aged <_65, 5 yr 12 deaths 60% (80-20) Quitters had worse baseline (1985) smokers at time of first Ml prognosis; differences between groups were apparent early and increased with time Perkins and Dick 119 UK patients who smoked at 5 yr 9 deaths 60% (1985) first MI Vlietstra et al. 11,605 patients in CASS who smoked 5 yr By risk®luartile: Total mortality: Quitters had worse baseline (1986) at tinte CI ID was diagnosed by prognosis; exclusion of (best) I: 13 40% those with mixed smoking angiography 2: 21 40% behavior and close followup 3: 44 50% reduced likelihood of (worst) 4: 156 20% misclassification of overall: 234 40% (50-20) exposure; also, hospitalization for MI was substantially reduced in former smokers ~ I lermanson et al. 3,045 CASS patients with Cl ID aged 5.3 yr lor 35-54 yr: NR 40% (50-30) Reanalysis of a subset of ~ .(1988) 35-54 MI or death patients analyzed by ~ Vlietstra (1986) W 1,893 CASS patients with Cl ID aged 55-59 yr: 99 30%(50-20) ~ ?55 60-64 yr: 92 30% (50-10) 65-69 yr: 48 30% (60-)) >70 yr: 21) 70% (8O-30) tJ W W

Current c%arette smokers Ex-cigprette smokers 2.0r- All cardiovascular .. diseases (330-334, 400-468) 0 . 1.0 Coronary heart disease (420) 0 2.0 r FIGURE 5.--Vlortality ratio for current and former cigarette smokers by years of smoking cessation. US Veterans Study. 1954-69 NOTE: Ex-smokers includes onty former ci~=arette smokers who stopped smokins for reasons other than physician's orders. ~ SOURCE: Roqot and Murray (1980). smokers. 60 percent lower than among current smokers. A more detailed analysis was not possible because only 10 cases occurred among former smokers. In a second report using the combined data from the Framinaham Study and the Albany cohort (Doyle et al. 1964). the relative risk for former versus never smokers _ 220 TIIVIN 438618

TABLE 5.-Continued Reference Population Followup . Von der Lippe and 1,330 participants in the 17mo Lund-Johansen Norwegian timolol trial who (1982) smoked at time of MI Rtannevik, 1,330 participants in the Norwegian 17 mo Gunderson, timolol trial who smoked at time of MI Abrahamsen (1 985) Shapiro, llowat, 142 MI survivors aged <45 <_I0 yr Sinbh(1982) Aberg et al. (1983) 983 Glitelxrrg male smokers <_ 10.5 yr at time of MI Cases among former smokers 31 deaths in those who stopped smoking before entering the trial 37 deaths in those who stopped in the first months of the trial 44 recurrent noiifatal MI N lt 104 recurrent nonfatal MI; 80 CI ID deaths Daly et al. (1983) 374 Dublin men, smokers at time of Mean 7.4 yr, 80 deaths MI diagnosis or angina <_ 13 yr Reduction in risk compared with persistent smokers" None 10% 33% reduction; 8% in quitters, 12% in persistent smokers 80% (former and never smokers vs. persistent smokers) 30%; difference between groups increased with time 60% overall; 40% first 6 yr; 80% 7-13 yr Comments Study not designed to examine effects of smoking cessation; no details provided on possible confounding Former and never smokers considered together, not separately 30% quitters had worse predicted prognosis at baseline; no further assessment of smoking beyond 3 mo after initial MI Followup began 2 yr after M1, when smoking status was assessed

three annual examinations with persistent smokers. In this analysis, which would be affected to a lesser extent by misclassification, former smokers had a 65-percent reduction in risk compared with persistent smokers (95-percent CI, 37-80). A trial using a somewhat similar design was conducted in Oslo, Norway (Hjermann et al. 1981; Hjermann, Holme, Leren 1986). Males aged 40 to 49 were screened for coronary risk, and normotensive men at high risk of CHD due to elevated serum cholesterol, smoking, and other risk factors were identified. The participants had no clinical CHD at the time of randomization to the intervention or control group (N=604 and N=628, respectively). The intervention consisted of advice and instruction on altering diet and reducing smoking. Participants were examined at least annually during the 5 years of followup. * After 5 years, fatal and nonfatal CHD was reduced in the interventiongroup by 47 percent. There was greater success in reducing cholesterol in this trial than in inducing smoking cessation. The mean serum cholesterol was ap- proximately 13 percent lower in the intervention group than among the controls. However, only 25 percent of the smokers in the intervention group and 17 percent in the control group quit entirely, although many reduced the amount smoked. There was an inverse relation between CHD incidence and percentage change in tobacco con- sumption, but this did not attain statistical significance. The authors calculated that approximately 25 percent of the difference in CHD incidence between the two groups was attributable to differences in smoking. A second report (Hjermann, Holme, Leren et al. 1986) included followup through 102 months. Statistically significant reductions among the intervention group com- pared with the control group were seen for fatal coronary events (reduced 59 percent), total coronary events (reduced 44 percent), and•total cardiovascular events (reduced 61 percent). The World Health Organization European Collaborative Trial in the multifactorial prevention of CHD was conducted at several sites in Europe. Pooled results were reported from centers in the United Kingdom. Belgium, Italy, and Poland (WHO European Collaborative Group 1983): separate reports have also been published from centers in the United Kingdom (Rose. Tunstall-Pedoe, Heller 1983) and Belgium (Komitzer et al. 1983). A total of 66 factories involving 49,781 men were randomized to a multifactorial risk factor reduction program or to the control group. The reduction of levels of risk factors varied considerably among the centers. Overall. the reduction in risk factor levels was modest, and there was no significant decline in CHD endpoints in the intervention aroup. The effect on CHD was broadly correlated with changes in risk factors. There was no specific analysis on the impact of smoking cessation. The Belgian center was the largest in the European Collaborative Trial. Fifteen pairs of factories were randomly allocated to the intervention or control groups, which included 19,409 men aged 40 to 59 years. The intervention included advice about smoking cessation and reduction of hypertension and elevated cholesterol. Subjects were screened as part of the trial, but referred to theirown physicians fortherapy. After 6 years, there was a 24.5-percent reduction in fatal and nonfatal CHD in the intervention group compared with the control group (p=0.03) (Kornitzer et al. 1983). The rates in the intervention and control groups continued to diverQe throuQhout the followup 228 TIMN 438626

other studies. ex vivo platelet aggregation was not related to cigarette smoking (Pittilo et al. 1984: Dotevall et al. 1987; de Lorgeril et al. 1985; Madsen and Dyerberg 1984). In one large study, aggregation in response to ADP stimulation was actually somewhat greater in nonsmokers (Meade et al. 1985). Studies of the effect of smoking on platelet production of thromboxane. which mediates the aggregatory effect, have also been inconsistent. In some studies, smoking was found to acutely increase thromboxane blood levels, which reflect the capacity to produce thromboxane in response to stimula- tion, and urinary metabolites, which reflect the normal steady-state production (Toivanen, Ylikorkala, Viinikka 1986; Marasini et al. 1986; Fischer et al. 1986). However, serum thromboxane B2 levels were found to be similar among chronic smokers compared with nonsmokers in another study (Dotevall et al. 1987). The serious limitations of ex vivo aggregability measurements in the evaluation of in vivo platelet activity have been noted (Fitzgeraid, Oates, Nowak 1988). These researchers measured urinary excretion of a thromboxane metabolite and found elevated levels in chronic smokers that were reduced to the level of nonsmokers after aspirin administra- tion. suggesting a platelet origin of the excess excretion (Nowak et al. 1987). The lack of a consistent relation between smoking and ex vivo tests of platelet aggregability despite the demonstration that platelets of smokers adhere more readily to endothelium has led to the suggestion that smoking inhibits the production in arterial walls of prostacyclin. an inhibitor of platelet aggregation (Madsen and Dyerberg 1984). Reinders and coworkers (1986) demonstrated that the production of prostacyclin by cultured human endothelial cells is impaired by incubation with cigarette smoke condensate. Pittilo and colleasues (1982) also found that smoking reduces endothelial cell synthesis of prostacyclin in rats. Thus, in vivo smoking-related effects on platelet function may be mediated in part by an interaction with endothelium. Fibrinogen levels have been found to be elevated among smokers in numerous cross-sectional studies (Meade et al. 1986; Kannel, D'Agostino. Belanger 1987; Wil- helmsen et al. 1984; Dotevall et al. 1987: Belch et al. 1984; Balleisen et al. 1985). Fibrinogen levels, in turn, are stronQly related to risk of CHD and stroke (Meade et al. 1986; Kannel. D'Agostino, Belanger 1987; Wilhelmsen et al. 1984). Smoking cessa- tion resulted in a decrease in fibrinogen levels after 4 weeks among 9 female smokers (Harenber2 et al. 1985) and after 8 weeks among 14 male smokers (Ernst and Matrai 1987). In the latter study, the levels after 8 weeks were similar to those among never smokers. When fibrinogen was remeasured after 5 years, values had decreased to the levels of never smokers amons men who had stopped smoking and had increased among those who started or resumed smoking (Meade. Imeson, Stirling 1987). In multivariate analyses of data from the Framingham Study (Kannel, D'Agostino. Belanger 1987) and Northwick Park Study (Meade et al. 1986) that both included cigarette smoking as well as fibrinogen levels, fibrinogen retained a clear independent association with risk of CHD, whereas the effect of smoking was substantially reduced after the inclusion of fibrinogen in the model. This analysis suggests that elevated fibrinogen levels may mediate a quantitatively important part of the effect of smoking on CHD risk. Other clotting abnormalities, such as increased plasma viscosity and reduced red cell deformability. that tend to promote thrombus formation have also been observed in smokers (Belch et al. 1984). In addition. levels of plasminogen, which promotes lysis 194 TIMN 438592

be quite valuable in assessing the time course for the decline in risk. However, the lack of detailed data on fatal cases is a potential limitation of the case-control approach. In a case-control study of women in the Nurses Health Study cohort, Willett and coworkers (1981) identified 263 women who reported a nonfatal MI on the baseline Nurses Health Study questionnaire in 1976 when they were 30 to 55 years of age. Their smoking histories were compared with randomly selected controls corresponding in age with a case-control ratio of 1:20. Women who were former smokers did not experience increased risk of MI, with a relative risk compared with never smokers of 1.0 (95-percent confidence interval (CI), 0.7-1.6). In contrast, current smokers had a significantly elevated threefold higher risk of MI. When duration of abstinence was assessed, it appeared that those who quit either I to 4 or 5 to 9 years earlier had a nonsignificantly elevated risk of 1.5, and those who quit 10 years or more earlier had a relative risk of 0.6. Because there were only 29 cases among former smokers, the estimates for risk by duration of abstinence are not precise. Rosenberg. Kaufman, Helmrich, and Shapiro (1985) specifically analyzed the impact of smoking cessation on risk of first MI among 4,648 men less than 55 years of age, using a hospital-based case-control design. Men with known preexisting heart disease were excluded. The 2,775 controls were mostly persons with fracture or sprain, disk disorders, and gastrointestinal disorders thousht not to be related to cigarette smoking. There were 1.873 cases and 2,775 controls. For current smokers (smoked within the past year), the age-adjusted relative risk was 2.9 (95-percent CI, 2.4-3.4) and for past smokers overall, it was 1.1 (95-percent CI, 0.9-1.4). The relative risk for those who had not.smoked for 12 to 23 months was 2.0 (95-percent CI, 1.1-3.8). For those with longer durations of abstinence, the relative risk was 1.1 (95-percent CI, 0.9-1.4) (Figure 2). The risk was increased for those smoking more cigarettes per day among current smokers as well as recent quitters. For longer durations of abstinence, the amount previously smoked appeared to have little impact. These investigators also examined the effect of quitting within categories of other risk factors; in general, there were no marked differences other than for diabetics among whom the benefits of cessation appeared to be greater. The same group of investigators (Rosenberg, Kaufman. Helmrich, Shapiro 1985) addressed the possibility that continuing smokers and former smokers may differ in their underlying risk of heart disease. They found that those who quit had a slightly higher risk profile. Hence, the benefit of cessation in this study cannot be attributed to overall better health among those who quit. Rosenberg and associates (1985) also conducted a hospital-based case-control study of first nonfatal MI amonQ women less than 50 years of age (Rosenberg, Kaufman, Helmrich, Miller et al. 1985). Women who smoked in the year before admission were classified as current smokers. Participants consisted of 555 cases and 1,864 controls who were hospitalized for trauma, orthopedic disorders, and other conditions thought to be unrelated to smokina. Current smokers had relative risks increasinQ from 1.4 to 7.0, depending on the number of cigarettes smoked per day. In contrast, former smokers (at least 1 year of abstinence) had the same risk as never smokers, with a relative risk of 1.0 (95-percent CI, 0.7-1.6). In a recent report, Rosenberg, Palmer, and Shapiro (1990) further examined the decline in risk of MI amonQ women who stopped smoking. Cases included 910 women 203 TIMN 438601

period. No specific analysis was conducted to assess the independent effect of smoking cessation on risk of CHD. The multifactor primary prevention trial in Goteborg, Sweden focused on reduction of hypertension, elevated serum cholesterol, and smoking (Wilhelmsen et al. 1986). A random sample of 10,004 men aged 45 to 55 years was included in the intervention group. and 2 other random samples of the same size were identified as controls. Of those invited to participate in the intervention group, 7,495 attended the first screening examination. At the outset, within the intervention and control groups combined, 20.6 percent were former smokers. After 4 years, the proportion of former smokers in- creased to 27.7 percent, and after 10 years to 39.4 percent in the intervention group. In the control group, the percentage of former smokers also increased-to 22.3 percent at 4 years and to 36.1 percent at 10 years. The differences achieved for other risk factors between the intervention and control groups were also quite small. After 10 years, there were virtually no differences in fatal and nonfatal outcomes between the groups. The center in the United Kingdom was also large (Rose, Tunstall-Pedoe, Heller 1983), with 12 pairs of factories and 18.2 10 men aged 40 to 59 years. There were only very modest changes in risk factors other than cigarette smoking. The reported number of cigarettes smoked per day in the intervention group decreased by 16 percent, but the proportion of current cigarette smokers decreased by only 4 percent. Rose and Hamil- ton (1978) stated that whereas self-report of cessation is likely to be reasonably accurate, reported decreases in smoking are probably exaggerated. With such small net changes in risk factors, it is not surprising that there was virtually no difference in the rate of CHD between the two groups. Only one trial has attempted to assess the effect of advice for smoking cessation without intervening for other risk factors simultaneously. In theory, trials of this design can provide the clearest indication of the effect of such advice in the absence of other effects. Participants were selected from a cohort of 16,016 from the Whitehall Civil Servants Study (Fuller et al. 1983). From this aroup, 1,445 high-risk male smokers aaed 40 to 59 were randomized to a normal care group or the intervention group that received antismoking advice. At year one, 51 percent of the intervention group reported that they were not smoking, and at year three. 36 percent reported the same. In the normal care group, the corresponding percentages were 10 and 14 percent. A third of the quitters reported smoking cigars or a pipe. It is important to note that the question- naire response rate at 3 years was 64 percent in the intervention group and 70 percent in the normal care group (Rose and Hamilton 1978). The 9-year response rate was 83 percent. At that point, 55 percent of responders in the intervention group reported quitting, as did 41 percent in the normal care group. Despite the similarity of smoking prevalence of the two groups, at 10 years CHD mortality decreased by 18 percent in the intervention group. This difference did not attain statistical significance (95-percent CI, -43 to +18 percent) (Rose et al. 1982). Smoking Cessation and CHD Risk Among Persons With Diagnosed CHD Studies examining smoking cessation and CHD risk among persons with diagnosed CHD may be less prone to some of the methodologic pitfalls discussed in Chapter 2. TIMN 438627 229

was 30 percent lower than that for current smokers. Overall, the 224. diabetic men experienced a very high risk of CHD; among this group the risk for former smokers was 30 percent higher than for current smokers (based on 10 cases among the current smokers). These data are generally consistent with other studies in the overall findings, but suggest that smoking cessation may not have the same benefit for diabetics as for the general population; however, this finding is based on small numbers, and the severity of diabetes was not considered in the analysis. This study did not provide any information on the time course of the decline in risk after cessation. It is also likely that during the long followup period, a substantial percentage of current smokers quit smoking. The effect of differences in coronary risk factors other than smoking was examined in quitters and persistent smokers by Friedman and colleagues (1979). As expected, there were a number of differences between quitters and persistent smokers when they were studied at a time in which individuals in both groups were smoking. A followup analysis of this same population was conducted to assess the impact of quittine on risk of CHD and to evaluate the effect of differences between these groups that might alter CHD risk (Friedman et al. 1981). Smoking was assessed by questionnaire at ap- proximately annual multiphasic health checkups given at the Kaiser-Permanente Medi- cal Centers in San Francisco and Oakland, CA. There were 9,394 persistent smokers, 2.856 persistent quitters (those who denied smokina at 2 sessions after an examination when they were currently smoking), and 12,697~never smokers. The cohort was monitored for ara average of 4 years for a total followup of 188,436 person-years. The age-, sex-, and race-adjusted death rates (perr thousand person-years) associated with CHD were 2.6 among smokers, 1.4 among quitters, and 1.6 amonQ never smokers. After adjustment for baseline differences, quitters had a risk of fatal CHD that was 55 percent lower (95-percent CI, 74-22) compared with persistent smokers. By excluding individuals with frank coronary disease at baseline, a slightly higher benefit for quittinQ was demonstrated. Further adjustment for measur,es~of smoking intensity slightly attenuated the reduction in risk to 47 percent, suggesting that only a small part of the apparent benefit of quitting is attributable to the fact that quitters were less intense smokers at initiation of smokinQ. Only the number of cigarettes smoked had any measurable impact: depth of inhalation and duration of smoking had no effect. Except for women durinQ the first half of this century, most smokers beein to smoke durina adolescence; thus, duration is very highly correlated with age in most populations. These findings generally confirmed previous results from the same study (Friedman, Dales, Ury 1979). The Seven Countries Study (Keys 1980) provided a valuable resource for analysis of risk factors for CHD. A total of 16 cohorts of inen, aged 40 to 59, livinQ in 7 countries, were examined and monitored for 10 years for CHD incidence. The cohorts were assembled between 1958 and 1964, and consisted of 12.096 men free from CVD. In each grouping of cohorts, former smokers had a lower risk of CHD than did current smokers. However, oniy about 28 cases of CHD death among former smokers were reported: therefore, no detailed analysis was possible. Data on the health effects of smoking cessation are also available from the Health Insurance Plan of Greater New York. The incidence of MI was ascertained over a ~~~ -_ __- TIMN 438620

References ADAMI, H.O., LUND, E., BERGSTROM, R., MEIRIK, 0. Cigarette smoking, alcohol consumption and risk of breast cancer in young women. British Journal nfCanrer 58(6):832- 837. December 1988. AMERICAN CANCER SOCIETY. Unpublished tabulations provided by L. Garfinkel from the Cancer Prevention Study II. August 1989. ARCHIMBAUD, E.. MAUPAS, J., LECLUZE-PALAZZOLO, C., FIERE. D., VIALA, J.J. Influence of cigarette smoking on the presentation and course of chronic myelogenous leukemia. Cancer 63(10):2060-2065, May 15, 1989. AUGUSTINE, A., HEBERT, J.R., KABAT, G.C.. WYNDER. E.L. Bladder cancer in relation to cigarette smoking. Cancer Research 48:4405-4408, August 1, 1988. AUSTIN. H.. COLE. P. Cigarette smoking and leukemia. Jntrrnal of Chronic Diseases 39(6):417-421, 1986. BARON. J.A. Smoking and estrogen-related disease. Ameriran Jorunal of Epidemiology 119(1):9-22, January 1984. BARON. J.A.. BYERS, T.. GREENBERG, E.R., CUMMINGS. K.M., SWANSON, M. Cigarette smoking in women with cancers of the breast and reproductive organs. Journal of the National Cancer Institute 77(3):677-680. September 1986. BARTON. S.E.. MADDOX. P.H.. JENKINS. D.. EDWARDS, R.. CUZICK. J., SINGER. A. Effect of cigarette smoking on cervical epithelial immunity: A mechanism for neoplastic change? Lancet 2(8612):652-654, September 17. 1988. BERG. J.W.. SCHOTTENFELD, D.. RITTER. F. Incidence of multiple primary cancers. III. Cancers of the respiratory and upper digestive system as multiple primary cancers. Jow•nal of the National Cancer Institute 44:263-274. 1970. BERGGREN. G., SJOSTEDT, S. Preinvasive carcinoma of the cervix uteri and smoking. Acta Obstetricia et Gvnecologica Scandinavica 62(6):593-598, 1983. BLOT, W.J.. MCLAUGHLIN, J.K.. WINN. D.;vt., AUSTIN, D.F., GREENBERG. R.S.. PRESTON-NIARTIN; S., BERNSTEIN, L.. SCHOENBERG, J.B.. STEMHAGEN. A.. FRAUMENI. J.F. JR. Smoking and drinking in relation to oral pharyngeal cancer. Cancer Research 48:3282-3287, June 1. 1988. BRINTON, L.A., SCHAIRER, C.. HAENSZEL. W.. STOLLFY, P., LEHMAN. H.F., LEV INE. R., SAVITZ. D.A. Cigarette smoking and invasive cervical cancer. Jow•nal of the American Mediral Association 255(23):3265-3269. 1986. BRINTON. L.A., SCHAIRER. C., STANFORD. J.L.. HOOVER. R.N. Cigarette smoking and breast cancer. American Journal of Epidemiology 123(4):614-622, April 1986. BRISSON. J., ROY, M., FORTIER. M., BOUCHARD, C., MEISELS, A. Condyloma and intraepithelial neoplasia of the uterine cervix: A case-control study. American Journal of Epideminlo,t r 128(2):337-342, August 1988. BROCK, K.E., BERRY, G.. MOCK, P.A.. MACLENNAN. R.. TRUSWELL, A.S., BRINTON. L.A. Nutrients in diet and plasma and risk of in situ cervical cancer. Journal of the National Canrerlnstitute 80(8):580-585. June 15, 1988. BROWNSON, R.C.. BLACKWELL. C.W., PEARSON. D.K.. REYNOLDS. R.D.. RICHENS. J.W., PARERMASTER. B.W. Risk of breast cancer in relation to cigarette smoking. Archives of Internal Medicine 148( I):140-144, January 1988. BROWNSON, R.C., CHANG, J.C.. DAVIS. J.R. Occupation, smoking, and alcohol in the epidemiology of bladder cancer. American Journal of Public Health 77(10):1298-1300, October 1987. B URCH, J.D., ROHAN, T.E., HOWE, G.R., RISCH. H.A., HILL, G.B., STEELE, R., MILLER, A.B. Risk of bladder cancer by source and type of tobacco exposure: A case-control study. International Journal of Cancer 44:622-628, 1989. 179 TIMN 438579

TABLE 8.-Conlinued Relative risk compared with never smokers' Reference Population Followup Cases among fbrmersmokers Outcome Former Current smokers smokers ACS (unpublished ACS CPS-11(50-State Study)` 4 yr (1982-86) NR Mortality due to Men <21 cigkiay tabulations) , cerebrovascular Quit <1 yr 3.94 2.43 disease 1-2 yr 1.11 3-5 yr 1.55 6-10 yr 1.64 11-15 yr 0.62 L>16yr 0.72 Men ~21 cig/day Quit <1 yr 0.37 2.07 1-2 yr 1.43 3-5 yr 1.39 6-10 yr 2.27 I 1-15 yr 2.34 >16yr 1.92 Women <20 cig/day Quit <1 yr NR 1.77 I-2 yr 1.92 3-5 yr 0.79 6-10 yr 0.59 1 I-15 yr 1.23 ?16yr 0.93

100 90 0 0 0 0 1 2 3 4 5 TIME (YR) 0 Quitters o Continuers FIGURE 6.-Effect of smoking cessation on survival among men with documented coroqary atherosclerosis; pooled survival among quitters (0) (N=1,490) and continuers (A) (N=2,675) SOURCE: Vlietstra et al. (1986). . strata, this difference was of borderline sianificance in a life table analysis (p=0.076). After exclusion of crossovers (14 smokers quit_6 months after the arrest, and 2 quitters resumed smokinQ), the benefit of cessation was slightly more pronounced (p=0.048). Analysis of data from a trial of practolol also provided information on the effects of smoking cessation after MI (Green 1987). There were 855 never smokers, 1,344 persistent smokers, and 851 individuals who quit smoking after the entry MI. Those who stopped smoking had a worse outcome initially than persistent smokers, and the benefit from cessation did not appear until 2 years after the event. When events in the first 6 weeks after the index MI were excluded, the benefits of cessation appeared at about 18 months. By 24 months, those who stopped had a 30-percent CHD risk reduction. As in other studies, former smokers when compared with continuing smokers tended to have more severe MI, with significantly more pulmonary congestion noted when x-rayed and significantly greater occurrence of faster dysrhythmia. This supports the view that those with a worse MI are more likely to quit, and it explains why quitters in the study had a worse initial outcome. In a trial of rehabilitation after MI, 147 patients in a Swedish hospital were routinely invited to participate in a rehabilitation program; 158 patients in a comparable hospital were not (Hedback and Perk 1987). The cardiovascular experience in the intervention 238 TIMN 438636

TABLE 8.-Continued Relative risk compared with never smokers° Reference Popul:uion Followup Cases among former smokers Outcome Former Current smokers smokers Wolf et al. (1988) Framingham Study: 4,255 26 yr N/A Stroke and transient Risk significantly Men: men and women ischemic attack lower than that of 1.42 stroke current smokers 1.56 brain infarction Women: olditz et al. (1988) urses I lealth Study: 118,539 yi 5 ubarachnoid 1.61 stroke 1.86 brain infarction 3.0 (1.3-6.6) 1-14 cig/day: 4.3 US women aged 30-55 hemorrhage 15-24 cig/day: 5.1 Thromboembolic _25 cig/day: 10.3 1.3 (0.7-6.6) 1-14 cig/day: 1.8 stroke 15-24 cig/day: 3.2 Total stroke _25 cig/day: 3.1 1.5 (1.1-2.2) 1-14 cig/day: 2.5 15-24 cig/day: 2.9 US DI II IS (1989) ACS CPS-1(25-State Study) 6 yr (1959-65) NR >25 cig/day: 3.8 Mend 35-64 yr: 1.02 (0.83-1.25)" 1.79 (1.55-2.08)" _>65 yr: 0.93 (0.80-I .08)a 1.15 (1.02-I.30)"

TA13Lr 7.-Continued Relative risk as compared with never smokers° Reference Source and case-control numbers Outcome Strokes among former smokers Former smokers Current smokers Bell and Symon England: 106 men, 1,628 Subarachnoid hemorrhage NR Men 1.92 3.89 (1979) women; general UK population 1965 Women 2.52 3.72 Cullaborative Group US: 430 cases (15-44 yr); Thrombosis 21 1.14 1.18 for the Study of Stroke in Young Women (1975) 429 hospital controls; 451 neighborhood controls I leniorrhage 26 1.76 3.27 Donnan ci al. (1989) 422 consecutive cases; 422 Cerebral ischemia 145 2.0(1.3-3.1) 3.7 community controls Quit <2 yr 3.2 2-5 yr 3.1 5-10 yr 2.1 >10 yr 1.7 NbTE: NR=not repueed. "95°/r, conliJence interval shown in puremhezes when available. 1'ReLuive risk calctduted frum data presented in original paper.

100 -1 aD ` 90~ , .6- 0 ~ .~ 70 + 0 5 10 15 20 5 10 15 20 Years of Follow-up FI(;Ultl; 8.-Survival free of stroke in cigarette smokers (dotted line), never smokers (solid line), and former smokers (dashed line), aged 60, using Cox proportional hazard regression model, among men and women SOURCE: Wolf et al. (1988).

Cohort Studies Data from prospective cohort studies are summarized in Table 2. The British Physicians Study of Doll and Hill (1954. 1956) was one of the important early studies that established the link between smoking and risk of CHD and the health benefits of cessation. The study is based on a survey of 40,637 British physicians who responded to a 1951 questionnaire inquiring about smoking behavior. A second questionnaire was mailed to men in 1957-58 and to women in 1960-61; the response rate was 98 percent. The 10-year followup (Doll and Hill 1964) used the updated data to assess risk among former smokers. Additional questionnaires were distributed in 1966 and 1972, with response rates of 96 and 98 percent, respectively. The 20-year followup of 34,440 men (Doll and Peto 1976) showed a reduction in CHD mortality among former smokers. The benefits were more apparent in the younger age group, and the excess risk declined with increasing duration of abstinence. In men aged 30 to 54 years, the relative risk among former smokers of I to 4 years' duration was 1.9 compared with never smokers; relative risk further declined to 1.4 to 1.3 with a maximum of 20 years' duration of abstinence. In contrast, persistent smokers had a relative risk of 3.5. In this study. those who quit had smoked about 10 percent fewer cigarettes per day before quitting than did persistent smokers. ~ The British Physicians Study also included 6,194 women. for whom the data were reported separately (Doll et al. 1980). These women completed questionnaires in 1951. 1961, and 1973. In contrast to most studies among adults, a substantial minority of nonsmokine women in this cohort initiated ci-arette smokine between 1951 and 1961. Thus, the rates of smoking-related diseases among those classified as never smokers are likely to be overestimated because never smokers, defined according to the 195 1 data, included a proportion of subsequent current smokers. Overall, the relative risk of CHD mortality among formerr smokers was 0.9 compared with 1.0 to 2.2 among current smokers, depending on the amount smoked. Because there were only 26 cases among former smokers. a detailed analysis was not performed. The first large-scale American Cancer Society (ACS) cohort was assembled in 1952 when 187,783~men aged 50 to 69, living in 9 States, completed a questionnaire related primarily to smoking (Hammond and Horn 1958a,b). The men were enrolled by over 22,000 ACS volunteers each of whom was asked to enroll 10 individuals, excluding those who were seriously ill. There was no further update of cigarette use. These men were studied for fatal outcomes for an average of 44 months, for a total of 667.753 person-years. Cause of death for 11.870 individuals was determined by death certifi- cate. Compared with never smokers, the relative risk of death due to CHD among current smokers of less than I pack per day was 1.75. Amona former smokers of less than I pack per day, those quitting within the previous year had a relative risk of 2.09, those quitting I to 10 years earlier had a risk of 1.54, and those quitting for more than 10 years had a relative risk of 1.09. A similar pattern was observed among smokers of 1I pack or more per day: amonj current smokers, the relative risk was 2.2; among quitters within the past year, 3.00; among quitters of I to 10 years, ?.06; and among quitters of more than 10 years, 1.60 (Figure 3). The authors speculated that the elevated TIMN 438603 205

group was favorable, and when the specific effect of smoking cessation was examined among the 82 patients from both groups who quit after MI, approximately 15.9 percent died in the subsequent 5 years compared with 30.6 percent among the persistent smokers and 11.8 percent among the never smokers. The influence of smoking cessation on frequency of restenosis after coronary angio- plasty was assessed by comparing 84 persistent smokers with 76 individuals who stopped at the time of angioplasty (Galan et al. 1988). Patients were reexamined angiographically after an average of 7 months. Restenosis was significantly higher in persistent smokers (55 vs. 38 percent, p=0.03). Several other studies (Fleck et al. 1988; Vandormael et al. 1987) failed to find an association between smoking at angioplasty and subsequent restenosis, but those studies did not consider the impact of cessation at the time of angioplasty. Although the mechanisms of restenosis are not clear, the findings of Galan and coworkers (1988) are consistent with a fairly rapidly acting process for decreased risk after cessation. As part of the British Regional Heart Study described above, investigators also monitored 1.515 men with evidence of CHD but without MI and 428 men with evidence of prior MI at entry (Phillips et al. 1988). Smoking behavior was assessed at baseline, and the men. aged 40 to 59, were studied for an average of 7.5 years. There was no update of the smoking information. After adjustment for age and other risk factors. for those with non-MI CHD at baseline. the relative risk comparing former with never smokers was 1.4: for current smokers, it was 2.1. For those with a history of MI, the relative risk for former smokers was 1.7; and for current smokers, it was 1.9. The degree of misclassification that may have occurred during the followup period is difficult to assess. No information is available on the duration of abstinence or the dearee of severity of CHD as distributed by smoking status. In a community-based followup of 325 post-MI patients in Baltimore, MD, Goldberg. Szklo, and Chandra (1981) found that after control for several clinical and sociodemographic factors, survival among those who quit at the time of MI was substantially improved. The 1-, 5-. and 10-year survival rates among those who quit were 99, 97. and 95 percent, respectively: in contrast, the rates among persistent smokers were 98. 84, and 51 percent, respectively. Despite the lack of updates on smokinQ behavior, there was a tr.end for diverQina survival between the two groups. Summary of Smoking Cessation and CHD Risk Within the past 40 years, large amounts of data regarding the effect of smoking cessation on CHD risk have been accumulated from numerous studies. However diverse in design and location, these studies consistently find that the risk of CHD is reduced among former smokers compared with those who continued to smoke. The data are compatible with a rapid, partial decline in risk, followed by a more gradual decline reachina levels of never smokers after a prolonged period. The initial decline appears to occur within 1 year of cessation or perhaps even less and constitutes a reduction of about one-half or more of the excess risk associated with current smoking. The remaining decline in excess risk is more gradual, with the risks reaching those of never smokers only after a number of years of smoking abstinence. This pattem of TIMN 438637 239

TABLE 7.-Case-control studies of smoking cessation and risk of stroke Reference Source and case-c:ontrol numbers Outcome Bonita et al. (1986) New Zealand: 132 cases; 1,586 community controls Bonita (1986) New Z.ealand: I 15 cases; 1,586 community controls Bell and Ambrose Scotland: 236 cases; general (1982) population control; (sample from survey by Tobacco Research Council) Taha, Ball, England: 178 cases, compared Illingworth (1982) to UK population Stroke, excluding suhatrachnoid hemorrhage Subarachnoid hemorrhage 73.3%, of consecutive series with smoking data recorded Survived subarachnoid hemorrhage • Relative risk as compared with never smokers" Strokes among ti>rmer smokers Former smokers Current smokers NR 1.4 (0.8-2.6) 2.4 NR 1.0 (0.5-I .9) 3.8 10 Men Hemorrhage 0.19 0.16 4 Infarction 0.14 0.88 I t lemorrhagic infarction 0.63 1.14 3 Women Hemorrhage 0.58 0.76 1 Infarction 0.33 1.99 0 Hemorrhagic infarction NR 3.00 7 Men h 2.1 4.7 12 Women 1.5 2.6

TABLE 4.-Continued Ref'erence 1'upulation Intervention Outcome Cases among former smokers Rose, 12 pairs of factories in UK, Diet, antisnwking, Nonfatal MI 403 Tunstal I-Pedoe, 1812 1O men aged 40-59 hypertension control and CI ID deaths I Ieller (1983) Rose et al. (1982) 1,445 healthy British civil Antismoking advice CFID deaths 49 servants all smoking at high CI ID risk Wilhelmsen et al. 10,004 random Oiiteborg Antihypertensive, dietary, Mujor CIiD NR (1986) men aged 45-55 antismoking advice Overall effect of intervention 4% net reduction in prevalence of current smoking, virtually no difference in outcome between the two groups 19% reduction in intervention group No difference Effect of smoking cessation (nonrandomized) No specific analysis of ex-smokers 19% CHD reduction in group offered antismoking advice, not statistically significant Intervention achieved only small differences between the groups for smoking and other risk factors NOTE: CIID=coronary hcart disease; MRfIT=Muhiple Risk Facior Intervention Trial; Ml=mya:ardial infarctiun; NR=not reported.

EX-C3GAiiET7E SMOKEAS IN 1952 Stll Stopped Stopped Stdl smokinq smoklnq smoking imoking in 1952 <1 yr 2i0 yr in 1952 2.09 2.20 •.`~ 206 Smoked _1 ppd OBSERVED 709 EXPEOTFD 709 23 so 40 604 1s 64 40 11 52 25 275 6 31 25 FIGURE 3.-Mortality ratios due to coronary artery diseases; rates for men who have stopped smoking are compared with those for men who never smoked and those for men still smoking in 1952 NOTE: ppd=packs/day. SOURCE: Hammond and Hom (1958b). risk among recent quitters reflected the inclusion of men who stopped smoking because of early symptoms of heart disease. A second cohort study, the ACS Cancer Prevention Study I(CPS-I) (formerly called the ACS 25-State Study), was undertaken bet~veen 1959 and 1972. Recruitment was by family, and eligible families had at least one person aged 45 or older. All family members aged 35 or older were asked to participate in the study; more than I million persons were enrolled. In a 6-year followup of 358.534 men free of diagnosed serious illness, clear reductions in risk of CHD mortality were observed among former smokers compared with current smokers (Hammond and Garfinkel 1969). Among those smok- ing less than 1 pack per day, the relative risk among current smokers was 1.90. Among those who stopped in the previous year, the relative risk was 1.62, and among those 214 TIMN 438612

had more severe coronary artery disease. The median score among never smokers and current smokers was 0.2 and 2.8, respectively. For past smokers, the data were presented by duration since quitting. There was no evidence for a trend of decreased effect by increasing time since cessation. The median score for those quitting within the previous 5 years was 5.0; for 5 to 10 years, 5.0; and for 10 years or more, 7.5. Coronary atherosclerosis was positively correlated with lifetime number of cigarettes smoked among both current or past smokers. In this study, past smokers had a slightly worse coronary risk profile than other groups. No information was provided about past or concurrent illness that may have motivated the former smokers to quit. Nonetheless, this study supports the view that cigarette smoking is a risk factor for atherosclerosis and that a substantial duration of abstinence may be necessary to appreciably reduce its extent. Weintraub and coworkers (1985) evaluated smoking history in 1,349 coronary arteriography patients. Of these patients, 984 had significant coronary disease (75 percent or more obstruction). Amount of current smoking was not a significant predictor of serious obstruction after total pack-years were considered. On average, the risk for such obstruction increased by about 1 percent per pack-year. ~ Cross-sectional studies of arteriographic findinas can be difficult to interpret because patients undergoing angiography are clearly not representative of the general popula- tion. Nonetheless, these studies support the view that smoking causes an increase in atherosclerosis and that very recent quitting has little impact on coronary stenosis. Fried. Moore, and Pearson (1986) studied the effects of smoking by assessing the coronary diameter in 31 men who had normal coronary arteriograms. Men with any detectable stenosis in the main coronary arteries or more than 25 percent in any coronary branch were excluded to assess the effects of smoking on the caliber of coronary arteries in the absence of atherosclerosis. These researchers found that after adjustment for alcohol intake (which is associated with wider arteries), current and former smokers had 40 to 50 percent narrower arteries than did never smokers. The past smokers had somewhat narrower arteries than current smokers although this was not statistically signifcant. Of the 1 1 ex-smokers. 6 had quit in the previous year. This study sua~ests the possibility of another persisting effect of smoking. apart from promoting atherosclerosis, not rapidly reversed by cessation. ~ Studies of Smoking Cessation and Risk of MI Among Healthy Persons Case-Control Studies Table I summarizes data from case-control studies (Willett et al. 1981: RosenberQ, Kaufman. Helmrich, Miller et al. 1985: LaVecchia et al. 1987: Rosenbera. Palmer, Shapiro 1990), of men and women from the United States and abroad. Prospective studies of CHD are generally considered less prone to bias than case-control studies, although case-control studies are probably less susceptible to misclassification result- ing from resumption of smoking among former smokers. For example, an individual diagnosed with a recent MI can probably recall his or her smoking status just before the infarction with considerable accuracy (Chapter 2). Thus, case-control studies may 200 TgAIN 438598

CHAPTER 7 SMOKING CESSATION AND NONMALIGNANT RESPIRATORY DISEASES 275 TIMN 438672

1982; Taha, Ball, Illingworth 1982; Bell and Symon 1979), population smoking rates rather than a true concurrent control group were used for comparison purposes. Despite these limitations, the risk of stroke among former smokers has been consistently lower than that among current smokers. Data for subarachnoid hemorrhage (Bell and Symon 1979; Taha, Ball, Illingworth 1982) show a persistent elevation in risk among former smokers compared with never smokers; however, this risk is lower than among current smokers. Prospective Cohort Studies To date, a total of 14 prospective cohort studies have reported sufficient detail to categorize former smokers as a specific subgroup monitored for incidence of stroke. These studies have obtained information on smoking status at baseline through inter- view or self-administered questionnaire and have observed populations for 2 years (Nomura et al. 1974) to 26 years (Wolf et al. 1988). Other cohort studies have reported the relation between cigarette smoking, and stroke but have not included sufficient details to categorize ex-smokers as a unique exposure group. In each of the studies included in Table 7, the risks among former smokers and among current smokers are reported compared with the risk among never smokers. The earlier prospective studies tended not to show a positive relation between smoking and stroke, and in several studies, the risk amonQ past smokers was higher than that among current smokers. In a multivariate analysis of data from the Whitehall Civil Servants Study (18,403 male British civil servants), the relative risk of stroke was 2.2 among current smokers of 15 cigarettes per day compared with never smokers, whereas the relative risk among former smokers was 1.5 (Fuller et al. 1983). Among British women, current smokers experienced a 3.0 relative risk of subarachnoid hemorrhage, and former smokers experienced a 2.3 relative risk (Vessey, Lawless, Yeates 1984). Lower elevations in risk were found among individuals experiencing ischemic strokes. No excess risk of stroke was observed amonQ 2,748 current or former smokers, residents of Cook County, IL (Ostfeld et al. 1974), or in 47,423 residents of Washington County, MD (Nomura et al. 1974). Doll and Peto (1976) studied 34,440 male British physicians for 20 years and updated information on cigarette smoking after 6 and 15 years. These researchers used similar methods for studying female British physicians among whom smokina status was updated after 10 years (Doll et al. 1980). Only slight elevations in risks of stroke were seen among male current or former smokers, and no excess risk was found among female current smokers. Similarly, Okada and colleagues (1976) found no significant elevation in risk of stroke among current or former smokers in a Japanese population. ~ In 14 cohort studies published after 1980, the relative risks amonQ former smokers were lower than those reported for current smokers (Table 7). Ro~ot and Murray (1980) observed U.S. veterans and defined the population of former smokers as those who had stopped smoking for reasons other than a doctor's orders. These former smokers had a relative risk of 1.02; current smokers had a relative risk of 1.32. In a study of 7,895 Hawaiian men of Japanese ancestry (Abbott et al. 1986), 658 smokers who quit in the first 6 years, of followup were monitored for another 6 years; '' I.ll~IN 438647 _ 249

among men was attributable to smoking; among women, 6 percent was estimated to be attributable to smoking. Studies of Smoking Cessation and Risk of Cerebrovascular Disease In this Section, data from cross-sectional, case-control, prospective cohort, and intervention studies are reviewed. As discussed in Chapter 2, misclassification of former smokers because of recidivism during the followup period is a general concern in prospective studies. However, case-control studies of stroke are limited by the relatively high fatality rate for incident cerebrovascular events, particularly for sub- arachnoid hemorrhage. This often excludes many incident cases or forces the use of proxy information from next of kin or other relatives. In all epidemioloeic studies of past smoking and risk of stroke, careful classification of stroke by pathophysiologic type is important. Details of the relation between past smoking and risk of stroke are presented in Tables 7 and 8 for each type of stroke reported by investigators. Cross-Sectional Studies In a cross-sectinal analysis of 1,692 black and white men and women admitted for diagnostic evaluation of the carotid arteries. Tell and coworkers (1989) reported a signiticant relation between cicrarette smokin- an the thickness of carotid artery plaque assessed usina B-mode ultrasonoQraphy. Based on self-report, patients were charac- terized as either nonsmokers (never smoked or quit more than 10 years earlier), former smokers (quit between 10 years and I month earlier), or current smokers. After adjusting fora patient's age, race, sex, and history of diabetes mellitus and hypertension, the mean plaque scores differed significantly among the three smoking groups. The mean difference in plaque thickness compared with that which could be expected was -0.31 mm for nonsmokers, 0.04 mm for former smokers, and 0.32 mm for current smokers. The absolute difference in mean plaque scores between nonsmokers and current smokers was 0.63 mm (95-percent CI. 0.45-0.81 mm), between nonsmokers and former smokers. 0.35 mm (95-percent CI, 0.17-0.54 mm), and between former and current smokers, 0.27 mm (95-percent CI. 0.08-0.47 mm). These data suggest a slower rate of progression of atherosclerosis among persons who have quit smoking compared with those who continue to smoke. ~ In a cross-sectional study of cerebral blood flow levels in 268 neurologically normal volunteers, Roaers and coworkers (1985) observed that subjects who quit smoking had significantly higher cerebral perfusion levels than subjects who continued to smoke. Case-Control Studies Case-control studies addressing the relation between smoking and risk of stroke are summarized in Table 7. In many other published case-control studies, former smokers have not been specifically identified as a distinct exposure group. In those studies that identify former smokers, the number of cases has been very small or unspecified except for the study by Donnan and colleagues (1989). In several studies (Bell and Ambrose 246 TIMN 438644

TABLE 8.-Continued Relative risk compared with never smokers° Reference Population Cases among Former Current Followup former smokers Outcome smokers smokers ACS (unpublished Women 220 cig/day tabulations) Quit <1 yr 0.29 2.33 (continued) 1-2 yr 0.51 3-5 yr 0.71 6-10 yr 0.84 11- 15 yr 0.23 >_16yr 0.73 NOTE: N/A=not applicable; ACS CPS-t and -II=Anlerlcnn Cancer Society Cancer Preventlon SlulheX I and II; NR=not retxorted; ICD=lntemational Classification of Disease. °95`7o confidence interval shown in parentheses when available. t'Rclative risk calculated from data presented in original paper. `Relalive risk reported by Shinton and Beevers (1989) d Datn for women funncr smokers not presented separalely. `Excluding those with a histury uf cancer, heart Jisease, or stmke a1 enrollment.

'TAF3LE 8.-Continued Relative risk compared with never smokers" Reference Population Followup Cases aimong former smokers Outcome Former Current smokers smokers Doll et al. (1980) British physicians: 6,194 22 yr NR Death due to cerebral 1.18 1-14 cig/day: 0.93 women thrombosis 14-24 cig/day: 0.45 ?25 cigJday: 0.19 Rogot and Murray US veterans: 248,046 men 15 yr 1,279 Stroke lCD 330-343 1.02 1.32 (1980) (7th revision) Fuller et al. (1983) Whitehall civil servants: 10 yr 34 Stroke mortality 1.52 1-9 cig/day: 1.0` 18,403 men aged 40-64 10-19 cig/day: 2.0 Z20 cig/day: 2.3 Vessey, Lawless, 17,0(>n UK women aged I0-16 yr 2 Subarachnoid 2.3e 3.0 Yeates (1984) 25-39 4 Nonhemorrhagic 1.3 1.4 Abbott et al. (1986) 1lonofulu I leart Study: 1,895 12 yr; 11 Thromboembolic 1.6 (0.7-3.8) 3.00 men of Japanese origin; 658 6 yr 3 I lemorrhagic 1.8 (0.4-9.0) 6.10 smokers who quit in first 6 yr Total 1.5(1.0-2.3) 3.50 Welin et al. (1987) 789 men living in Goihenburg, 678 examined 18.5 yr; I I yr NR Excluded subarachnoid hemorrhage 1.18b 1.67 Carstensen, 25,159 Swedes 16 yr 124 Cerebrovascular mortality 1.10 1-7 g/day: 0.9 Pershagen, Eklund ICD 430-438 8-15 g/day: 0.9 (1987) >{ 5 gjday: 1.1

SMOKING CESSATION AND CEREBROVASCULAR DISEASE Stroke is the third leading cause of death in the United States. It is also a major cause of morbidity, with approximately 400,000 Americans suffering strokes each year (Graves 1989). The two major types of stroke are ischemic strokes due to occlusion of a vessel by an embolus or thrombus and hemorrhagic strokes resulting from subarach- noid or parenchymal hemorrhage. The terms cerebrovascular accident and stroke are nonspecific and usually refer to clinical syndromes resulting from cerebral infarction or hemorrhage. A thrombotic or embolic stroke may be caused by atherosclerotic disease of the extra- or intracranial blood vessels. Embolization from the heart or extracranial arteries is also an important cause of stroke. In the Framingham Study, atherothrombotic brain infarction (referred to in this Chapter as ischemic stroke) accounted for 52.9 percent of strokes (Wolf et al. 1988). Improved diagnostic methods have provided a better categorization of the causes of stroke. ~ The 1964 Report of the Surgeon General (US PHS 1964) noted a moderate increase in the mortality rate from cerebrovascular disease in cigarette smokers compared with nonsmokers in the original ACS 9-State Study (Hammond and Horn 1958a.b) and the U.S. Veterans Study (Dorn 1959). In the 1971 Report, six major prospective epidemiologic studies were reviewed (US DHEW 1971). Cigarette smokers in these studies experienced increased stroke mortality compared with nonsmokers. The 1980 Report noted that women who smoke have an increased risk of subarachnoid hem- orrhage (US DHHS 1980). The 1983 Report reviewed the data associating cigarette smoking with stroke and found an increased risk of stroke among smokers that was most evident among younger age groups (US DHHS 1983). It also noted that female cigarette smokers have an increased risk of subarachnoid hemorrhace and that the concurrent use of ciaarettes and oral contraceptives greatly increased this risk. The 1989 Report of the Surgeon General reviewed four additional large cohort studies that addressed the relation between cicarette smokinc, and risk of stroke and concluded that cigarette smoking is a cause of stroke (US DHHS 1989). In a recent meta-analvsis, Shinton and Beevers (1989) summarized the relation between cigarette smokinQ and stroke using 32 separate case-control and cohort studies. The overall relative risk of stroke associated with cigarette smokins was 1.5 (95-percent CI. 1.4-1.6). Relatve risks differed considerably for the subsets of stroke: cerebral infarction 1.9, cerebral hemorrhaQe 0.7, and subarachnoid hemorrhage 2.9. Relative risks decreased with increasing age: for persons less than 55 years of aae, the relative risk was 2.9; for those aged 55 to 74 years, the relative risk was 1.8: and for those 75 years and older, the relative risk was 1. 1. A dose-response relation was observed between the number of cigarettes smoked and risk of stroke. and women had a slightly greater relative risk than men (RR= 1.72 vs. 1.43). Based on the data from ACS CPS-II, the 1989 Report of the Surgeon General estimated that 51 percent of cerebrovascular disease deaths among men aged less than 65 years were attributable to cigarette smoking, and among women of the same age, 55 percent of cerebrovascular disease deaths were attributable to smoking (US DHHS 1989). For persons 65 years of age or older, 24 percent of cerebrovascular disease TIMN 438643 245

TABLE 5.-Studies of the effect of smoking cessation on persons with diagnosed CIiD Reference Population Mulcahy et al. 190 Dublin men aged <60 who (1977) smoked at time of first coronary insufficiency or MI Daly et al. (1987) 373 men aged <60 who smoked at time of first MI or unstable angina and survived 2 yr Sparrow, Dawber, Framingham f leart Study: 195 cohort and Colton (1978) members who snwkedat time of lirst MI I lubert, Ilolford, Framingham 1lcart Study: subjects Kannell (1982) with angina Followup Cases among former smokers Reduction in risk compared with persistent smokers" 5 yr '13 deaths 50% Average 9.4 yr; NR 10% for sudden death; <_ 16 yr 40% for total mortality 6 yr 10 deaths 40% <_26 yr NR 10-yr followup: W Salonen (1980) Nonh Karelia, Finland: 523 men 3 yr 26 deaths; 00 aged <65 who smoked at first MI 22 CI ID dealhs N . . \IZ <60 yr 90% ?60 yr 60% 26-yr followup: <60 yr 70% _60 yr 10% 40%r, (60-10) 40% (60-0) Comments Smokers (N=42) who reduced cig/day also had a lower mortality compared with persistent smokers No further classification of smoking; some Df same patients as in Daly 1983 Only 25 cases in baseline smokers, so estimates are statistically unstable Followup began 6 mo after Mt; apparent benefit more pronounced in first 6 mo of followup (60%) N W

their age-adjusted relative risk for total stroke was 1.5 compared with never smokers (95-percent CI, 1.0-2.3). Risks were similar for ischemic and hemorrhagic strokes. Concurrently, current smokers had a relative risk of 3.5 compared with never smokers. Former smokers had a significant reduction in risk of total stroke compared with current smokers (p<0.05). This analysis suggests that after adjusting for other risk factors, former smokers may be at increased risk of stroke. This residual risk may be due to the irreversibility or slow reversibility of the underlying mechanisms of smoking- attributable stroke, or the resumption of smoking among former smokers. Welin and colleagues (1987) followed 789 men born in 1913 for 18.5 years. Smoking information was updated during a followup examination after 6 years. Investigators then identified a subgroup of former smokers who were monitored for 12 years. Among these former smokers, the relative risk of stroke was 1. 18 compared with 1.67 for current smokers. Wolf and coworkers (1988) studied 4,255 men and women in the Framingham Study and updated cigarette smoking information at 2-year intervals. Among current smokers, the relative risks of overall stroke were 1.42 for men and 1.61 for women. During the 26 years of followup. 50 percent of the normotensive smokers quit smoking compared with 44 percent of the hypertensive smokers (p<0.05). Former smokers had a significantly lower risk compared with current smokers. This relation was observed amono men and women in each of the blood pressure categories. Benefits of smoking cessation were observed in the hypertensive and notmoten5ive subjects. ~ In the Nurses Health Study, current smoking was strongly associated with risk of both subarachnoid hemorrhage and thromboembolic stroke (RR= 10.3 and 3.1, respectively, for 25-cigarettes or more per day) (Colditz et al. 1988). The relative risks for former smokers were substantially lower. As described in the 1989 Report of the Surgeon General, the relative risks of stroke for smokers showed an increase when CPS-II data from 1982 to 1986 were compared with CPS-1 data from 1959 to 1965 (US DHHS 1989). These studies. usinQ the same design and methods, showed an increase in the relative risk of death from stroke among current smokers for men aQed 35 to 64 years from 1.79 in 1959-65 to 3.67 in 1982-86. For women of the same age, the relative risk increased from 1.92 to 4.80. The number of former smokers among women in CPS-I was too small to report these data separately. However, for males, the relative risk of stroke amons former smokers has shown little increase and remained only slightly higher than among never smokers. The reasons are unclear for the stronger associations between cigarette smoking and risk of stroke noted in more recent studies. However, this tendency for higher relative risks in the more recent studies has been documented for a wide variety of smoking- related diseases (US DHHS 1989). One likely explanation is that the effect of smoking is related to duration of smoking, and the cohorts of persons (especially women) who started smokin2 before aQe 20 are onlv now reachin2 middle and late adulthood (Garfinkel and Stellman 1988). Controi of hypertension has improved in the United States durina the last decade, and the incidence of stroke has declined. Thus, smoking may now play a relatively greater role in the etiology of this disease than it did in earlier periods when uncontrolled hypertension was more common. 250 TIMN 438648

with higher previous daily cigarette consumption from 0.94 for those smoking less than 10 cigarettes per day to 1.34 for those smoking 40 cigarettes or more per day compared with never smokers (Figure 7). Data from ACS CPS-II also address this relationship (Table 8). Within each level of previous smoking, the risk of stroke was clearly lower for former smokers than for continuing smokers, except among men who smoked 21 cigarettes or more per day. Other studies have had too few former smokers to classify them according to previous number of cigarettes smoked. <10 cig/day 10-20 cig/day 21-39 cig/day M Ex-Smokers 0 Current Smokers >40 cig/day FIGURE 7.-Mortalitv ratios for stroke for current smokers and ex-smokers compared with never smokers, by daily cigarette consumption, US Veterans Studv, 1954-69 SOURCE: Adapted from Roeot and Murray (1980). Effect of Duration of Abstinence The relation between duration of abstinence and risk of stroke has been addressed in only a few studies. In a case-control study that included 145 former smokers who suffered stroke. Donnan and coworkers (1989) observed that the relative risk of stroke declined monotonically over the 10 years following quittinQ; at the end of 10 years, a significant excess risk of stroke was still evident. Using 5-year intervals, Rogot and Murray (1980) reported the mortality ratios for those who had abstained. Assuming that an individual classified as a former smoker at the beainnine of the study would remain a former smoker throughout the 15 years of 252 TIMN 438650

TABLE 6.-Studies of smoking cessation and risk of death due to aortic aneurysm Standardized mortality ratios compared with never smokers Reference Population Followup Cases among Iornlersmoker5 Fonnersmokers Currentsmokerti Doll and t'eto (1976) British physicians: 34,440 men 20 yr 30 3.2 5.2 Doll el al. (1980) British physicians: 6,194 women 22 yr NR 3.0 1-14 cig/day: 1.3 t 5-24 cig/day: 1.3 Rogot and Murray US vetcrans: 293,958 nuun 15 yr 253 2.58 5.23 (1980) Carstensen. 25,129 Swedish men 16 yr 12 1.4 1-7 g/day: 1.7 Pcrshagen, Eklund 8-15 g/day: 2.7 (1987) > 15 g/day: 3.0 US RkiI IS (l989) ACS CPS-1 (25-State Study) 6 yr NR Women 3.67° 4.64 Men 2.40 4.11 Nn'fE: NR=noi reponcd: ACS CPS-1=Arou'icun Cancer Pruvention Study I. ° tndicates current and rormer smukcr..

3-year interval among 110,000 individuals (Shapiro et al. 1969). A total of 613 cases of MI were reported among men aged 35 to 64 in this group. Compared with current smokers, those who quit in the preceding 5 years had a 50-percent lower risk; compared with never smokers, the relative risk was 1.0. As in other studies, the percent reduction in risk associated with smoking cessation tended to be lower in the older age groups, but a decreased risk associated with quitting was apparent among all ages. Many studies of smoking cessation have focused on middle-aged men and women. Even as recently as the late 1970s, current smoking was considered to be a minor risk factor for CHD beyond age 65 (US DHEW 1979), and the benefits of cessation among older persons have been questioned (Seltzer 1974, 1975). Jajich, Ostfeld, and Freeman (1984) assessed the effect of quitting among 2,674 recipients of public assistance aged 64 to 75 in Cook County. IL. Of the 2,674 individuals studied, 270 were past smokers, 873 were current smokers, and 1,248 were never smokers. Participants were screened at baseline and monitored for 4 years for CHD mortality. Overall, former smokers had a relative risk of CHD mortality of 1.11 (based on 20 exposed cases), whereas current smokers had a relative risk of 1.94. The number of cases was inadequate for a detailed analysis of the effect of duration of abstinence. Persons with heart problems were not excluded at baseline. Approximately one-third of the CHD deaths were among those with such a history; therefore, it is likely that the apparent benefits of quitting may be understated because of the tendency of such individuals at high risk to quit because of illness. These data provide some evidence that the benefits of cessation extend to older adults. The British Regional Heart Study (Cook et al. 1986) monitored 7,735 men aged 40 to 59 who were randomly selected from general practice lists in the United Kinadom. .The men were screened at baseline and studied for 5 to 7.5 years for incidence of fatal and nonfatal CHD; in this interval, there were 336 CHD outcomes. Those with CHD at baseline were not excluded. Compared with never smokers, quitters had a relative risk of approximately 2.5; compared with current smokers. the relative risk was approximately 30 percent lower. Men who quit smoking within the previous 5 years had a relative risk of approximately 3.3. compared with 3.6 among persistent smokers. Those who had quit more than 5 years earlier had a relative risk of approximately 2.3. but there was no evidence for a trend of decreasing risk with increasing duration since cessation. Even those who had quit 20 or more years earlier had an elevated risk. After adjustment for other risk factors, the relative risk in this group was 1.6 (p=0. l 1). As expected. the prevalence of CHD at baseline among quitters was siznificantly higher than for either current or never smokers. Presumablv, the diagnosis of disease provided a motivation to quit. When these men were excluded, the relative risks were attenuated. Nonetheless. for those who had quit in the previous 5 years, the relative risk was still elevated at 3.2. The total years of smoking was suggested as the most important variable. It was also suggested that cessation lowered risk primarily by preventing the accumulation of further years of smoking. It is noteworthy that although results of this study are adequate to show an elevated risk among past smkers, the number of cases among former smokers is too small to provide precise estimates of risk at the various durations since quitting. For example, there are only 11 cases in the aroup that quit 20 or more years earlier. TIMN 438621 223

TABLE 5.-Continued Reference Population Followup Cases among former smokers Reduction in risk compared with persistent smokers" I lallstrom, Cobb, 310 survivors of out of hospital arrest, Mean 47.5 mo 35% for fatal recurrent Ray (1986) smokers at that time cardiac arrest Green (1987) 2,199 men who smoked at time of MI 2 yr NR 30% for CI ID I ledback and Perk 157 smokers at time of MI 5 yr 13 fatal and 50% (1987) nonfatal CI ID Galan et al. (1988) 160 patients re-angiographed after Mean 7 mo 31 % decreased for angioplasty restenosis Phillips et al. 530 male British former smokers with Mean 7.5 yr 33% for fatal or nonfatal (1988) non-MI CIID CIiD 175 former smokers with M1, aged 10% 40-59 Survival Goldberg 325 post-MI patients < IOyr , _ Quit at MI Not quit Szklo, Chandra (1981) I yr 99% 98%. 5 yr 97% 84% IUyr 95% 51% Comments Borderline statistical significance Trial of rehabilitation including smoking cessation Groups were similar at baseline No update of smoking data; no assessment of severity of baseline C!!D Independent of multiple risk factors; no update of smoking status NOTE: CI ID=corunary heart diseise; Mt=myucardud infxrclion; NR=nut rcpurteJ; C ASS=Coronary Artery Srvgery Study. "95'%, couliJence iutCrval.Ahowo in parentheses when available.

The Multiple Risk Factor Intervention Trial (MRFIT) was designed to test whether reduction of diastolic blood pressure, serum cholesterol, and cigarette smoking decreases the incidence of CHD (Hughes et al. 1981; MRFIT Research Group 1986; Grimm 1986). Men aged 35 to 57 were screened; of those in the upper 15 percent of CHD risk (based on coefficients from the Framingham Study), but without overt CHD, 6,428 were randomized to special intervention, and 6,438 were assigned to usual care. Men in the special intervention group were given intensive instructions concerning diet and smoking cessation and were treated for hypertension. Those in the usual care group were referred to their regular source of medical care. The difference in total cholesterol between the two groups was only half that expected; because of better than anticipated hypertension treatment in the usual care group, the difference in blood pressure was also substantially less than expected. At the outset. 59 percent of the participants were current cigarette smokers. After 12 months, 31 percent of the smokers in the interven- tion group had quit (verified by thiocyanate (SCN) levels) compared with 12 percent of the smokers in the control group. At the end of the 6-year trial, 46 percent of smokers in the intervention group had quit compared with 29 percent in the control aroup. Mortality resulting from CHD was only 7 percent lower in the special care group, a difference that did not approach statistical significance. The authors suggested that the small decrease in risk was due in part to the smaller than anticipated differences in risk factor levels between the two groups and that some of the benefit in risk factor reduction might possibly have been counterbalanced by an unfavorable response to antihyperten- sive therapy in some of the hypertensive patients (MRFIT Research Group 1982). Within the intervention group, those who quit in the first year had a multivariate- adjusted relative risk 50 percent lower than that of persistent smokers; in the control group, adjusted relative risk 30 percent lower than that of persistent smokers. In this trial. risk of sudden CHD death was reduced 65 percent among quitters compared with persistent smokers. Because all participants were seen at least annually. the possible misclassification of smokina status was minimized. The 10.5-year followup data from MRFIT have recently been published (MRFIT Research Group 1990). Deaths due to CHD were 10.6 percent lower in the special intervention group (95-percent CI, ?3.7 to 4.9) compared with the usual care group (two-sided p value=0.24). This reduction in risk was largely attributable to a 24.3- percent lower risk of death due to acute MI (2-sided p value=0.04). Total cardiovascular mortality was 7.1 percent lower after 10.5 years in the special intervention group compared with the usual care group (p>0.05). In one analysis not based on randomized groups, CHD mortality rates of smokers who had quit within the first 12 months of the trial and of those who were still smokinQ at that time were compared (Ockene et al. 1990). Quitters had a 37-percent reduction in mortality. After adjustment for other CHD risk factors, the reduction was 42 percent (95-percent CI. 16-60). The slightly greater benefit observed after adjustment for risk factors indicates that there was little confoundina and that it was in the direction that would tend to underestimate the benefit of cessation. This analysis ignored any changes in smoking status after the first annual examination. To the extent that either some of the quitters resumed smoking or some of the current smokers quit, that analysis would yield an underestimate of the benefits of cessation. A second analysis compared quitters who remained abstinent at the first TIMN 438625 "7

TABLE 8.-Prospective cohort studies of smoking cessation and risk of stroke Reference Population Ostfetd ct al. (1974) 2,748 Cook County. IL residents receiving old age assistance abed 65-74 Nomura et al. (1974) 47,423 Washington County, MD residents Fullowup 3 yr 2 yr morbidity 8 (women) ~ Doll and Peto (1976) British physicians: 34,440 20 yr Cases among former smokers 23 27 (men) NR men Okada et al. (1976) 4,186 Japanese 6 yr NR Relative risk compared with never smokers° Outcome Former smokers Current smokers All strokes 0.91 b 1-9 cig/day: 1.29 10-19 cig/day: 0.85 _20 cig/day: 0.81 Thrombosis 1.03b 0.79 l lemorrhage 0.79 0.86 Undifferentiated 1.00 1.30 Total 0.97 0.90 Thrombosis 1.08 1.14 I lemorrhage 2.00 0.91 Undifferentiated 1.14 0.36 Total 1.26 0.92 Cerebral thrombosis rnortality 1.22 1.24 Ccrebrovascular attacks Relative risk of cerebral hemorrhage in nonsmokers was lower than in smokers or ex-smokers, but the difference was not statistically significant tJ U W

-----~ •AIRWAYS OBSTRUCTION , , ~ , , a , , , , , , , , , .........................................................................I FIGURE i.--Nonproportional Venn Diagram of the interrelationship among chronic bronchitis, emphysema, asthma, and airways obstruction. SOURCE: Snider (1988). Researchers in the United States and the United Kingdom tend to separate asthma from the other obstructive airways diseases and to deemphasize the importance of cigarette smoking in this particular clinical entity. However. the data suggest that cigarette smokinQ may influence asthma and that allergy and airway hyperresponsive- ness, strongly associated with asthma, may play a role in the development of fixed airflow obstruction (O'Connor, Sparrow, Weiss 1989). The generally accepted model of the pathogenesis of COPD is based on the results of longitudinal investigations of lung function (Fletcher and Peto 1977; Becklake and Permutt 1979; Burrows 1981; Speizer and Tager 1979) (Figure 2). The model suggests that disease development is preceded by a lona latent period durina which lunQ function declines at an accelerated rate. 280 TIMN 438675

between initial and repeat operations was significantly shorter in those who continued to smoke (Mann-Whitney test, p<0.05). Those who stopped or reduced smoking attained a significant improvement in overall survival by 12 months. A second series of 160 patients was studied for 8 years after their first hospital admission. Those who were smoking at the time of referral had a significantly poorer survival pattern than those who had stopped smoking or had reduced smoking. Similar results were observed by Jonason and Bergstrom (1987) who studied 343 consecutive patients with intermit- tent claudication and by Faulkner, House, and Castleden (1983) who studied 133 patients. A retrospective record review was undertaken at Mayo Clinic to identify nondiabetic patients with a clinical diagnosis of arteriosclerosis obliterans, and Juergens, Barker, and Hines (1960) reported the survival and amputation rates among these patients. Of 159 patients who smoked at the time of diagnosis and who survived 5 years, 88 continued to smoke and 71 abstained from smoking after diagnosis. Of the total number of patients who continued to smoke, 11.4 percent required an amputation within the 5-year period. In contrast, none of the abstainers required amputation during this period. In a recent retrospective 5-year followup study, Ameli and colleagues (1989) reported the rates of amputation and patency of 136 arterial reconstructions performed for lower limb ischemia. Of 121 patients. 103 smoked before the operation. and of the smokers 43 postoperatively discontinued smoking. The 34 patients who continued to smoke more than 15 cigarettes per day had a fivefold increase in risk for amputation at 2 years and a threefold increase in risk for amputation at 5 years compared with the 87 nonsmokers (including never and former smokers) and smokers of 15 cigarettes or less per day (p=0.013). Five years after surgery, 28 percent of patients smoking more than 15 cigarettes per day had undergone amputation compared with 1 1 percent of the patients who were nonsmokers or smoked 15 cigarettes or less per day. The effect of smoking on the patency of femoropopliteal vein bypass grafts used for treating peripheral arterial occlusion was studied among 157 patients monitored for I year (Wiseman et al. 1989). Patients who continued to smoke, identified by elevated serum SCN-, had a graft patency of 63 percent after 1 year compared with 84 percent among nonsmokers (p<0.0?). However, the analysis did not separate never smokers from those who stopped smoking near or at the time of sur2ery (p<0.02). Only serum fibrinosen levels were a stronger predictor of graft failure than serum SCNi-. Summary Overall, these studies show a lower risk of peripheral artery disease among former smokers compared with current smokers and a consistent reduction in complications of peripheral vascular disease among patients who stop smoking. Those who quit have improved performance and improved overall survival. 244 - TININ 438642

CONTENTS Introduction .......................................................279 Part i. Smoking Cessation and Respiratory Morbidity ...................... 285 Respiratory Symptoms ............................................. 285 Clinical Studies ................................................ 285 Cross-Sectional Studies of Populations .............................. 288 Occupational Groups ............................................. 296 Longitudinal Studies ................................... . . . . . . . . . 299 Clinical Studies of Possible Mechanisms ............................ 304 Respiratory Infections ............................................. 305 Smoking Cessation and Respiratory Infection ........................ 307 -------~--.-_ _----~ Part II. Pulmonary Function Among Former Smokers ...................... 308 Cross-Sectional Population Studies of FEV I ........................ ... 308 Pulmonary Function Studies After Smoking Cessation ................... 316 Changes in Spirometric Parameters After Cessation ................... 319 Tests of Small Airways Function .................................. 323 Diffusing Capacity Among Former Smokers ......................... 327 Other Measures ................................................. )28 Longitudinal Population-Based Studies ................................. 28 Part III. Airway Responsiveness, Cigarette Smoking, and Smoking Cessation .. 337 Mechanisms of Heightened Airway Responsiveness in Smokers and Former Smokers ....................................................... 338 Cross-Sectional Studies ............................................ 338 Longitudinal Studies ............................................... 39 Clinical Studies ................................................... 340 Part IV. Effects of SmokinQ Cessation on COPD Mortality ................. 341 Part V. Former Smokers With Established Chronic Obstructive Pulmonary Disease .......................................................... ~5 Effect of Smoking Cessation on FEV i Decline Among COPD Patients ...... 345 Effect of Smokina Cessation on Mortalitv AmonQ COPD Patients ........... 47 Conclusions ....................................................... 349 References ........................................................3 5l TIMN 438673 277

Smoking Cessation and Development of Peripheral Artery Disease Two studies provide sufficient detail to calculate the risk of peripheral vascular disease among former smokers compared with current smokers. Jacobsen and coworkers (1984) compared a consecutive series of 53 patients with intermittent claudication with age-matched controls free from symptoms of claudication. All patients with claudication were either current or former smokers. Among former smokers, the risk of developing peripheral arterial disease was 50 percent lower than that of current smokers. Hughson, Mann, and Garrod (1978) reported risk factors for intermittent claudication among 54 patients and 108 controls. Smoking was the risk factor most strongly associated with the development of intermittent claudication. Former smokers had an estimated 58-percent lower risk than that of current smokers. Smoking Cessation and Prognosis of Peripheral Artery Disease In a study of 91 men with mild intermittent claudication monitored for at least 6 months, patients who stopped or decreased smoking had slightly less progression of symptoms during 2.5 years of followup. but this finding was not statistically significant (Cronenwett et al. 1984). Changes in treadmill exercise tolerance were assessed among 41 patients suffering from intermittent claudication who continued to smoke durina the followup period and among 16 patients who stopped smoking after the first test and remained nonsmokers until the end of study (Quick and Cotton 1982). The maximum treadmill walking distance did not change significantly among continuing smokers (23 meter improvement, p=0.17). However, among those who stopped smoking, the improvement in maximum treadmill distance was statistically significant (86.2 meters, p=0.02). The two groups were not compared directly. During a 6-year period, the risk of developing pain at rest was studied in 224 consecutive nondiabetic patients with intermittent claudication (Jonason and RinQgvist 1985). The cardiovascular risk profiles were almost identical for 30 never smokers and 34 patients who stopped smoking within 1 year after initial examination. These two groups were combined and compared with 160 patients who continued to smoke. The cumulative percentage of patients with pain at rest after 6 years was 8 percent among those who had stopped smoking within I year after the initial examination or who were never smokers; among smokers and those who stopped smoking more than 1 year after the initial examination, 21 percent developed pain at rest (p<0.03 after adjustments for difference in presence of multiple stenoses at baseline). These data are difficult to interpret because never and former smokers were combined, but suggest that the rate of development of rest pain is decreased among former and never smokers compared with those who continue to smoke. In a followup study of 60 patients who underwent operation for intermittent claudica- tion, those who stopped or reduced smoking after referral had a much improved prognosis (Hughson et al. 1978). At baseline, clinical characteristics or the number of cigarettes smoked did not differ between those patients who decreased or stopped smoking and those who continued to smoke during the followup period. The interval TIMN 438641 243

decline in excess risk is compatible with multiple effects of smoking on the process of developing CHD, including both short-term influences on platelets and other factors relating to thrombosis which may be more rapidly reversible and long-term increases in atherosclerosis which are only slowly reversible. Persistent smokers may differ from those who quit in other ways that could affect the risk of developing CHD. A number of investigators have examined whether such differences would account for some or all of the decline in risk among those who stop smoking. The risk profiles of quitters and persistent smokers vary among studies: In some studies, there are no material differences; however, in other studies, quitters have a healthier profile; the opposite is true for still other studies. In the studies of primary prevention, none of these differences could explain even a minor portion of the decreased risk among quitters. Most studies of cessation after an MI have found that quitters had a higher baseline risk; however, their risk decreased compared with persistent smokers. Thus, both in primary and secondary prevention studies, confound- ing effects of other risk factors do not explain the apparent benefits of cessation. To the contrary, in many studies, the decrease in risk is even more pronounced after adjustment for baseline characteristics. Only a few studies have examined the impact of smokinQ cessation in relation to various other CHD risk factors. No data are available to suggest that the relative risks differ substantially in the presence or absence of other CHD risk factors; that is, the percentage reduction in risk most likely occurs across risk factor categories. However, because individuals at high risk for other reasons such as family history, hypertension, or elevated cholesterol have higher rates of CHD, a given percentage decrease in risk among these individuals is a greater absolute decrease than among those with a lower risk profile. Hence, it is of especially great importance to achieve high rates of cessation among individuals who are otherwise at high risk for CHD. Most data on the effects of smoking cessation are derived from white males, but sufficient information is available about women to indicate that the findings are similar for both sexes. Less is known about the effects of cessation among minority groups; however, there is no reason to believe that the benefits of cessation would be any different for these groups. Several studies have examined the effect of smokina cessation after age 60 on subsequent CHD risk. Data are now available that demonstrate that the benefits of cessation extend to older adults as well as to young and middle-aged adults for both primary (Table 3) and secondary prevention (Hermanson et al. 1988). Although the relative risks of CHD among current smokers tend to be lower among older persons than among younger persons, smoking cessation among older persons can have a~reater absolute effect because their rates of CHD are so much hiQher. ~ Considerable data address the effects of smoking cessation amona individuals with diagnosed CHD. A reduction in risk of further CHD-related morbidity and mortality that accompanies smoking cessation has been conclusively demonstrated. Cigarette smoking is considered the leading modifiable CHD risk factor; overwhelming evidence demonstrates that cessation reduces that risk substantially. y 240 TIMN 438638

SMOKING CESSATION AND AORTIC ANEURYSM Abdominal aortic aneurysm refers to the dilatation or expansion of the aorta because of degenerative or inflammatory destruction of the components of the arterial wall. Most abdominal aortic aneurysms are a result of atherosclerosis, although other conditions cause abdominal aortic aneurysms. The preponderance of evidence from autopsy studies reviewed in the 1983 Report of the Sur-eon General suggests that cigarette smoking aggravates or accelerates aortic atherosclerosis (US DHHS 1983). In addition, epidemiologic studies published up to that time indicated that smokers had elevated death rates from ruptured abdominal aneurysm compared with nonsmokers (Hammond and Garfinkel 1969; Hammond and Horn 1958a.b; Kahn 1966; Weir and Dunn 1970). Mechanisms whereby smoking causes atherosclerosis are reviewed in this Chapter. Studies of Smoking Cessation and Risk of Aortic Aneurvsm Several of the larger prospective cohort studies reviewed above have reported results for mortality by cause of death. The data on mortality among former smokers from abdominal aortic aneurysms reported in five prospective cohort studies are summarized in Tabie 6. A consistent pattern is seen amonQ men in these studies, with an excess risk of mortality approximately 50 percent lower among former smokers than among current smokers. However, excess risk among former smokers has remained about two to three times hieher than that among never smokers. A similar pattern was also present for women in ACS CPS-II. AlthouQh data for women are limited, Doll and associates (1980) reported 11 deaths due to aortic aneurysm occurring during 22 years of followup among 6,194 women. Overall, these data indicate that former smokers have a reduced risk of death from aortic aneurysm compared with current smokers. More detailed analyses by duration of smokinQ abstinence have not been presented. SMOKING CESSATION AND PERIPHERAL ARTERIAL OCCLUSIVE DISEASE The peripheral arteries include those branches of the aorta that supply the upper and lower extremities and the abdominal viscera. Most peripheral arterial occlusive disease results from atherosclerosis, althouah other conditions may cause obstruction of these arteries. Symptomatic atherosclerosis of peripheral arteries occurs most often in the vessels of the lower extremities. The 1983 Report of the Surgeon General reviewed risk factors and epidemiologic data relating to the etiology of peripheral artery disease (US DHHS 1983). In that Report, an extremely strong association between cigarette smokinQ and diagnosis of peripheral artery disease was observed (US DHHS 1983). Cigarette smokine was the strongest risk factor for peripheral artery disease in the Framingham Study (Kannel, McGee, Gordon 1976). In this Section, the impact of smoking cessation on risk of developing peripheral artery disease is reviewed. In addition, the influence of cessation on treadmill time, rest pain, progression to amputa- tion, and survival among patients with diagnosed peripheral artery-disease is discussed. TIMN 438639 241

// i1 ~i a I 1 s 5 10 15 20 30 40 AGE-years t I 50 60 I 70 I eo FIGURE 2.-Theoretical curves depicting varying rates of decline of FEV 1 NOTE: Curves A and B represent never smokers and smokers, respectively, declining at normal rates. Curve C shows increased decline without development of COPD. Rates of decline for former smokers are represented by curves D and E for those without and with clinical COPD, respectively. Curves F and G show rates of decline with continued smokins after development of COPD. SOURCE: Speizer and Tager (1979). Several features of this conceptual model merit emphasis in relation to smoking. First, disease development may occur as a result of factors that accelerate decline in adult life, lead to less than maximal growth, or both. Second, because of the extremely long latent period from the onset of smoking to disease development, factors important in childhood and young adulthood cannot be addressed in longitudinal studies that begin in adulthood. Third, longitudinal studies of children and adults have shown that pulmonary function levels are very stable over time with tracking correlations ranging from 0.70 to 0.90. This high degree of longitudinal correlation, consistent with both environmental and genetic determinants of disease, demonstrates the importance of previous level of function as a major determinant of future disease risk. Research on risk factors for COPD was reviewed extensively in the 1984 Report of the Surgeon General (US DHHS 1984). The review leads to several general findings 281 TIMN 438676

were more likely to quit than the lighter smokers. Former smokers had a lower rate of subsequent CHD. There was a suggestion that older persons benefited less; however, this finding could not be confirmed because only a small fraction of the 25 older smokers actually quit. . Salonen (1980) monitored a Finnish cohort of men less than 65 years of age whose smoking behavior was assessed 6 months after MI. Of these, 352 were never smokers, 302 were persistent smokers, and 221 quit smoking within 6 months after MI. Three years after MI, quitters had a 40-percent reduction in risk of total mortality (95-percent CI, 10-60 percent) and of CHD death (95-percent CI, 10-60 percent) compared with persistent smokers. The reduction in risk was more pronounced in earlier periods; between 6 months and 1 year, mortality was reduced by 60 percent (95-percent CI, 10-80 percent). It is possible that the apparent decline in benefit may represent misclassification because current smokers continued to quit but were still analyzed as current smokers. The benefits of quitting were strongest among those with the best prognosis after infarction. Of post-MI deaths, 28 percent were estimated to be at- tributable to continued smoking. As part of the Norwegian trial of timolol use after MI, mortality of the 1,884 participants was ascertained over an average of 17 months according to smoking status. Virtually no differences were observed (Von der Lippe and Lund-Johansen 1982). Across both the timolol and placebo groups. 8 percent of the nonsmokers died, compared with 8 percent of those who stopped smoking before entry into the trial. 7 percent among those who quit in the first month of the trial, and 8 percent amona persistent smokers. However, there was a reduction in reinfarctions, 8 percent among those who quit in the first month of the trial compared with 12-percent among persistent smokers (Ronnevik, Gundersen, Abrahamsen 1985). Shapiro. Howat, and Singh (1982) monitored 142 patients who survived a first MI that occurred when the patient was younger than age 45. Of these patients. 50 who continued to smoke more than 20 ciQarettes per day had substantially higher mortality rates (58-percent 10-year mortality by life table methods) than did the 61 never and former smokers (12-percent mortality). The survival curves began to diverge 1 year after MI. Unfortunately, data were not presented separately for former smokers, and apparently there were only a small number of never smokers. Abera, and colleaaues (1983) studied 983 men aged 67 years or less who were listed in the MI Register of Gotebora between 1968 and 1977. The men were smokers within 3 months of their initial MI. who survived hospitalization. Not all men listed in the Register were included in the study, but the selection process did not introduce bias. Quitting was defined as not smoking 3 months after the infarction. Followup began at that point and continued for 10.5 years. The 542 males who had stopped smoking by 3 months after infarction had a significantly worse prognosis, based on predischarge characteristics, than did the 441 persistent smokers. Those who quit had substantially more left ventricular failure and higher peak enzyme levels durinQ hospitalization. Based on these and other preinfarction and hospitalization variables, those who quit had a predicted 2-year mortality that was 8 to 9 percent higher than that of persistent smokers. However, despite this slightly worse baseline prognosis, quitters had a significantly lower mortality than did persistent smokers. Overall, the 5-year mortality 235 TININ 438633

was significantly reduced among quitters, with a cumulative mortality rate 30 percent lower. The effect was somewhat stronger among those aged 50 or older than among younger men, but was significant in both age groups. The cumulative 5-year reduction in recurrence of MI was 30 percent. These estimates almost certainly underrepresent te true effect of cessation for two reasons: quitters at baseline had a distinctly worse prognosis, and smoking cessation was defined only at the point 3 months after infarction. It is likely that some of the smokers quit at a later point; this would tend to dilute the smoking group with ex-smokers who enjoy a lower risk. Thus, the rates of mortality and reinfarction among truly persistent smokers would be underestimated in this study. The two groups began to diverge for both endpoints after as little as 1 year postinfarction, and the differences increased with time. This report confirmed and extended initial findings from that study (Wilhelmsson et al. 1975). Several studies have monitored patients with aneiographically diagnosed coronary disease. Kramer and coworkers (1983) studied 278 men with sequential coronary angiograms. These researchers found that neither cigarette smoking at the initial or followup examination nor smoking cessation was predictive of progression of atherosclerosis. ~ Daly and colleagues (1983) studied 217 men who stopped smoking after a first diagnosis of unstable angina or MI and 157 persistent smokers. Smoking status was defined 2 years after the first diagnosis. As in the AberQ study (1983), those who quit tended to have a more serious diagnosis than the persistent smokers. However, quitters enjoyed substantial protection compared with persistent smokers. For total mortality, risk was reduced by 60 percent among those who quit smoking compared with continuing smokers; for fatal reinfarction, risk was also reduced by 60 percent. During the first 6 years of followup, the reduction in risk was 40 percent (95-percent CI. 10-60 percent), but in the followup period of 7 to 13 years, the benefits of quitting were more marked, with a reduction in risk of 80 percent (95-percent CI. 50-90 percent). The benefits of quitting were more marked among those with less severe initial disease. In this study, quitters had a lower cumulative mortality than did never smokers with these diagnoses. Those never smokers may have had more coronary risk factors other than smokina which may be less amenable to change than smoking. In a later study with some of the same patients, Daly and coworkers (1985) found that 1 year after the initial event. 241 quitters had a 40-percent lower prevalence of anQina compared with 143 persistent smokers. However, by 6 years of followup, the prevalence of anQina was the same in both groups and remained similar throughout the followup period of 17 years. Green (1985) noted that the prevalence of angina 6 months after infarction amona 851 ex-smokers was equivalent to that among smokers. How- ever, it is unclear whether the ex-smokers were smoking at the time of the event. Most studies of the effect of post-MI cessation have been conducted among men. Johansson and colleaaues (1985) examined 156 women in GoteberQ, younger than 65, who were smokers at the time of their first MI. The definitions and criteria were the same as those in the study by Aberg and coworkers (1983). Three months after infarction. 75 women continued to smoke and 81 had stopped. As in the GoteberQ S tudy of men (Abero et al. 1983), women who quit had more severe infarctions. Despite the worse prognosis normally associated with the higher enzyme elevations and other 236 TIMN 438634

MEN Cigarette Smokers •-•••-••• Pipe & Cigar Smokers Ex-Smokers ----- Never Smokers FIGURE 5.-Prevalence of cough and phlegm by smoking group NOTE: Persons with grade 2 cough and phiegm had both symptoms and at least one symptom for _3 mo/yr. SOURCE: Payne and Kje(sberg( t964). 293 Z`I.11IN 438688

indications of severity, the quitters had a significantly better survival. The reduction in .risk compared with smokers remained at 60 percent (95-percent CI, 20-80 percent), and after adjustment for prognostic features before and during the infarction, the reduction remained at 60 percent. When compared with never smokers, the relative risk among quitters was 1.1. The reinfarction rate was slightly, though not significantly, higher among persistent smokers. Similar findings for a rapid benefit were observed in the small study of Perkins and Dick (1985). For 5 years, these researchers monitored 52 patients (including 11 women) who stopped smoking at the time of the infarction and 67 persistent smokers (of whom 18 were women). Men who quit had a 50-percent reduced risk of death; for women it was 60 percent lower. As part of the Coronary Artery Surgery Study, the effect of smoking cessation on risk of clinical CHD outcomes was assessed in men with documented coronary atherosclerosis by angiography (Vlietstra et al. 1986). The death rates among 1,490 quitters were compared with those of 2,675 persistent smokers and 2.912 never smokers. Men who were quitters at baseline but who subsequently resumed smoking and those who were smokers initially but later stopped were excluded from the analysis. Hence, this study was largely free of misclassification. As in most of the other studies, the quitters had slightly worse prognoses than did the persistent smokers. At every level of risk, however, quitters had a significantly better 5-year survival. Overall, the reduction in risk (from Cox regression) was 40 percent (95-percent CI, 20-50 percent). The benefit was slightly more pronounced among those with the worst baseline prognosis. Overall, the 5-year survival rate among quitters was similar to that of never smokers (85 vs_ 87 percent, respectively). Nearly all the benefit was attributable to a decreased rate of CHD death. After adjustment for prognostic score. the rate of hospitalization for MI was substantially higher among persistent smokers than among quitters (11.3 vs. 7.1 percent, respectively). For both fatal and nonfatal endpoints, the rates began to diver~e substantially after about 1 year (Figure 6). Because of the careful study design and the unusually large number of cases, the results of this study must be accorded considerable weight. In an extension of the analysis of survival data from the Coronary Artery Surgery Study, the effects of smoking cessation were examined in a population of individuals aged 55 and older with angiographically documented coronary disease (Hermanson et al. 1988). As in the previous report. persistent smokers were defined as those 1,086 smokers who did not quit throughout the 6-year followup period, and quitters were those 807 who stopped smoking 1 year before the baseline angiogram and who did not resume smoking during followup. The experience of 3,045 younger subjects aged 35 to 54 years was also examined. At every age, quitters had better survival rates than did persistent smokers, and there was no evidence that the benefit was attenuated with increasing age. Employing a different approach, Hallstrom, Cobb, and Ray (1986) studied a cohort of 310 men who smoked and were discharged from the hospital after an episode of out-of-hospital cardiac arrest. After the arrest, 219 men continued to smoke and 91 men quit. During the average 47.5 months of followup, 67 persistent smokers and 18 former smokers died of a recurrent cardiac arrest. After adjustment across baseline risk TIMN 438635 237

WOMEN AGE 10 20 30 40 50 Cigarette Smokers • Ex-Smokers ......... Pipe & Cigar Smokers ----- Never Smokers 60 70 FIGURE S. (Continued)-Prevalence of cough and phlegm by smoking group NOTE: Persons with grade 2 cough and phlegm had both symptoms and at ieast one symptom for _3 mo/yr. SOURCE: Payne and Kjelsberg(1964). TIMN 438689 294

GRAVES, E.J. National Hospital Discharge Survey. Vital and Health Statistics Series 13(99):1-60, April 1989. GREEN, K.G. Does stopping smoking delay onset of angina after infarction? (Letter.) British Medical Journal 291(6504):1281, November 2, 1985. GREEN, K.G. Falsely favourable early prognosis for continuing smokers following recovery from acute myocardial infarction. Information from the multi-centre practolol trial. British Journal of Clinical Practice 41(6):785-788, June 1987. GRIMM, R.H. JR. The drug treatment of mild hypertension in the Multiple Risk Factor Intervention Trial. A review. Drugs 3 1 (Supplement 1): 13-21, 1986. HAAPANEN. A., KOSKENVUO, M.. KAPRIO, J., KESANIEMI, Y.A., HEIKKILA, K. Carotid arteriosclerosis in identical twins discordant forcigarette smoking. Circulation 80(1): 10-16, July 1989. HALLSTROM, A.P., COBB. L.A., RAY, R. Smoking as a risk factor for recurrence of sudden cardiac arrest. New England Journal of Medicine 314(5):271-275, 1986. HAMMOND, E.C., GARFINKEL, L. Coronary heart disease, stroke, and aortic aneurysm. Factors in etiology. Archives of Environmental Health 19:167-182, August 1969. HAMMOND, E.C., HORN, D. Smoking and death rates-Report on forty-four months of follow-up of 187,783 men. I. Total mortality. Journal of the American iYledical Assoc•iation 166(10):1159-1172, March 8, 1958a. HAMMOND, E.C., HORN, D. Smoking, and death rates-Report on forty-four months of follow-up of 187.783 men. II. Death rates by cause. Journal of the American Medical Association 166(11):1294-1308, March 15, 1958b. HARENBERG, J., STAIGER. C., DE VRIES, J.X., WEBER, E., ZIMMERMAN. R.. SCHET- TLER, G. The effects of a combination of ciaarette smoking and oral contraception on coagulation and fibrinolysis in human females. Klinische Woc•hensc•hrift 63(5):221=?24. March 1. 1985. HARTZ, A.J., ANDERSON, A.J., BROOKS, H.L.. MANLEY, J.C., PARENT. G.T., BAR- BORIAK, J.J. The association of smoking with cardiomyopathy. New England Journal of Medicine 311(19):1201-1206, November 8. 1984. HARTZ, A.J., BARBORIAK, P.N.. ANDERSON, A:J.. HOFFMANN. R.G., BARBORIAK, J.J.. Smoking, coronary artery occlusion, and nonfatal myocardial infarction. Journal of the American Medical Association 246(8):851-853, August 21, 1981. HEDBACK, B., PERK. J. 5-year results of a comprehensive rehabilitation programme after myocardial infarction. European Heart Journal 8(3):234-242. March 1987. HENNEKENS, C.H., LOWN, B., ROSNER. B., GRUFFERMAN, S., DALEN, J. Ventricular premature beats and coronary risk factors. American Journal of Epidemiology 112(1):93-99, July 1980. HERMANSON, B.. OMENN, G.S., KRONMAL, R.A., GERSH. B.J. Beneficial six-year outcome of smoking cessation in older men and women with coronary artery disease: Results from the CASS Registry. New England Journal of Medicine 319(21-):1365-1369. November 24, 1988. HJERMANN, I. Smoking and diet intervention in healthy coronary high risk men. Methods and 5-year follow-up of risk factors in a randomized trial. The Oslo Study. Journal of the Oslo City Hospitals 30(1):3-17, January 1980. HJERMANN, I., HOLME, I., LEREN. P. Oslo study diet and antismokinQ trial: Results after 102 months. American Journal of:Yledicine 80(Supplement'_A):7-12, February 14, 1986. HJERMANN, I., HOLME, I., VELVE BYRE, K.. LEREN, P. Effect of diet and smokin; intervention on the incidence of coronary heart disease. Lancet 2(8259):1303-1310, Decem- ber 12, 1981. 265 TIMN 438663

hemorrhage are not available from that study. Because oral contraceptive preparations used today provide substantially lower doses, the risk of cardiovascular disease as- sociated with their use and their interaction with cigarette smoking may be different than observed for the early high-dose preparations. Effect of Smoking Cessation After Stroke In contrast with CHD, in which the focus after MI is prevention of recurrent disease, the center of attention after a major cerebrovascular event is rehabilitation. For CHD, substantial evidence shows the benefits of abstaining from smoking after onset of CHD. Comparable data are not available on the benefits of abstinence after stroke. Summary Risk of stroke resulting from occlusion of the cerebral arteries and from subarachnoid hemorrhage is increased approximately twofold to fourfold among current smokers compared with never smokers. After cessation, the excess risk decreases steadily. In some studies, the risk of stroke among former smokers becomes indistinguishable from that of never smokers within 5 years: in other studies, this decrease did not occur until after 10 years or more of smoking abstinence. The reduced risk of stroke among persons who stop smoking is independent of the amount previously smoked and other known risk factors for stroke. Similar reductions in riSk of stroke after cessation are seen among men and women, but few data are available for minority populations. CONCLUSIONS 1. Compared with continued smoking. smoking cessation substantially reduces risk of coronary heart disease (CHD) among men and women of all aQes. 2. The excess risk of CHD caused by smoking is reduced by about half after l year of smoking abstinence and then declines gradually. After 15 years of abstinence, the risk of CHD is similar to that of persons who have never smoked. 3. Among persons with diagnosed CHD, srriokinQ cessation markedly reduces the risk of recurrent infarction and cardiovascular death. In many studies, this reduction in risk of recurrence or premature death has been 50 percent or more. 4. Smoking cessation substantially reduces the risk of peripheral artery occlusive disease compared with continued smoking. 5. Among patients with peripheral artery disease, smoking cessation improves exercise tolerance, reduces the risk of amputation after peripheral artery suroerv, and increases overall survival. ~ 6. Smoking cessation reduces the risk of both ischemic stroke and subarachnoid hemorrhage compared with continued smoking. After smoking cessation, the risk of stroke returns to the level of never smokers: in some studies this has occurred within 5 years, but in others as long as 15 years of abstinence were required. .1 260 TIMN 438658

In many instances, studies are primarily of individuals who were smokers up to the time of the infarction. Such a major health event can be a powerful motivation to quit smoking permanently. Moreover, the timing of quitting often coincides with the infarction and is therefore ascertained quite accurately. Because those with a prior diagnosis of CHD are at such high risk for another event, the estimates of effect can be relatively precise, even with a modest number of individuals under study. One difficulty in interpreting these studies is in the comparison of quitters with never smokers. Never smokers who suffer MI tend to have a worse CHD risk factor profile (apart from smoking) than smokers (Mulcahy 1983). However, most of the other risk factors are less amenable to change than smoking. After smoking is removed as a risk factor among former smokers, the effect is often a better prognosis than that for never smokers. Several of these issues and a review of the literature prior to 1983 are discussed by Mulcahy (1983). This researcher found that studies were quite consistent in showing that quitters had about half the risk of recurrent MI or CHD death compared with persistent smokers (Mulcahy 1983). Nearly all studies of this issue have indicated a benefit of cessation (Table 5). A cohort of 213 patients who survived for 28 days a first attack of coronary insufficiency or MI was studied for 5 years (Mulcahy et al. 1977). Of these, 190 were smokers at the time of the event. Of the 89 who stopped, the cumulative 5-year death rate was 14.6 percent. Of the 42 who reduced cigarette use, the rate was 14.2 percent. However, among the 59 persistent smokers, 28.8 percent died within 5 years. Nearly all of the deaths were associated with CHD. This study was extended by further accrual of patients and followup of 551 men less than 60 years of age (Daly et al. '1987). Of the 406 current smokers at the time of the event. 140 had stopped by year two. Those quitters had a 10-percent reduction in risk of sudden death'and a 40-percent reduction in risk of total mortality compared with those who continued to smoke. A 1978 report from the Framingham Study (Sparrow, Dawber, Colton 1978) com- pared the survival of 56 individuals who quit smoking after a first MI with 139 who continued to smoke after the diagnosis. Within 2 to 3 years after diaQnosis, former smokers had a significantly better survival rate than persistent smokers. The 6-year mortality rate (estimated by life table methods) was 18.8 percent among quitters compared with 30.4 percent among persistent smokers. When the risk of recurrent MI was assessed, the authors found that former smokers had a lower risk than persistent smokers, with a 6-year reinfarction rate of 15.5 percent in quitters versus 21.5 percent among smokers. However, with only eight reinfarctions among the quitters, the differences were not statistically significant. The rate of decline in risk could not be assessed because of the small samples.. Framingharn Study investigators (Hubert. Holford, Kannel 1982) conducted a long- term followup study of 130 subjects with angina pectoris. They found that smoking status at the examination ascertaininQ angina was modestly associated with subsequent risk of a later, more serious CHD outcome. Apparently, the change in smoking behavior explained this finding. Of the angina patients who smoked, 14 percent quit between the onset of disease and the biennial examination when the diagnosis was confirmed. Another 29 percent quit during the followup period. In this cohort, the heavier smokers 230 TIMN 438628

of respiratory symptoms and the level of pulmonary function (Burrows et al. 1980; Kauffman et al. 1986). Cigarette smokers exhibit elevations of the peripheral blood eosinophil count (Taylor, Gross et al. 1985), although it is unknown if allergen-induced and cigarette smoking-induced eosinophilia occur by similar or different mechanisms. Eosinophils in peripheral blood are also related to clinical correlates of emphysema (Nagai, West, Thurlbeck 1985). Cigarette smoking has also been associated with increased levels of airway respon- siveness (Woolcock et al. 1987; Sparrow et al. 1987; Burney et al. 1987). Several mechanisms could explain the relationship between cigarette smoking and increased airway responsiveness, including smoking-associated reduction in prechallenge level of lung function, chronic airway inflammation due to smoking, and smoking-induced impairment of epithelial function. The potential central role of cigarette smoking in parenchymal and airways inflammation is depicted in Figure 3. BRONCHIOLAR ~' NARROWING AIRWAY INFLAMMATION ClGARETTE Z SMOKING , HYPERRESPONSIVENESS ALVEOLAR 1NFLAMMATION EMPHYSEMA FIGURE 3.-Hypothesized mechanisms by which airway hyperresponsiveness may be associated with developing or established COPD without necessarily being a preexisting risk factor NOTE: COPD=chronic obstructive pulmonary disease. SOURCE: O'Connor. Sparrow. Weiss (1989). When considered in this pathophysiologic framework, the potential consequences of smokinQ cessation on the degree of impairment and future risk of COPD vary with the extent of irreversible changes at cessation and with host characteristics of the quitting smoker. In adults, ciQarette smoking cessation is associated with a siowina of FEV 1 decline to the rate of never smokers (Figure 2). To the extent that airwav and alveolar inflammation have caused reversible epithelial and parenchymal inflammation, pul- monary function could improve after cessation, particularly if heightened airway responsiveness and bronchiolitis can resolve. To the extent that cigarette smoking has caused permanent damage to lung structure (e.g., emphysema), those changes are AIRWAY 284 TIMN 438679

INTRODUCTION Obstructive airways diseases constitute a heterogeneous group of disorders that include but are not limited to emphysema, asthma, chronic bronchitis, and chronic obstructive pulmonary disease (COPD). These four clinical conditions are the most prevalent of the obstructive airways diseases and are responsible for substantial mor- bidity and mortality. Over 18 million Americans suffer from asthma, and about 12 million Americans have COPD, which is the fifth leadin8 cause of death and the most rapidly increasing cause of death among adults older than 65 years (Feinleib et al. 1989). The 1984 Report on the health consequences of smoking reviewed information on chronic obstructive lung diseases (US DHHS 1984). The Report concluded that "ci?arette smoking is the major cause of chronic obstructive lung disease in the United States for both men and women. The contribution of cigarette smoking to chronic obstructive lung disease morbidity and mortality far outweighs all other factors" (US DHHS 1984, p. 8). Approximately 84 percent of COPD mortality among men and 79 percent among women is attributable to cigarette smoking (US DHHS 1989). The annual toll of smokinor-attributable COPD in the United States is estimated to be 57,000 deaths (US DHHS 1989), which are responsible for more than 500,000 years of potential life lost before the average life expectancy (Davis and Novotny 1989). The nosology of obstructive airways diseases has been evolving since the CIBA Foundation Guest Symposium in 1959. one of the first attempts to create a standardized classification. For the purposes of this Chapter. emphysema refers to pathologic abnormal permanent enlargement of the airspaces distal to the terminal bronchiole, accompanied by destruction of airspace walls and without obvious fibrosis (American Thoracic Society 1987). Chronic bronchitis refers to chronic cough and/or sputum production for at least 3 months per year for 2 consecutive years. Asthma has been defined as "a disease characterized by increased responsiveness of the airways to various stimuli and manifested by slowina down of forced expiration. which changes in severity either spontaneously or as a result of therapy" (American College of Chest Physicians, American Thoracic Society Joint Statement 1975). The term COPD is used to describe persistent obstructive ventilatory impairment as determined by a test of pulmonary'ventilatory function (O'Connor, Sparrow, Weiss 1989). Overlap of these conditions is extremely common. although discrete cases of each can be identified (Figure 1). It is estimated that 60 to 100 percent of COPD patients also have airways hyperresponsiveness (Klein and Salvaocrio 1966: Parker, Bilbo, Reed 1965; Ramsdell. Nachtwey, Moser 1982: Ramsdale et al. 1984; Bahous et al. 1984). Almost one-half of all asthmatics suffer from chronic bronchitis (Burrows et al. 1987), and asthma may be a risk factor for the development of chronic airflow obstruction (Fletcher et ai. 1976: Schachter, Doyle, Beck 1984; Buist and Vollmer 1987; Peat. Woolcock, Cullen 1987). Although the extent of emphysema, as documented by postmortem examination of the lungs. correlates significantly with the degree of fixed airflow obstruction, the correlation is modest, suggesting that emphysema alone does not fully explain the functional impairment in most persons with COPD (Cosio et al. 1977). TIMN 438674 279

CEDERLOF, R., FRIBERG, L., HRUBEC, Z., LORICH, U. The relationship of smoking and some social covariables to mortality and cancer morbidity. A ten year follow-up in a probability sample of 55,000 Swedish subjects age 18-69, Part 1/2. Stockholm, Sweden: Karolinska Institute, Department of Environmental Hygiene, 1975. CLARKSON, T.B.. BOND, M.G., BULLOCK. B.C., MCLAUGHLIN, K.J., SAWYER, J.K. A study of atherosclerosis regression in Macaca mulatta. V. Changes in abdominal aorta and carotid and coronary arteries from animals with atherosclerosis induced for 38 months and then regressed for 24 to 48 months at plasma cholesterol concentrations of 300 or 200 mg/dL. Experimental and Molecular Pathology 41(1):96-118, August 1984. COLDITZ, G.A., BONITA, R., STAMPFER, MJ., WILLETT, W.C., ROSNER, B., SPEIZER, F.E., HENNEKENS, C.H. Cigarette smoking and risk of stroke in middle-aged women. New England Journal of Medicine 318(15):937-941, April 14, 1988. COLLABORATIVE GROUP FOR THE STUDY OF STROKE IN YOUNG WOMEN. Oral contraceptives and stroke in young women. Associated risk factors. Journal of the American Medical Association 231(7):718-722, February 17, 1975. COOK, D.G., POCOCK. S.J:, SHAPER, A.G., KUSSICK, S.J. Giving up smoking and the risk of heart attacks. A report from the British Regional Heart Study. Lancet 2(8520):1376-1380, December 13. 1986. CRIQUI, M.H., BROWNER, D., FRONEK, A., KLAUBER, M.R.. COUGHLIN, S.S., BAR- RETT-CONNOR, E., GABRIEL, S. Peripheral arterial disease in large vessels is epidemiologically distinct from small vessel disease. An analysis of risk factors. American Journal of Epidemiology 129(6):1110-1119, June 1989. CRONENWETT. J.L., WARNER, K.G., ZELENOCK, G.B., WHITEHOUSE. W.M. JR., GRAHAM. L.M.. LINDENAUER, S.M.. STANLEY, J.C. Intermittent claudication: Current results of nonoperative management. Archives of Surgery 119(4):430-436, April 1984. DALY, L.E.. GRAHAM. I.M.. HICKEY, N., MULCAHY, R. Does stopping smoking delay onset of angina after infarction? British Medical Journal 291(6500):935-937, October 5, 1985. DALY. L.E., HICKEY, N., GRAHAM, I.M., MULCAHY. R. Predictors of sudden death up to 18 years after a first attack of unstable angina or myocardial infarction. British Heart Journal 58(6):567-571, December 1987. DALY, L.E., MULCAHY, R., GRAHAM. LM., HICKEY, N. Long term effect on mortality of stopping smoking after unstable ancrina and myocardial infarction. British Medical Journal 287(6388):324-326, July 30, 1983. DAVIS, J.W.. SHELTON, L., EIGENBERG, D.A.. HIGNITE. C.E., WATANABE. I.S.. Effects of tobacco and non-tobacco cigarette smoking on endothelium and platelets. Clinical Pharmacology and Therapeutics 37(5):529-533, May 1985. DAVIS, J.W.. SHELTON, L., HARTMAN, C.R., ET AL. Smoking-induced changes in endothelium and platelets are not affected by hydroxyethylrutosides. British Journal of Experimental Pathology 67(5):765-771, October 1986. DE LORGERIL. M., REINHARZ, A., BUSSLINGER, B., REBER. G.. RIGHETTI, A. Acute influence of cigarette smoke in platelets, catecholamines and neurophysins in the normal conditions of daily life. European Heart Journal 6(12):1063-1068, December 1985. DE PARSCAU. L., FIELDING, C. Abnormal plasma cholesterol metabolism in cigarette smokers. Metabolism 35(11):1070-1073, November 1986. DEWOOD, M.A., SPORES, J., NOTSKE, R.. MOUSER. L.T., BURROUGHS, R., GOLDEN, M.S., LANG, H.T. Prevalence of total coronary occlusion during the early hours of transmural myocardial infarction. New England Journal of Medicine 303(16):897-902, October 16, 1980. 262 TIMN 43866

MEN 100 90 80 70 60 50 40 30 20 10 L I 30 i 40 1 50 i 60 10 20 AGE WOMEN x ¢ xw Ha 100 90 80 70 60 50 40 30 20 10 ------------- •"' ---------- ~ --<--------- 1 I I 1 1 70 , 10 20 30 40 50 60 70 AGE Cigarette Smokers - -•-•-- Ex-Smokers ......... Pipe & Cigar Smokers -------- --• Never Smokers FIGURE 6.-Prevalence of dyspnea by smoking group SOURCE: Payne and Kjelsberg (196•3). 295 TIMN 438690

FORTMANN, S.P., HASKELL, W.L., WILLIAMS, P.T. Changes in plasma high density lipoprotein cholesterol after changes in cigarette use. American Journal of Epidemiology 124(4):706-710, October 1986. FREEDMAN, D.S., SRINIVASAN, S.R., SHEAR, C.L., HUNTER, S.M., CROFT, J.B., WEBBER, L.S., BERENSON, G.S. Cigarette smoking initiation and longitudinal changes in serum lipids and lipoproteins in early adulthood: The Bogalusa Heart Study. American Journal of Epidemiology 124(2):207-219, August 1986. FREEDMAN, D.S., SRINIVASAN, S.R., SHEAR, C.L., WEBBER, L.S., CHIANG, Y.K., BERENSON, G.S. Correlates of high density lipoprotein cholesterol and apolipoprotein A-I levels in children. The Bogalusa Heart Study. Arteriosclerosis 7(4):354-360, July-August 1987. FRIED, L.P., MOORE, R.D., PEARSON, T.A. Long-term effects of cigarette smoking and moderate alcohol consumption on coronary artery diameter. Mechanisms of coronary artery disease independent of atherosclerosis or thrombosis? American Journal of Medicine 80(1):37-44, January 1986. FRIEDMAN, G.D., DALES, L.G., URY, H.K. Mortality in middle-aged smokers and non- smokers. New England Journal of Medicine 300(5):213-217, February 1, 1979. FRIEDMAN, G.D., PETITTI, D.B., BAWOL. R.D., SIEGELAUB, A.B. Mortality in cigarette smokers and quitters. Effect of base-line differences. New England Journal of Medicine 304(23):1407-1410. June 4, 1981. FRIEDMAN. G.D., SIEGELAUB, A.B., DALES, L.G., SELTZER. C.C. Characteristics predictive of coronary heart disease in ex-smokers before they stopped smoking: Comparison with persistent smokers and nonsmokers. Journal of Chronic Diseases 32:175-190, 1979. FULLER, J.H., SHIPLEY, M.J., ROSE, G.. JARRETT, R.J., KEEN, H. Mortality from coronary heart disease and stroke in relation to degree of glycaemia: The Whitehall Study. British Medical Journal 287(6396):867-870, September 24, 1983. FUSTER, V., CHESEBRO. J.H., FRYE, R.L., ELVEBACK, L.R. Platelet survival and the development of coronary artery disease in the young adult: Effects of cigarette smoking, strong family history and medical therapy. Circulation 63(3):546-551, Mafch 1981. GALAN, K.M., DELIGONUL. U., KERN. M.J.. CHAITMAN, B.R., VANDORMAEL, M.G. Increased frequency of restenosis in patients continuing to smoke cioarettes after percutaneous transluminal coronary angioplasty. American Journal of Cardiolo-y 61:260-263. 1988. GALEA, G., DAVIDSON, R.J.L. Haematological and haemorheoloeical changes associated with cigarette smoking. Journal of Clinical Pathology 38(9):978-984, September 1985. GARFINKEL, L., STELLMAN, S.D. Smoking and lung cancer in women: Findings in a prospective study. Cancer Research 48(23):6951-6955, December 1, 1988. GOLDBERG, R., SZKLO. M., CHANDRA, V. The effect of cigarette smoking on the long-term prognosis of myocardial infarction. American Journal of Epidemiology 114(3):431, Septem- ber 1981. GOLDBOURT, U.. YAARI, S., COHEN-MANDELZWEIG, L., NEUFELD. H.N. High-den- sity lipoprotein cholesterol: Correlation with biochemical, anthropometric. behavioral, and clinical parameters in 6.500 Israeli men. Preventive Medicine 15(6):569-581, November 1986. GOMO, Z.A.R. The effect of age, sex, alcohol consumption and cigarette smoking on serum concentrations of lipids and apolipoproteins in Zimbabwean blacks. Atherosclerosis 61(2): 149-154, August 1986. GORDON, T., DOYLE, J.T. Alcohol consumption and its relationship to smoking, wei~ht, blood pressure, and blood lipids. The Albany Study. Archives of Internal Medicine 146(2): 262-265, February 1986. GORDON, T., KANNEL, W.B., MCGEE, D. Death and coronary attacks in men after giving up cigarette smoking. Lancet 2(7893):13ti5-1348, December 7, 1974. 264 TIMN 438662

symptoms, by cigarette smoking status and gender Current smokers Former smokers Never smokers Age Symptom? (number of Reference subjects) Male (%) Female (%) Male (%) Female (%) Male (%) Female (%) Cough 3 mo/yr Wilhelmsen 50(339) 36.2 - 8.2 - 4.8 and Tibblin (1966) Weiss et al. 50-69 (287) 41.0 - 9.0 - 11.0 (1963) Fletcher and 40-59 (363) 19.9 - 13.0 - 0.0 Tinker(1961) Mueller et al. 20-69 (892) 13.0 20.0 5.0 10.0 9.0 5.0 (1971)b Manfreda, Nelson. Cherniack (1978) 25-54 (256)c 25.4 20.3 8.1 - 8.3 - 25-54 (246)d 31.5 31.7 2.9 10.0 4.0 4.0 Schenker, 17-74 (5,670) - 9.l` - 7.5 - 5.6 Samet, Speizer 17.Of (1982)b 31.81 Phlegm 3 mo/yr Wilhelmsen and 11.5 - 1.4 - 1.2 Tibblin (1966) Fletcher and 17.6 - 16.9 - 7.5 Tinker(1961) Mueller et al. 18.0 10.0 12.0 5.0 4.0 1.0 (1971) b Manfreda. Nelson, Chemiack (1978) 25-54 (256)c 16.9 10.2 10.8 - 0.0 0.0 25-54 (246)d 24.7 25.4 5.7 5.0 4.0 4.0 Hawthorne 45-64 36.2 23.0 16.1 10.9 10.1 6.7 and Fry (1978) Miller et al. Male (mean): 40.8 28.4 14.7 6.9 12.1 0.4 (1988)h 42.0 (1.169) Female (mean): 42.9 (1.169) 289 TIMN 4386,84

KANNEL, W.B., D'AGOSTINO, R.B., BELANGER, A.J. Fibrinogen, cigarette smoking, and risk of cardiovascular disease: Insights from the Framingham Study. American HeartJournal 113(4):1006-1010, April 1987. KANNEL, W.B., MCGEE, D., GORDON, T. A general cardiovascular risk profile: The Framingham Study. American Journal of Cardiology 38(l):46-51, July 1976. KEYS, A. Smoking habits. In: Keys, A. (ed.) Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, Massachusetts: Harvard University Press, 1980, pp. 136-160. KHAW, K.-T., BARRETT-CONNOR, E., SUAREZ, L., CRIQUI, M.H. Predictors of stroke- associated mortality in the elderly. Stroke 15(2):244-248, March-April 1984. KLEIN, L.W. Cigarette smoking, atherosclerosis and the coronary hemodynamic response: A unifying hypothesis. Journal of the American College of Cardiology 4(5):972-974, Novem- ber 1984. KONO, S., IKEDA, M., TOKUDOME, S., NISHIZUMI, M., KURATSUNE, M. Smoking and mortalities from cancer, coronary heart disease and stroke in male Japanese physicians. Journal of Cancer Research and Clinical Oncology 110(2):161-164, 1985. KORNITZER, M., DE BACKER, G., DRAMAIX, M., KITTEL, F., THILLY, C., GRAFFAR, M., VUYLSTEEK, K. Belgian Heart Disease Prevention Project: Incidence and mortality results. Lancet 1(8333):1066-1070, May 14, 1983. KRAMER, J.R., KITAZUME, H., PROUDFIT, W.L., MATSUDA, Y., WILLIAMS, G.W., SONES, F.M. JR. Progression and regression of coronary atherosclerosis: Relation to risk factors. American Heart Journal 105(l):134-144, January 1983. KRUPSKI, W.C., OLIVE, G.C., WEBER, C.A., RAPP, J.H. Comparative effects of hyperten- sion and nicotine on injury-induced myointimal thickening. Surgery 102(2):409-415, August 1987. KULLER, L., MEILAHN, E., TOWNSEND, M., WEINBERG, G. Control of cigarette smoking from a medical perspective. Annual Review of Public Health 3:153-178, 1982. LAVECCHIA, C., FRANCESCHI, S., DECARLI. A., PAMPALLONA, S., TOGNONI, G.. Risk factors for myocardial infarction in young women. American Journal of Epidemiology 125(5):832-843, May 1987. LANNERSTAD, 0., ISACSSON, S.-O.. LINDELL, S.-E. Risk factors for premature death in men 56-60 years old. A prospective study of men bom in 1914, living in Malmo, Sweden. Scandinavian Journal of Social Medicine 7(1):41~7, 1979. living LAYDE, P.M., BERAL, J., KAY, C.R. Further analyses of mortality in oral contraceptive users. Royal College of General Practitioners' oral contraceptive study. Lancet 1(8219):541-546, March 7, 1981. LOSCALZO, J., BRAUNWALD, E. Tissue plasminocren activator. New England Journal of Medicine 319(14):925-931, October 6, 1988. LUSBY, R.J., BAUMINGER, B., WALTERS, G., DAVIES, P.W., SKIDMORE, R.. BAIRD, R.N. Cigarette smoking induced vasoconstriction in habitual smokers with and without arterial occlusive disease. In: Greenhalgh, R.M. (ed.) Smoking and Arterial Disease. Woodstock, New York: Beekman Publications, Inc., August 1981, pp. 218-225. MADSEN, H., DYERBERG, J. CiQarette smoking and its effects on the platelet-vessel wall interaction. Scandinavian Journal of Clinical Laboratory Investigation 44(3):203-206, May 1984. MALINOW, M.R., BLATON, V. Regression of atherosclerotic lesions. Arteriosclerosis 4(3):292-295, May-June 1984. MAOUAD, J., FERNANDEZ, F., BARRILLON, A.. GERBAUX, A., GAY, J. Diffuse or segmental narrowing (spasm) of the coronary arteries during smoking demonstrated on angiography. American Journal of Cardiology 53(2):354-355, January 15, 1984. TIMN 438665 26-7

DOLL, R., GRAY, R., HAFNER, B., PETO, R. Mortality in relation to smoking: 22 years' observations on female British doctors. British Medical Journa1280(6219):967-971, April 5, 1980. DOLL, R., HILL, A.B. The mortality of doctors in relation to their smoking habits. A preliminary report. British Medical Journal 1(4877):1451-1455, June 26, 1954. DOLL, R., HILL, A.B. Lung cancer and other causes of death in relation to smoking. A second report on the mortality of British doctors. British Medical Journa12:1071-1081, November 10, 1956. DOLL, R., HILL, A.B. Mortality in relation to smoking: Ten years' observations of British doctors. British Medical Journal 1:1460-1467, June 6, 1964. DOLL, R., PETO, R. Mortality in relation to smoking: 20 years' observations on male British doctors. British Medical Journal 2:1525-1536, December 25, 1976. DONNAN, G.A., ADENA, M.A., O'MALLEY, H.M., MCNEIL, J.S., DOYLE, A.E., NEILL, G.C. Smoking as a risk factor for cerebral ischaemia. Lancet 2(8664):643-647, September 16, 1989. DORN, H.F. Tobacco consumption and mortality from cancer and other diseases. Public Health Reports 74(7):581-593, July 1959. DOTEVALL. A., KUTTI, J., TEGER-NILSSON, A.C., WADENVIK. H.. WILHELMSEN, L. Platelet reactivity, fibrinogen and smoking. European Journal of Haematoloqy 38( l):55-59, January 1987. DOWNEY. H.F., BASHOUR, C.A.. BOUTROS, I.S., BASHOUR. F.A.. PARKER, P.E. Regional myocardial blood flow during nicotine infusion: Effects of beta adrenergic blockade and acute coronary artery occlusion. Journal of Pharmacology and E.rperimental Therapeutics 202(1):55-68. July 1977. DOYLE, J.T.. DAWBER. T.R., KANNEL, W.B., HESLIN, A.S., KAHN, H.A. Cigarette smoking and coronary heart disease. Combined experience of the Albany and Framinsham studies. New England Journal of Medicine 266( l 6):796-801, April 19, 1962. ~ DOYLE. J.T.. DAWBER, T.R.. KANNEL. W.B., KINCH, S.H.. KAHN, H.A. The relationship of cigarette. smoking to coronary heart disease. Journal of the American iLledical Association 190( l0):108-112, December 7, 1964. ERNST, E., MATRAI, A. Abstention from chronic cigarette smokino normalizes blood rheology. Atherosclerosis 64(1):75-77, March 1987. FAULKNER, K.W., HOUSE, A.K.. CASTLEDEN, W.M. The effect of cessation of smoking on the accumulative survival rates of patients with symptomatic peripheral vascular disease. Medical Journal of Australia 1(5):217-219, March 5, 1983. FISCHER, S., BERNUTZ, C., MEIER, H., WEBER, P.C. Formation of prostacyclin and thromboxane in man as measured by the main urinary metabolites. Biochimica et Biophysica Acta 876(2):194-199, April 15, 1986. FITZGERALD, G.A.. OATES, J.A., NOWAK, J. Cigarette smoking and hemostatic function. American Heart Journal 115:267-271, 1988. FLECK, E., REGITZ, V., LEHNERT, A., DACIAN, S., DIRSCHINGER, J., RUDOLPH, W. Restenosis after balloon dilatation of coronary stenosis, multivariate analysis of potential risk factors. European Heart Journal 9(Supplement C):15-18, March 1988. FLODERUS, B., CEDERLOF, R., FRIBERG, L. Smoking and mortality: A 21-year follow-up based on the Swedish Twin Registry. International Journal of Epidemiology 17(2):332-340, June 1988. FOLTS, J.D., BONEBRAKE, F.C. The effects of cigarette smoke and nicotine on platelet thrombus formation in stenosed dog coronary arteries: Inhibition with phentolamine. Cir- culation 65(3):465-470, March 1982. TIMN 438661 263

with regard to smoking. Cigarette smoking is associated with low levels of 1-sec forced expiratory volume (FEVI) in cross-sectional investigations (Knudson, Burrows, Lebowitz 1976; Burrows et al. 1977; Beck, Doyle, Schachter 1981; Dockery et al. 1988; US DHHS 1984), with accelerated decline of FEV I in longitudinal studies (Burrows et al. 1987; Beck, Doyle, Schachter 1982; Bosse et al. 1981; US DHHS 1984), and with increased mortality from COPD (Best 1966; Doll and Peto 1976; Hammond 1965; Hammond and Horn 1958; US DHHS 1984). The effects of cigarette smoking on lung function level or rate of decline and on mortality increase with the duration and amount of smoking (US DHHS 1984). Because the development of COPD in adults is associated with a long latent period, the age at which cigarette smoking might have a critical effect has not readily been addressed. Passive smoking impairs lung growth in children and thus, may limit maximal lung growth (Tager et al. 1983; US DHHS 1986). Smoking in adults may shorten the phase when lung function tends to plateau between the ages of 20 and 40 and/or may accelerate the decline in lung function (Tager et al. 1988). Cigarette smoking is the predominant cause of lung function decline at a rate greater than the annual volume loss of 20 to 30 mL associated with aging. Although cigarette smokina has been clearly established as the major risk factor for COPD, the interactions of the intensity of smoking with factors determining suscep- tibility have not been fully characterized. For example, Burrows and coworkers (1987) suggested that two subsets of COPD patients can be differentiated by the presence or absence of accompanying asthmatic features. According to this hypothesis, subjects with chronic asthmatic bronchitis have a better long-term prognosis, smaller cumulative exposure to tobacco smoke,, and greater prevalence of allergy and airway responsive- ness. The second group of patients has emphysema, poorer long-term prognosis. greater cumulative tobacco smoke exposure. and reduced prevalence of allergy and airway hyperresponsiveness (Burrows et al. 1987). Available data do not discriminate the relative contributions of cigarette smoking in these clinical subtypes of patients. Studies of the mechanisms~by which ciQarette smoking causes luna injury were reviewed extensively in the 1984 Report of the Surgeon General (US DHHS 1984). That Report and other reviews (Thurlbeck 1976; Snider 1989; Wright 1989) also cover the relationship between the structural changes associated with smoking and the severity of airflow obstruction. Cigarette smoking causes inflammation of both the airways and parenchyma of the lunQ; the resulting structural damage has functional consequences that can lead to the development of clinically diagnosed COPD if there is sustained smoking. Frank parenchymal damage is preceded by an increase in inflammatory cells in lunar parenchyma at the level of the bronchioli (Niewoehner, Kleinerman, Rice 1974). Both neutrophils and alveolar macrophages are important in the development of this inflammatory bronchiolitis. Although neutrophils store and release greater quantities of elastase than alveolar macrophages (Janoff et al. 1979). the macrophage may be an important cell in attracting neutrophils to the lunQ (Hunninghake and Crystal 1983). Cigarette smokinQ-induced bronchiolitis is associated with func- tional abnormalities detectable in the early stages only with sensitive tests of small airway function (Buist et al. 1979: Cosio et al. 1977;McCarthy, Craig, Cherniack 1976: Ingram and Schilder 1967; Ingram and O'Cain 1971). Even before significant em- 282 TIMN 438677

HOLME, I. On the separation of the intervention effects of diet and antismoking advice on the incidence of major coronary events in coronary high risk men. The Oslo Study. Journal of the Oslo City Hospitals 32(3/4):31-54, March-April 1982. HOLME, I., HELGELAND, A., HJERMANN, I., LEREN, P., LUND-LARSEN, P.G. Four and two-thirds years incidence of coronary heart disease in middle-aged men: The Oslo Study. American Journal of Epidemiology 112( l):149-160, July 1980. ~ HUBERT, H.B., EAKER, E.D., GARRISON, R.J., CASTELLI, W.P. Life-style correlates of risk factor change in young adults: An eight-year study of coronary heart disease risk factors in the Framingham offspring. American Journal of Epidemiology 125(5):812-831, May 1987. HUBERT, H.B.. HOLFORD. T.R., KANNEL, W.B. Clinical characteristics and cigarette smoking in relation to prognosis of angina pectoris in Framingham. American Journal of Epidemiology 115(2):231-242, February 1982. HUGHES, G.H., HYMOWITZ, N., OCKENE, J.K., SIMON, N., VOGT, T.M. The Multiple Risk Factor Intervention Trial (MRFIT). V. Intervention on smoking. Preventive Medicine 10(4):476-500, July 1981. HUGHSON, W.G., MANN, J.I., GARROD. A. Intermittent claudication: Prevalence and risk factors. British Medical Journal 1(6124):1379-1381, May 27, 1978. HUGHSON, W.G., MANN, J.I., TIBBS, D.J., WOODS. H.F., WALTON, I. Intermittent claudication: Factors determininc, outcome. British Medical Journal 1(6124):1377-1379, May 27, 1978. ~ HULLEY, S.B.. COHEN, R., WIDDOWSON. G. Plasma high-density lipoprotein cholesterol level. Influence of risk factor intervention. Journal of the American Medical Association 238(21):2269-227 I, November 21, 1977. JACOBSEN, B.K., THELLE. D.S. The Tromso Heart Study: Food habits, serum total cholesterol. HDL cholesterol, and triglycerides. American Journal of Epidemiology 125(4):622-630, April 1987. JAJICH, C.L., OSTFELD. A.M., FREEMAN. D.H. JR. Smoking and coronary heart disease mortality in the elderly. Journal of the American Medical Association 252(20):2831-2834, November 23-30. 1984. JENKINS, C.D.. ROSENMAN, R.H., ZYZANSKI. S.J. Cigarette smoking: Its relationship to coronary heart disease and related risk factors in the Western Collaborative Group Study. Circulation 38(6):l 140-11». December 1968. JOHANSSON. S... BERGSTRAND, R., PENNERT. K.. ULVENSTAM, G., VEDIN. A.. WEDEL. H., WILHELMSSON, C., WILHELMSEN. L., ABERG, A. Cessation of smoking after myocardial infarction in women. Effects on mortality and reinfarctions. American Journal of Epidemiology 121(6):823-831. June 1985. , JONASON, T., BERGSTROM. R. Cessation of smoking in patients with intermittent claudi- cation: Effects on the risk of peripheral vascular complications. myocardial infarction and mortality. Acta Medica Scandinavica 221:253-260, 1987. JONASON, T.. RINGGVIST, I. Factors of prognostic importance for subsequent rest pain in patients with intermittent claudication. Acta Medica Scandinavica 218(1):27-33, 1985. JUERGENS. J.L.. BARKER. N.W., HINES, E.A. JR. Arteriosclerosis obliterans: Review of 520 cases with special reference to pathogenic and prognostic factors. Circulation 21(2):188- 195, February 1960. KAHN, H.A. The Dom study of smoking and mortality among U.S. veterans: Report on eight and one-half years of observation. In: Haenszel, W. (ed.) Epidemiological Approaches to the Study of Cancer and Other Chronic Diseases. NCI Monograph No. 19. U.S. Department of Health, Education, and Welfare, U.S. Public Health Service, National Cancer Institute. January 1966, pp. 1-125. 266 TIMN 438664

unlikely to be reversible. Thus, the amount and duration of smoking, the relative extents of parenchymal and airway inflammation, and the degree of permanent structural damage are probably the key determinants of the level of function after smoking cessation. Even in the setting of established COPD, smoking cessation may potentially reduce the rate of functional loss. Former smokers may differ from continuing smokers with regard to host charac- teristics that potentially determine susceptibility to cigarette smoke. Because presmok- ing levels of atopy and airway responsiveness modify the short-term response to smoke. individuals with atopy or heightened airway responsiveness may be less likely to take up smoking, to reduce smoking, or to quit smoking if respiratory symptoms occur. This potential bias, termed the "healthy smoker effect" by O'Connor, Sparrow, and Weiss (1989), cannot be evaluated in cross-sectional studies. PART I. SMOKING CESSATION AND RESPIRATORY MORBIDITY Respiratory Symptoms Since the 1950s, strong evidence has accumulated documentino, increased respiratory symptoms in smokers of all ages compared with nonsmokers (US PHS 1964: US DHEW 1971, 1979; US DHHS 1984). Further, the number of cigarettes smoked per day is the strongest risk factor for the principal chronic respiratory symptoms including chronic cough, phlegm production, wheeze, and dyspnea (Lebowitz and Burrows 1977; Dean et al. 1978; Higgins, Keller, Metzner 1977; Huhti and Ikkala 1980; Higenbottam et al. 1980; Schenker, Samet, Speizer 1982). The widespread effects of chronic smoking on the luncr, includina decreased tracheal mucous velocity (Lourengo, Klimek. Borowski 1971; Goodman et al. 1978: Thomson and Pavia 1973), increased secretion of mucus on the basis of mucous cyland hypertrophy and hyperplasia (Thurlbeck 1976), chronic airway inflammation (Niewoehner, Kleinerman, Rice 1974), increased epithelial permeability (Jones et al. 1980: Minty, Jordon, Jones 1981; Mason et al. 1983), and emphysema (US DHHS 1984), underlie the development of these symptoms. Smoking cessation has been associated with a.reduction in respiratory morbidity, presumably through reversal of some of these pathophysiologic abnor- malities. Relevant evidence can be found in clinical studies, which involve followup of the symptoms of persons participating in smoking cessation clinics, and epidemioloQic studies. ~ Clinical Studies Buist and coworkers (1976) found that smoking cessation was associated with a dramatic reduction in respiratory symptoms within 1 month of cessation. These researchers assessed spirometry and respiratory symptoms for over 12 months in 75 cigarette smokers enrolled in a smoking cessation program. Subjects were divided into quitters (those who did not smoke during the entire 12-month period), modifiers (individuals who reduced their cioarette consumption by 25 percent), and nonmodifiers 285 TIMN 438680

followup. these investigators reported mortality ratios close to 1.0 for all durations except for 5 to 9 years after quitting. Based on 26 years of studying 4.255 men and women in the Framingham Study (Wolf et al. 1988). the risk of stroke among persons who stopped was significantly lower than that among persons who continued to smoke cigarettes. Furthermore, persons who quit smoking developed stroke at the rate of never smokers soon after discontinuing cigarette smoking (Figure 8). Wolf and coworkers (1988) estimated that the risk of stroke among smokers had decreased significantly 2 years after quitting and reverted to the level of never smokers within 5 years. These results persisted after controlling for age, blood pressure, serum cholestrol level, relative weight, left ventricular hyper- trophy on electrocardiogram, and blood glucose level. Thus, the reduction in risk after smoking cessation is not attributable to differences in other risk factors for stroke between those who quit and those who continue to smoke. In the Nurses Health Study (Colditz et al. 1988), a lower risk of stroke was observed with increasing time from cessation. Compared with the risk amonQ never smokers, the relative risk was 2.6 amonc, women who had stopped for less than 2 years (95-percent CI, 1.4-4.7). However, among women who had stopped for 2 years or more, the relative risk'was reduced to 1.4 (95-percent CI, 1.0-2.0). Women currently smoking 15 to 24 cigarettes per day had a relative risk of 2.9 compared with never smokers. Again, the elevation of the relative risk during the first 2 years after cessation is consistent with high recidivism among these women. Prospective data from ACS CPS-II showed that among men who quit smokinQ, the risk of stroke returned to that of never smokers after 11 years or more of smoking abstinence for those originally smoking fewer than 21 cigarettes per day. However, for men who previously smoked 21 cigarettes or more per day, the risk among former smokers did not return to the level of never smokers, even after 16 years or more of cessation. Among women who quit, the rate of decrease was much more rapid; by 3 to 5 years after cessation, the risk of stroke was similar to that of never smol4ers (Table 8). Oral Contraceptives and Smoking Cessation In two studies the risk of subarachnoid hemorrhage was augmented among cigarette smokers who also take oral contraceptives (Petitti and WinQerd 1978; Collaborative Group for the Study of Stroke in Young Women 1975). In the Collaborative Group Study of stroke among young women (1975), the cateQory of former smokers was not clearly defined: rather, a group of "once regular smokers" was compared with "never regular smokers." In this study there was no association between current smoking or former smoking and risk of thrombotic stroke. Overall, the relative risk for hemorrhagic stroke was 1.8 among once reQular smokers and 3.3 among current smokers. Within the group of once regular smokers, women currently using oral contraceptives had approximately twice the risk compared with women not usinar oral contraceptives. The Royal College of General Practitoners study of oral contraceptives did not separate former smokers from never smokers (Layde. Beral, Kay 1981). Hence, data to address the relationship among oral contraceptives, smoking cessation, and risk of subarachnoid 258 TIMN 438656

WILHELMSEN, L., SVARDSUDD. K.. KORSAN-BENGTSEN. K.. LARSSON, B., WELIN, L.. TIBBLIN, G. Fibrinogen as a risk factor for stroke and myocardial infarction. New England Jnurnal of Medicine 311(8):501-505, August 23, 1984. WILHELMSSON, C.. VEDIN. J.A., ELMFELDT.~D., TIBBLIN. G., WILHELMSEN. L. Smoking and myocardial infarction. Lancet 1(7904):415-420. February 22, 1975. WILLETT, W.. HENNEKENS. C.H., CASTELLI, W.P.. ROSNER. B.. EVANS. D., TAYLOR. J.. KASS. E.H. Effects of cigarette smoking on fasting trislyceride, total cholesterol, and HDL-cholesterol in women. American HeartJoicrnal 105(3):417-421, March 1983. WILLETT. W.C.. HENNEKENS, C.H.. BAIN, C.. ROSNER. B.. SPEIZER, F.E. Cigarette smoking and non-fatal myocardial infarction in women. American Journal of Epidemiology 113(5):575-582. May 1981. WILLETT. W.C.. GREEN, A., STAMPFER. M.J.. SPEIZER, F.E., COLDITZ. G.A.. ROSNER. B.. MONSON. R.R.. STASON. W., HENNEKENS, C.H. Relative and absolute excess risks of coronary heart disease among women who smoke cigarettes. New England Journal of Medicine 317(21):1303-1309, November 19, 1987. WINNIFORD, M.D.. WHEELAN, K.R., KREMERS. M.S., UGOLINI. V.. VAN DEN BERG, E. JR.. NIGGEMANN, E.H.. JANSEN, D.E., HILLIS. L.D. Smokins-induced coronary vasoconstriction in patients with atherosclerotic coronary artery disease: Evidence for adrenergically mediated alterations in coronary artery tone. Circulation 73(4):662-667. April 1986. WISEMAN. S.. KENCHINGTON. G.. DAIN, R.. MARSHALL. C.E., MCCOLLUM, C.N.. GREENHALGH. R.M., POWELL. J.T. Influence of smokina and plasma factors on patency of femoropopliteal vein grafts. British Medical Jnurna1299:643-646, September 9, 1989. W ITTEMAti.J.C.. WILLETT. W.C., STAMPFER. M.J.. COLDITZ. G.A.. KOK. F.1.. SACKS, F.M., SPEIZER, F.E.. ROSNER, B., HENNEKENS. C.H. Relation of moderate alcohol consumption and risk of systemic hypertension in women. American Journal offardinlo,t ti• 65(9):633-637. March 1. 1990. ~ WOLF. P.A.. D'AGOSTINO, R.B.. KANNEL. W.B.. BONITA. R.. BELANGER, A.J. Cigarette smokina as a risk factor for stroke: The Framinsham- Study. Journal of the American MedicalAssnciatinn 259(7):1025-1029, February 19,1988. WORLD HEALTH ORGANIZATION EUROPEAN COLLABORATIVE GROUP. Multi- factorial trial in the prevention of coronary heart disease. 3. Incidence and mortality results. European Heart Journul 4:141-147. 1983. ZIMMERMAN, M., MCGEACHIE. J. The effect of nicotine on aortic endothelium. A quantitative ultrastructural study. Atherosclerosis 63:33~ 1. 1987. TIMN 438671 273

TABLE 3.--Continued Continuing smokers Former smokers Never smokers Symptoms Age Reference (mean) Lost No changea Gained Lost No change' Gained Lost No changea Gained Comstock et al. Net change: 5.0 Net change: -5.0 Net change: -2.0 (1970)O1 Sharp et al. 13.4 77.0 9.6 11.1 78.7 10.2 7.3 88.4 4.3 (1973)"' 'No change indicates that respiratory symptoms were either consistently absent or consistently present. bOnly females, cough and/or phlegm, 5-yr study period. `l.ight=570cig/wk; moderate=71-140 cig/wk; heavy=more than 140 cig/wk. °Fonner smokers defined as those who stopped between baseline and followup. `Mules only, 5-6-yr followup. rMules only, fornier studies defined as those who stopped between baseline and followup, 7-yr followup. 'Fonner studies defined as those who stopped between baseline and followup, 1.5-yr followup. °ppd=packs/day. 'Grade 2 or 3 dyspnea. JDyspnea not defined. kDyspnea at ordinary pace. 'Wheeze not defined. mCVer wheeze.

TAtfLE 1.-l:ontinued Current smokers Former smokers Never smokers Symptoms' Reference Age (number of subjects) Male (%) Female (%) Male (%) Female Male (%) (%) Female (%) Schenker. Samet. 7.2e 6.7 4.5 Speizer (1982)b 16.7t 24.81 Lebowitz 14-96 (2,857) 11.2 11.0 25.9 12.6 45.5 35.8 and Burrows (1977) Dvsanea gmde, Wilhelmsen and 24.7 21.9 20.2 Tibblin (1966Y Weiss et at. (196V 44.0 46.0 36.0 Fletcher et al. (1959) Grades 2 or more 40-59 23.5 29.0 25.0 23.1 10.0 31.4 Fletcher and Tinker(1961) Grade 3 or more 8.7 6.5 2.5 Mueller et al. (1971) b Grade 2. 29.0 32.0 14.0 41.0 22.0 32.0 Grade 3. or more 7.0 13.0 4.0 11.0 6.0 7.0 Manfreda. Nelson. Cherniack (1978)k Grade 2 or more 25-54 (256)` 5.6 22.1 5.4 6.1 8.3 7.0 25-54 (246)d 12.3 17.5 5.8 5.0 4.0 12.0 Hawthorne and 13.2 18.6 9.9 20.5 7.0 13.2 Fry (1978}' Miller et al. (1988)h Grade 2 9.3 15.6 7.1 12.7 3.0 9.5 Grade 3 3.0 8.9 3.3 11.5 0.4 2.6 Schenker. Samet. Speizer (1982) b G rade 3 .e .2 .9 6:1 17.6e 290 T"IIV1N 438685

Summary of Observational Studies In a meta-analysis of cohort and case-control studies of cigarette smoking and stroke (Shinton and Beevers 1989), the overall relative risk of stroke among former smokers was 1.17 compared with never smokers (95-percent CI, 1.05-1.30). This estimate is based on a summary of 18 relative risks from 13 studies that separately identified former smokers (Kahn 1966; Doll and Peto 1976; Abbott et a1.1986; Colditz et al. 1988; Ostfeld et al. 1974; Kono et al. 1985; Khaw et al. 1984; Vessey, Lawless, Yeates 1984; Bell and Symon 1979; Bell and Ambrose 1982; Bonita et al. 1986; Bonita 1986; Taha, Ball, Illingworth 1982). As observed for the relation between current smoking and stroke, the risk among former smokers was greater when the analysis was repeated using only those studies with stroke occurring before age 75 (RR= 1.47, 95-percent CI, 1.15-1.88 compared with never smokers). By comparison, the relative risks for current smokers were 2.9 for those younger than 55 years and 1.8 for persons aged 55 to 74 years. Thus, although a modest elevation in risk persisted among younger former smokers, this relative risk was substantially less than that which was observed among current smokers. Intervention Studies Intervention trials described above provide little direct evidence relating to change in risk of stroke after smoking cessation. Only the trial of smoking cessation conducted among 1,445 British men used a single intervention (Rose et al. 1982). During 10 years of followup, five men in the normal care group died because of stroke, and seven men in the intervention group died because of stroke. The small numbers in each group and the small difference in smoking cessation rates between the intervention and control groups limit any conclusion reQardina the impact of smoking cessation in this popuia- tion. ~ ~ , Other intervention studies have included management of -hypertension and cholesterol as well as smoking cessation programs. As discussed under randomized trials of smoking cessation and CHD, these multiple interventions make drawing conclusions difficult regarding the relation between smoking cessation and risk of stroke (Steinbach et al. 1984; Wilhelmsen et al. 1986; MRFIT Research Group 1982, 1986; Salonen, Puska, Mustaniemi 1979; Hjermann 1980; Holme 1982). In a nonrandomized intervention, Rogers and colleagues (1985) measured changes in cerebral artery blood flow among volunteers who were encouraged to abstain from cigarettes. Cerebral perfusion was improved after smoking abstinence. Influence of Prior Levels of Smoking Using data from the followup of 248,046 U.S. veterans monitored for 15 years, Rogot and Murray (1980) reported the mortality ratio for stroke among former cigarette smokers who stopped smoking for reasons other than a physician's orders according to the level of prior cigarette smoking. Based on 1,279 strokes among past smokers, the mortality ratio for stroke among former smokers relative to never smokers increased TIMN 438649 251

1'A13I,E 3.--Change (%) in presence of respiratory symptoms, longitudinal studies, by cigarette smoking status Continuing smokers Former smokers Never smokers Symptoms Age Reference (mean) Lost No change" Gained Lost No change4 Gained Lost No changea Gained .otgh 3 m~~/vr Woolf and Zamel (1980)t' ashkin et al. Smokers Light: 43.2t1.7` Moderate: 39.1 t I.1 I leavy: 38.6 t0 .9 18.0 66.0 .3 77.6 16.0 4.1 2.0 4.3 85.0 2.7 13.0 .0 5.0 86.0 9.0 (1984)~ Smokers Male: 45.1 Female: 46.9 Quiners Male: 43.4 Female: 45.6 Comstock et al. 40-59 Net change: 1.0 Net change: -21.0 Net change: 3.0 (1970f Sharp et al. 43-58 10.7 78.0 11.3 16.7 78.5 4.8 4.5 90.8 4.7 (1973)r g Friedman and 20-79 Siegelaub White male 21 plxlh 7.6 85.5 6.9 10.1 89.3 0.6 (1980)g Whitefentale?I ppd 7.4 85.2 7.4 5.0 92.5 2.5 Black male -1 plxl 5.5 89.2 5.3 1.3 97.4 1.3 Pl.th,i;tn 3 !1 o Black female -I ppd 5.0 89.7 5.3 2.9 96.6 1.5 Tashkin et al. (1984) 8.8 77.9 13.3 7.7 86.3 6.0

based study in Tecumseh, MI. Payne and Kjelsberg (1964) reported age- and sex- specific prevalence rates for cough and phlegm production that were comparable for former and never smokers (Figure 5). In contrast, sex-specific rates of dyspnea were highest among former smokers and increased with age (Figure 6). More recent studies have also found lower prevalence of respiratory symptoms among former smokers and documented sex-specific differences among smoking• categories (Table i). Mueller and colleagues (1971) showed that male former smokers had fewer symptoms than current smokers, including cough for 3 months per year, grade 2 dyspnea, and wheezing. Only sputum production for 3 nionths per year was higher among male former smokers than among never smokers. Female former smokers had lower prevalence rates for cough and phlegm production but higher rates for dyspnea and wheezing than current smokers. Rates for female former smokers were generally higher than those for male former smokers. Manfreda, Nelson, and Cherniack (1978) studied subjects from urban and rural communities in Canada, and found very similar overall and sex-specific prevalence rates for these respiratory symptoms among former smokers. In this study, however, female former smokers had prevalence rates between those of current and never smokers for all symptoms. In three separate surveys, Hawthorne and Fry (1978) evaluated the association among smoking, respiratory symptoms, and cardiopulmonary mortality in 11,295 men and 7,491 women from southwest Scotland. Former smokers had prevalence rates for phlegm production and wheezing intermediate to those of current and never smokers. Male former smokers reported shortness of breath as often as male never smokers, whereas female former smokers had an increased prevalence of dyspnea compared with current smokers of either sex. Miller and colleaQues (i988) determined sex-specific prevalence rates for a wide range of respiratory symptoms in a stratified'random sample from the general popula- tion of Michican. Mean aQe for the three smoking groups was comparable. Male current and former smokers had similar lifetime cigarette pack consumption (9.09 x 103 vs. 9.93 x 103), whereas female current smokers had almost twice the cigarette consumption of former smokers (8.32 x 103 vs. 4.50 x 103). The prevalence rates of persistent sputum and wheezing were lower among male former smokers compared with current smokers. In contrast, the prevalence~of dyspnea was similar for male former and current smokers, and findincrs were similar among females. Furthermore, female former smokers had higher rates for dyspnea than males but lower rates for all other respiratory variables assessed. Schenker, Samet. and Speizer (1982) evaluated the effect of smokinQ status on respiratory symptoms of 5,686 women. Age-adjusted prevalence rates for chronic cough, chronic phlegm, and wheeze most days or nights among former smokers were between those for current and never smokers. Grade 3 dyspnea was reported more often by former smokers than current smokers of I to 24 cigarettes per day or by never smokers. Several reports have addressed the occurrence of symptoms in an epidemiologic study in Tucson, AZ (Lebowitz and Burrows 1977; Paoletti et al. 1985). Cross- sectional analyses, based on the first survey of the population, indicated that former smokers had a hiaher prevalence of chronic phlegm production than did never smokers 292 TIMN 438687

Cross-Sectional Studies of Populations The results of community-based studies have shown lower prevalence of respiratory symptoms among former smokers compared with current smokers (Table 1). Two early investigations evaluated symptoms of chronic nonspecific lung disease among smoking groups. Ferris and Anderson (1962) studied a random sample of subjects, aged 25 to 74, from an industrial town in New Hampshire. Using spirometry and interviewer-ad- ministered questionnaires, these researchers recorded lung function and symptoms associated with chronic nonspecific respiratory disease in 1,167 individuals. Chronic nonspecific respiratory disease was considered present if (1) phlegm production was reported six or more times per day for 4 days per week for 3 months per year for the past 3 years (chronic bronchitis); (2) if a diagnosis of asthma had been made and was still present; (3) if wheezing or whistling in the chest occurred most days or nights; (4) if shortness of breath occurred while walking at subject's normal pace on level ground; or (5) if an FEV 1 less than 60 percent of forced vital capacity (FVC) was noted (chronic obstructive lung disease). Age-standardized prevalence rates per 100 for chronic nonspecific respiratory disease showed that both male and female ex-smokers had rates of abnormality similar to those of never smokers and lower than those of current smokers (for males, 18.1 vs. 8.4 vs. 50.3. and for females, 17.2 vs. 19.2 vs. 31.0 for never smokers, ex-smokers, and current smokers, respectively). In 1967, a resurvey of the population using a slightly different random sample was performed (Ferris et al. 1971). Again, the age-standardized rates were less for both male and female ex- smokers than for current smokers. Mueller and colleagues (1971) studied a random sample of one-fifth of the population of Glenwood Springs, CO. Symptoms of chronic nonspecific lung disease, comparable with those defined by Ferris arid colleagues (1971), were reported by 20 percent of 55 male former smokers and by 9 percent of 22 female ex-smokers. These percentages were between those of current and never smokers. Age trends were not apparent among males; the small sample size precluded analysis for females. In the mid-1960s, two surveys assessed the effects of smoking on respiratory symptoms in older men (Table 1). Wilhelmsen and Tibblin (1966) analyzed data from 339 men aged 50 years, born in 1913 and living in Goteborg. an industrial town in Sweden. Of 73 former smokers, the percentages with morning cough for 3 months per year, sputum for 3 months per year, and wheezing other than from colds were lower than those for 182 current smokers of less than or greater than 15 g of tobacco per day and similar to those of 84 never smokers. Dyspnea when walking fast or up a small hill was reported most frequently by current smokers of more than 15 g of tobacco per day; all other groups showed comparable percentages of subjects reporting this symptom. Weiss and coworkers (1963) studied 350 consecutive men, aged 50 years or older, undergoina routine examination in the Philadelphia Pulmonary Neoplasm Research Project (N=6,137). Fifty-three percent of former cigarette smokers (N=68) reported one or more symptoms of couah, wheeze, or dyspnea compared with 57 percent of current smokers (N=183) and 42 percent of never smokers (N=36). Furthermore, former smokers complained of cough as frequently as never smokers (9 vs. t 1 percent) and complained of dyspnea as often as current smokers (46 vs. 44 percent). Only 20 288 TIMN 438683

SCHOENENBERGER, J.C. Smoking change in relation to changes in blood pressure, weight, and cholesterol. Preventive Medicine 11:441-453, 1982. SCHOLL, J.M., BENACERRAF, A., DUCIMETIERE, P., CHABAS, D., BRAU, J., CHAPELLE, J., THERY, J.L. Comparison of risk factors in vasospastic angina without significant fixed coronary narrowing to significant fixed coronary narrowing and no vaso- spastic angina. American Journal of Cardiology 57:199-202, February 1, 1986. SELTZER, C.C. Cigarette smoking and longevity in the elderly. Medical Counterpoint 6(2):29-33, February 1974. SELTZER, C.C. Smoking and coronary heart disease in the elderly. American Journal of the Medical Sciences 269(3):309-315, May-June 1975. SHAPIRO, L.M., HOWAT, A.P., SINGH, S.P. The mortality and morbidity of young survivors of myocardial infarction. Quarterly Journal of Medicine 51(203):366-371, S ummer 1982. SHAPIRO, S., WEINBLATT, E., FRANK, C.W., SAGER, R.V. Incidence of coronary heart disease in a population insured for medical care (HIP): Myocardial infarction, angina pectoris, and possible myocardial infarction. American Journal of Public Health 59(6)(Supplement 2):1-101, June 1969. SHINTON, R., BEEVERS, G. Meta-analysis of relation between cigarette smoking and stroke. British Medical Journal 298(6676):789-794, March 25, 1989. ~ SPARROW, D., DAWBER, T.R., COLTON, T. The influence of cigarette smoking on prognosis after a first myocardial infarction. A report from The Framingham Study. Journal of Chronic Diseases 31(6/7):425-432, 1978. ~ STAMFORD, B.A., MATTER, S., FELL, R.D., PAPANEK, P. Effects of smokinQ cessation on weight gain, metabolic rate, caloric consumption, and blood lipids. American Journal of Clinical Nutrition 43(4):486-494, April 1986. STAMPFER, M.J., GOLDHABER, S.Z., YUSUF, S., PETO, R., HENNEKENS. C.H. Effect of intravenous streptokinase on acute myocardial infarction: Pooled results from randomized trials. New England Journal of Medicine 307(19):1180-1182, November 4, 1982. STEERING COMMITTEE OFTHE PHYSICIANS' HEALTH STUDY RESEARCH GROUP. Final report on the aspirin component of the ongoing Physicians' Health Study. New England Journal of Medicine 321(3):129-135, July 20, 1989. STEINBACH. M., CONSTANTINEANU, M., GEORGESCU, M., HARNAGEA, P., THEODORINI, S., GALFI, L.. DAMSA, T., SCHIOIU, L.. MITU. S., POPESCU. A., ET AL. The Bucharest Multifactorial Prevention Trial of Coronary Heart Disease-Ten year follow-up: 1971-1982. Revue Roumaine de Medecine Interne 22(2):99-106, April-June 1984. STEINBERG, D., PARTHASARATHY, S., CAREW, T.E., KHOO. J.C., WITZTUM, J.L. Beyond Cholesterol: Modifications of low-density lipoprotein that increase its atherogenicity. New England Journal of Medicine 320(14):915-924, April 6, 1989. ~ TAHA, A., BALL. K.P., ILLINGWORTH. R.D. Smoking and subarachnoid haemorrhage. Journal of the Royal Society of Medicine 75(5):332-335, May 1982. TELL. G.S., HOWARD, G., MCKINNEY, W.M., TOOLE. J.F. Cigarette smoking cessation and extracranial carotid atherosclerosis. Journal of the American Medical Association 261(8):1178-1180, February 24, 1989. TOIVANEN, J., YLIKORKALA, 0., VIINIKKA, L. Effects of smoking and nicotine on human prostacyclin and thromboxane production in vivo and in vitro. Toxicology and Applied Pharmacology 82(2):301-306, February 1986. TUOMILEHTO, J., TANSKANEN, A., SALONEN, J.T., NISSINEN, A., KOSKELA, K. Effects of smoking and stopping smoking on serum high-density lipoprotein cholesterol levels in a representative population sample. Preventive Medicine 15(1):35-45, January 1986. rrV4N 438669 271

0.8 0.7 -a 0.6-! 0.5-1 0.4-1 0.3-3 O © i i 1 L1• 19 NONMODIFIERS 23 MODIFIERS I- r r r -p- - - -- - --Cr• 13 QUITTERS ~ ~ 0) -r 0 3 6 9 TIME FROM INITIAL (MO) ! 12 FIGURE -I.---Symptom ratio (number of observed symptoms to number of possible symptoms) in nonmodifiers, modifiers, and quitters at each test period; symptoms are cough, sputum production, wheezing, and shortness of breath SOURCE: Buist et al. (1976). In summary, studies of participants of smoking cessation clinics have shown that respiratory symptoms have disappeared rapidly on quitting, even after 20 pack-years of exposure. Limited studies of asthmatics have provided conflicting results. 287 TIMN 438682

but a lower prevalence compared with current smokers (Table 1). When examined within age groups. the prevalence of chronic phlegm tended to be higher among older male former smokers with substantial past consumption of ciQarettes, suggesting that symptoms may not revert quickly to those of never smokers. 4 To evaluate the effect of cumulative tar consumption on respiratory symptoms and lung function in the Tucson population, Paoletti and coworkers (1985) studied the predictive value of estimated tar exposure and pack-years on respiratory symptoms of 582 current smokers and 621 former smokers. Tar exposure was calculated from the Federal Trade Commission data on tar yield of each type of cigarette smoked and was used to classify retrospectively the smokers' exposures into categories of low and high tar pack-years as well as total tar (kilograms). Only current and former smokers with consistent consumption behavior were analyzed. Ex-smokers had lower prevalence rates of cough, chronic cough, phlegm. and chronic phlegm than did current smokers. Multiple logistic regression analysis was used to determine risk factors for any cough, any wheeze, and dyspnea. Statistical models for former smokers could not be derived using total pack-years, total tar estimates, age, or deep inhalation that sianificantly predicted respiratory symptoms among former smokers of either sex. ~ The low prevalence rates of symptoms among former smokers may have limited the modeling. Ballal (1984) analyzed the effect of depth of inhalation on respiratory symptoms in 75 former smokers as part of a larger study of the smoking behavior of 753 Sudanese medical practitioners. The proportion of former smokers complaining of any wheeze increased with degree of inhalation (slightly, moderately, or deeply), but the trend was not statistically significant. Small numbers and subject selection restrict the importance of this finding. In summary, cross-sectional population-based studies have generally shown that former smokers have reduced prevalence rates for cough, phlegm production, and wheezina compared with current smokers. Dyspnea may not completely reverse after cessation as shown by the comparable prevalence rates for current and ex-smokers in several studies. However. dyspnea may prompt cessation when sustained smokinQ has caused significant physiologic impairment. Differences in symptom rates by gender have been documented in former smokers; potential explanations include sex-specific differences in reporting, differences in smoking practices, or distinct underlying physiologic responses to cessation by gender. Although the relevant data are limited, reversal of most symptoms reflecting mucous oland hypertrophy and hyperplasia and airways inflammation appears to be rapid and not dependent on cumulative smoking at the time of cessation. Measures of past cigarette consumption have not been associated with current respiratory symptoms among former smokers. Occupational Groups Studies of grain elevator workers, dairy farmers. cedar mill workers, and persons exposed to dust, Qas, fumes, and asbestos have addressed the influence of occupation and smoking on respiratory symptoms (Table 2). Broder and coworkers (1979) and Dopico and colleagues (1984) compared respiratory symptoms in grain handlers with those of civic outside workers and of city workers, respectively. In both studies, former 296 TIMN 438691

respiratory symptoms by smoking and occupational exposure status Current smokers Former smokers Never smokers Symptomsa Mean age Occupationally Occupationally Occupationally Reference (Total) exposed Control exposed Control exposed Control Cough 3 mo/yr Broder et al. Grain elevator 67.0 - 38.0 - 23.0 (1979)b workers (A) 39±13(189) Grain elevator 59.0 - 23.0 - 15.0 workers (B) 41±13 (252) Civic outside 56.0 - 15.0 - 5.0 workers (B) 42±14(180) Chan-Yeung White cedar 30.7 12.3 - 8.5 et al. (1984) mill workers 44.3±14.1 (511) Nonwhite cedar 30.7 12.3 - 8.5 mill workers 39.6±9.1 (141) White office - 21.8 - 3.0 - 3.5 workers 43.2+-11.5 (394) Nonwhite office 21.8 - 3.0 - 3.5 workers 39.0±9.9 (46) Kilburn, Shipyard 55.0 - 33.0 - 33.0 Warshaw, Thornton workers 58(288) . • (1986) Michiean men 51.0 48.0 30.0 13.0 15.0 3.0 42(594) Phlegm 3 mo/yr Broder et al. 45.0 17.0 - 15.0 (1979)c Dopico et al. Grain handlers 42.0 32.0 - 37.0 (1984)d 4 L0±12.0 (310) Citv workers - 26.0 - 4.0 - 8.0 41.0±12.0 (239) 297 TIMN 438692

smokers had intermediate prevalence rates for cough, sputum production, wheeze, and shortness of breath compared with current and never smokers. Additionally, former smokers who were grain handlers had more acute and chronic symptoms than ex- smokers who were outside civil or city workers. For grain workers, length of employ- ment had no effect on the prevalence of respiratory symptoms within each smoking group. The results of these two studies differ in that the occupational effect was minimal and less than the smoking effect in the former investigation but significant and.greater in the latter. The choice of control subjects may explain this discrepancy. Babbott and colleagues (1980) assessed the respiratory symptoms of 198 Vermont dairy farmers and 516 nonmineral industrial workers. Former smokers were matched on age (mean 43 years) and years since cessation (mean 8 years). Chronic sputum production, wheezing, and dyspnea were more common among current smokers than among former or never smokers and more frequent among dairy farmers than industrial workers. Similar results were found by Chan-Yeung and coworkers (1984) in a study of 652 cedar mill workers and 440 control office workers. Korn and associates (1987), in a population sample of 8.515 white adults, showed that smoking and exposure to dust, gases, or fumes were independently associated with an increased prevalence of chronic cough. chronic phlegm, persistent wheeze, and breathlessness. Former smokers with Qas or fume exposure were more likely to have respiratory symptoms, particularly breathlessness, than exposed current or never smokers. A multiplicative relationship between smoking and occupational exposure was found for breathlessness but not for other symptoms. ~ Kilburn, Warshaw, and Thornton (1986) conducted an investigation of respiratory symptoms, cardiopulmonary diseases, and asbestosis among 338 male and 81 female shipyard workers and their families. In general, the study group had more symptoms than reported from a similarly stratified random sample of the Michigan population (Miller et al. 1988). The authors suggested that environmental influences in the Los Angeles area may explain the higher rates. Male shipyard workers who were former smokers had more couah, sputum production, and wheezing than shipyard workers who were current smokers, whereas the pattern was reversed for female shipyard workers. In summary, results from selected occupational groups support the findings from the community-based studies, although work exposures may interact with smoking in determining the occurrence of symptoms among former smokers (US DHHS 1985). The results of these investications may be affected by misclassification of exposures and by selection or recall bias. As in the community-based studies, limited descriptive information is provided on former smokers. Longitudinal Studies Numerous longitudinal population-based studies have found rapid resolution of most respiratory symptoms after smoking cessation (Table 3). A study by Woolf and Zamel (1980) indicated that 302 female former smokers with a mean cigarette consumption of 15 pack-years had dramatic resolution of respiratory symptoms within 5 years. These investigators defined former smokers as women who had not smoked for at least 1 year before entry into the study. Persistent former and never smokers were comparable in 299 TIMN 43869,4

WOMEN PACK-YEARS 0 10 20 30 40 50 60 Z85 CURRENT SMOKERS N. 1,495 <5 10 20 30 40 50 60 afi5 PACK-YEARS FIGURE 7. (Continued)-Sex-specific mean height-adjusted FEV I residuals versus pack-years for current and ex-smokers, and distributions of number of subjects by pack-years NOTE: FEVI=1-sec forced expiratory volume. SOURCE: Dockery et al. (1988).* • 318 "TIMN 438713

age; former smokers had a shorter duration of smoking in years than current smokers of 1/2 to 1 pack per day, but similar cumulative pack-years (11.5 vs. 15.0). More former and never smokers reported consistent absence of cough or sputum, dyspnea, or wheeze compared with current smokers. Thirteen percent of former smokers developed cough or phlegm during the study period compared with 9 percent of never smokers and 16 percent of smokers. At enrollment, smokers had more respiratory symptoms and were more likely to develop symptoms over the 5 years of the study. Similarly, in a large population study in the Los Angeles area, respiratory symptoms diminished among former smokers after only 5 years of abstinence (Tashkin et al. 1984). In this study, the following 4 smoking groups were defined: 278 persistent smokers; 414 never smokers; 106 quitters, subjects who smoked regularly at baseline but were nonsmokers at the conclusion of the study; and 294 former smokers, individuals who were regular smokers but had quit at least 2 years prior to baseline. The mean'age for female quitters (45.6 years) was comparable among the smoking cate4ories; the mean age for male quitters (43.4 years) was similar to the mean ages for current and never smokers; however, it was 6.2 years less than that for former smokers. Quitters and former smokers had smoked similar numbers of cigarettes per day (26.3 vs. 24.6 for males; 19.1 vs. 19.0 for females), but quitters had higher pack-years (38.6 vs. 26.8 for males; 27.4 vs. 16.2 for females). In addition, quitters had pack-years comparable with current smokers (38.6 vs. 40.5 for males; 27.4 vs. 30.9 for females). Over the 5 years of the study, quitters recovered from the symptoms of cough, sputum, and wheeze more frequently than continuing smokers. No difference in shortness of breath was found between the two aroups in the 5-year study period. Quitters and former smokers were not compared to determine the relative importance of cumulative exposure versus time since exposure on the observed reduction of symptoms among ex-smokers. Comstock and coworkers (1970) reported comparable findings in a study of respiratory symptoms in 670 male telephone company employees studied for 5 to 6 years. Symptoms of chronic cough. phlegm production, and wheeze decreased sicr- nificantly in quitters whose baseline prevalence for these symptoms was similar to persistent smokers but whose followup values were comparable to never smokers. Baseline and followup prevalence rates for breathlessness in quitters were equivalent to those of persistent smokers. Sharp and colleagues (1973) found similar trends in respiratory symptoms in 1,263 middle-aaed males from an industrial population surveyed in 1961 and again in 1968. Former smokers were defined as individuals who stopped smoking after entry into the study; previous smoking histories were not provided. Over the 7 years of the study, 72.3 percent of former smokers with persistent cough and 64.4 percent with persistent phlegm recovered from the symptoms. These rates of recovery were higher than for the other smoking groups with similar symptoms. Additionally, former smokers who originally complained of dyspnea and wheeze tended to lose these symptoms over the study period, but less dramatically (49-percent and 45.5-percent recovery, respective- ly). New reports of cough and phlegm were made by less than 10 percent of never and former smokers and 16 percent of continuing smokers, whereas new wheeze was found in 13.5 percent of former and 14.1 percent of continuing smokers. In contrast, dyspnea developed in 18.1 percent of former smokers and 22.4 percent of continuing smokers. TIMN 438698 303

MEN PACK-YEARS 0 10 20 30 40 50 60 70 80 90 2:95 C4 ~ -1000 -1 0'1 -250 -I EX-SMOKERS 300-i N-1 277 , 200 ~ 100 O U E---~~-; 0 300 CURRENT SMOKERS N-1,436 200 H 100 ~ ~~ U 0J , <5 10 20 30 40 50 60 70 80 90 2-95 PACK-YEARS FIGURE 7.-Sex-specific mean height-adjusted FEV l residuals versus pack-years for current and ex-smokers, and distributions of number of subjects by pack-years NOTE: FEV i_i-sec forced expiratory volume. SOURCE: Dockery et al. (1988). 317 TIMN 438712 ,

In a study of shorter duration. Friedman and Siegelaub (1980) confirmed the findings of Tashkin and coworkers (1984). Comstock and associates (1970), and Sharp and colleagues (1973). Over approximately 1.5 years of observation. 3.825 recent quitters more often reported decreased chronic cough but no exertional dyspnea when compared with 9.392 persistent smokers. Findings from two Finnish studies and one British study support the results of these North American investigations (Huhti and Ikkala 1980: Poukkula. Huhti, Makarainen• 1982: Leeder et al. 1977). In the 10-year study of Huhti and Ikkala (1980), respiratory symptoms increased in all groups of smokers except male quitters, who had lower prevalence of phlegm production and wheezing (Table 4). Similarly, in a 10-year followup of male pulp mill workers. Poukkula. Huhti, and Makarainen (1982) observed a decrease in respiratory symptoms only for quitters and only for cough and phlegm production. No explanation for the increase in symptoms over time for never smokers was provided in either study. During a 6-year period, Leeder and colleagues (1977) evaluated chronic cough and phlegm annually in 3,916 young married adults. Men who gave up smoking had a pro~ressive decline in the reporting of cough and phlegm. Only a small number of female ex-smokers were included. ~ In summary, the findings from these longitudinal studies agree with those from the cross-sectional surveys and sugaest that cough, phle~m production, and wheezing reverse after cessation, regardless of duration or quantity previously smoked. Dyspnea, however, may be less likely to resolve in subjects with lonaer smokina histories, possibly indicating irreversible damage induced by smoking up to time of cessation. Clinical Studies of Possible Mechanisms Few studies have investiQated the mechanisms by which respiratory symptoms improve 'after smoking cessation. Reversal of mucous gland hyperplasia and reduction in airway inflammation have been considered likely mechanisms but have not been documented. Recovery of epithelia] integrity has been shown in two small clinical studies of epithelial permeability (Minty, Jordan, Jones 1981; Mason et al. 1983). Improvement in tracheal mucous velocity, another possible mechanism by which respiratory symptoms may decrease after smoking cessation, has also been examined. Goodman and coworkers (1978) reported that five of nine younQ former smokers had tracheal mucous velocities that were comparable with aQe-matched never smokers. One subject had a minimally depressed velocity, and three had markedly depressed values. Only one subject was restudied 2 months after baseline and 9 months after cessation, and at that time, tracheal mucous velocity was found still to be reduced. Because subjects were not studied while smoking, the chanse after cessation could not be determined. Camner, Phiiipson, and Arvidsson (1973) studied tracheal velocity in subjects before and after smoking cessation. They found that in I 1 of 17 male.former smokers, tracheal mucous velocity improved 3 months after cessation and that in the remaining 6 former smokers, velocity was slower or similar when compared with baseline values. Improved tracheal mucous velocity may lead to less mucus in the airways and thereby reduce symptoms of cough and wheeze among former smokers. 304 TIMN 438699

TABLE 3.-Continued Continuing smokers Former smokers Never smokers Symptoms Reference Age (me,trl) Lost No change° Gained Lost No change" Gained Lost No change" Gained Comstock et a1. Net change: 4.0 Net change: -15.0 Net change: 0.0 (1970) Sharp et al. 15.4 86.2 6.4 10.2 77.0 ' 12.8 8.0 85.0 7.0 (1973) Dyspnea > grade 2 Woolf and Zame! 17.0 69.0 13.0 18.0 75.0 8.0 7.0 91.0 2.0 (1980)' Tashkin et al. 4.6 89.9 5.5 4.2 89.8 6.0 - - - (1984)' C.omstock et al. Net change: 2.0 Net change: 11.0 Net change: 2.0 (1970) Sharp et al. 11.0 72.8 16.2 14.4 72.8 12.8 10.2 79.8 10.0 (1973)k Friedman et al. (1973) White male ? I ppd Net change: -8.9 Net change: 4.8 White female _I ppd Net change: -11.8 Net change: -5.0 W )• z• Woolf and Zamel 18.0 71.0 11.0 0.0 96.0 5.0 5.0 91.0 4.0 (1980), Tashkin et a!. 11.2 77.8 11.0 13.7 82.1 4.2 - - - (1984 )I

cessation, former smokers had pneumococcal oropharyngeal adherence values com- parable with those of never smokers. The significance of these changes in specific components of host defenses to the risk of subsequent respiratory infections among former smokers has not been characterized. Mortality from influenza and pneumonia with respect to cigarette smoking has been assessed in several cohort studies (Table 5). Mortality from influenza and pneumonia was increased in eversmokers relative to neversmokers in the American CancerSociety Cancer Prevention Study I (ACS CPS-1) followup from 1959 through 1963 (Hammond 1965). In the British Physicians Study, current and former smokers had small excesses of mortal ity from pneumonia. but annual mortality rates from pneumonia increased with the amount smoked (47/100.000 for 1-14 g tobacco/day, 62/100.000 for 15-24 g tobacco/day. 91/100,000 for ?25 g/day) (Doll and Peto 1976). A similar exposure- response relationship was found in the U.S. Veterans Study (Rogot and Murray 1980). Findings from ACS CPS-II on age-adjusted mortality from influenza and pneumonia have been examined for the effects of active smokinQ and smoking cessation (Table 5). Male former smokers of fewer than 21 cigarettes per day have mortality ratios after 10 years of abstinence that are approaching unity. Male former smokers of more than 21 cigarettes per day have mortality ratios approaching unity after 15 years of abstinence, but much higher for shorter periods of abstinence. Female former smokers of any amount have mortality ratios that approach those of never smokers within 3 to 5 years of abstinence. The association between cioarette smokinQ status and mortality from influenza and pneumonia may partially reflect the effects of smoking on respiratory defense mechanisms including immune responses. The vulnerability of persons with cigarette- related cardiopulmonary diseases to respiratory infections may also contribute to the association. For example, Glezen, Decker, and Perrotta (1987) studied underlying diagnoses in patients hospitalized with acute respiratory disease during influenza epidemics in Houston, TX. Chronic pulmonary conditions were the most common underlying condition, and cardiac conditions were the next most frequent. PART II: PULMONARY FUNCTION AMONG FORMER SMOKERS Cross-Sectional Population Studies of FEVi EpidemioloQic studies have generally evaluated airflow obstruction based on FEVI, a spirometric parameter sensitive to airways and parenchymal effects. Cross-sectionall population studies, that is, studies in which luna function and cioarette smoking are measured at a single point in time, have demonstrated that cigarette smoking is a strong determinant of FEV [ level (US DHHS 1984). In those studies in which results from former smokers have been reported. the level of FEV 1 has Qenerally been between that of never smokers and current smokers (Table 6). Several studies have shown that the level of FEV i declines with increasing cumula- tive smokinQ amona former smokers as well as current smokers (Burrows et al. 1977; Beck. Doyle. Schachter 1981: Dockery et al. 1988). Burrows and colleagues (1977) 308 TIMN 438703

serum antigen and prawn antigen, respectively. Whether smokers have a lower in- cidence of hypersensitivity pneumonitis has not been adequately studied. Finally, smokers manifest a blunted immune response to influenza vaccination. Although smokers and nonsmokers have similar postvaccination titers at 3 months (Knowles, Taylor, Turner-Warwick 1981), current smokers have reduced titers at 1 year when compared with nonsmokers (Finklea et al. 1971; Mackenzie, Mackenzie, Holt 1976). In a large clinical trial comparing responses to killed and live attenuated vaccine, smokers had a decreased primary immune response to the killed vaccine (Mackenzie, Mackenzie, Holt 1976). Although effects of smoking on the immune system have been demonstrated, few studies have investigated the association between smoking and acute respiratory illnesses of presumed infectious etiology. Aronson and coworkers (1982) found that smoking was associated with an increased risk of acute respiratory tract illness. In addition, these investigators found that smoking increased the likelihood of having a lower respiratory tract illness and increased the duration of the symptom of cough. These findings corroborated the results of other investigations (Haynes, Krstulovic, Bell 1966; Peters and Ferris 1967; Parnell, Anderson, Kinnis 1966) that showed the same trend for increased respiratory infections among smokers compared with nonsmokers. In contrast, Pollard and associates (1975) found no difference in the incidence of respiratory illness observed among smokers compared with nonsmokers. Short fol- low-up of 9 weeks and selection of Naval recruits who had a high prevalence of acute respiratory disease as patients may explain the discrepancy in results. Kark, Lebiush, and Rannon (1982) studied an outbreak of influenza among 336 men serving in a milita'ry unit in Israel. They found that 68.5 percent of 168 current and occasional.smokers had clinically apparent influenza as compared with 47.2 percent of never and former smokers. Smokers and nonsmokers with influenza had comparable serologic response rates. Amone smokers, the attributable risk percentage for severe influenza, defined as illness resultins in bedrest or loss of workdays, was 40.6 percent (95-percent confidence interval (CI), 21.6-54.8 percent). Similar results have also been reported by several other researchers (Finklea. Sandifer, Smith 1969; MacKenzie, Mackenzie, Holt 1976; Kark and Lebiush 1981). Smoking Cessation and Respiratory Infection The relationship between altered immune and inflammatory functions and the occur- rence of respiratory infections among ex-smokers has not been extensively investiQated. This Section reviews available relevant studies. ~ Studies of animals have shown a return to normal immune and inflammatory function after cessation of cigarette smoke exposure (Holt and Keast 1977). Investigations of humans have yielded similar findings. Specifically, among former smokers, serum concentrations of IgG, IgA, and IQM (Hersey, Prendergast. Edwards 1983) and bronchoalveolar lavage cell numbers and percentages return to those of never smokers (Holt 1987). Additionally, Miller and coworkers (1982) found that within 6 weeks of smoking cessation, the number and function of T lymphocytes reverted to normal. Finally, Raman, Swinburne, and Fedulla (1983) found that 3 years after smoking TIMN 438702 107

140 ~ a 60 L MO AFTER CLINIC 200 r CVNC L N2/L 150 CC/TLC • 0 10 20 30 MO AFTER CLINIC Quitters 0- - - -0 Smokers FIGURE 9.-Mean values for the ratio of CVIVC, of CC/TLC, and slope for phase III of the single breath N2 test (AN2/L), expressed as a percentage of predicted values in 15 quitters and 42 smokers during 30 months after 2 smoking cessation clinics NOTE: Asterisks (*) denote a siQnificant difference from the initial value at p<0.05. CV=closing volume: VC=vital capacity; CC=closing capacity: TLC=total lung capacity. SOURCE: Buist, Nagy, Sexton (1979). 326 TIMN 438721

TA13LE 6.-Continued Reference Yearofstudy Location Population Findings 13oss~ et al. (1981) 1963 Boston, MA t{5O healthy male veterans followed tnitiat FEVt adjusted forage eck, Doyle, Schachter 972-74 ebanon and for 5 yr ,690 men and women, Never smokers 4.0 Former smokers 3.7 Current smokers 3.8 Residual FEV t(L) adjusted for age, (1981) Ansonia, CT; aged 7 and older height, weight ashkin et al. (1984) 973-75 Winnsboro, SC os Angeles, CA .092 men and 1,309 women aged Men Women Never smokers -(1.02 -(l.(12 Former smokers -0.12 4).20 Current smokers -0.22 -11.27 Initial adjusted level of FEV t aylor, Joyce ct al. 981-82 .ondon, UK 25-64 (Ollowed for 5 yr 27 men followed for 7.5 yr Men Women Nonsmokers 3.9 2.7 Former smokers 3.8 2.7 Current smokers" 3.6 2.5 FEV t as percentage of predicted All Reactors Nonreactors (I985) Nonsmokers 119.1 92.0 121.4 Former smokers 107.8 96.4 111.4 Current smokers 1(H).5 84.6 108.5

TABLE 7.-Continued Reference Population Followup Measure TLC FVCur VC FEVI FEVi /FVC Zamel, Leroux, 12 men and 14 62t6 days % change 1.2% 3.0% 4.0% Ramcharan women, mean age (1979) 36t9 yr Pride et ul. 8 male smokers who 4 yr No improvement in spiroinetric tests or MMEF (1980) thought easy to stop MMEF Vniax50 Vmax75 NOTG: TLC=tutal lung capacity; FVC=1i)rced vital eapacity; VC=vital capacity; FM=1-scc tiirced expir.uory volume; MMEt~=ntid-umxiomm expiratory flow. 'Average percentage change recalculated from individual values. b Perccntage cltange in reported mean values.

numbers of T lymphocytes are increased among smokers (Kaszubowski, Wysocki, Machalski 198 1; Robertson et al. 1983; Burton et al. 1983; Smart et al. 1986). Light and moderate smokers have increases in OKT3+ (total T cells) and OKT4+ (T-helper cells) (Hughes et al. 1985; Ginns et al. 1982), and heavy smokers have decreases in OKT4+ and increases in OKT8+ (T-suppressor cells) (Ginns et al. 1982; Miller et al. 1982). Additionally, functional changes in T lymphocytes from smokers have been observed (Whitehead et al. 1974; Suciu-Foca et al. 1974; Onari et al. 1980), but these findings remain controversial. Changes in serum components have also been reported. Smokers have higher levels of C5, C9, C 1 inhibitor (Wyatt, Bridges, Halatek 1981), C-reactive protein, and autoantibodies (antinuclear and rheumatoid factors) (Heiskell et al. 1962), but lower levels of specific immunoglobulins (IgG, IgM, and IgA) (Ferson et al. 1979; Vos-Brat and Rumke 1969; Kosmider, Felus, Wysocki 1973; Dales et al. 1974; Wingerd and Sponzilli 1977; Gulsvik and Fagerhol 1979; Gerrard, Heiner et al. 1980; Leitch. Lumb, Kay 1981; Andersen et al. 1982: Bartelik, Ziolo. Bartelik 1984; McSharry, Banham, Boyd 1985). As previously described, IgE is elevated in smokers (Burrows et al. 1981: Zetterstrom et al. 1981; Hailgren et al. 1982: Warren et al. 1982; Bonini 1982; Stein et al. 1983), and this increase may result from suppression of regulatory T-lymphocyte function (Holt 1987). Bronchoalveolar lavage has provided evidence on the noncellular and cellular com- ponents of the peripheral airways and alveoli among smokers and nonsmokers. Data have indicated that smokers appear to have normal or slightly elevated levels of IgA and IgG (Reynolds and Newball 1974; Warr and Martin 1977; Bell et al. 1981: Velluti et al. 1983: Pre, Bladier, Battesti 1980; Gotoh et al. 1983). Similarly, values for lysozyme (Harris et al. 1975), complement components (Robertson et al. 1976), and fibronectin ( Villiger et al. 1981) are elevated in lavage fluid from smokers. The total number of cells retrieved from lavage of smokers is increased with marked elevation in the percentages of activated macrophages and neutrophils (Hunninahake et al. 1979: Harris. Swenson. Johnson 1970). Absolute lymphocyte numbers remain unchanged, although T-cell function may be altered (Daniele et al. 1977; DeShazo et al. 1983). Recovered macrophages have increased chemotactic function (Warr and Martin 1974; Labedzki et al. 1983: Richards et al. 1984) and increased release of damaging products such as superoxide anions (Hoidal et al. 1979; Hoidal et al. 1980: Joseph et al. 1980; Hoidal and Niewoehner 1982: Greening and Lowrie 1983; Razma et al. 198-1), but diminished microbicidal activity (Martin and Warr 1977; Fisher et al. 1982: Ando et al. 1984). Smokers have been shown to have reduced specific immune responses to inhaled antigens in several occupational studies. Farmers who were never smokers had higher levels of serum precipitins to Micropolysporafaeni than farmers who smoked (Morgan et al. 1973: Morsan et al. 1975; Gruchow et al. 1981: Cormier and Belanaer 1989: Kusaka et al. 1989), whereas pigeon breeders who had never smoked had higher precipitating antibodies to pigeon 7 globulin compared with their smoking counterparts (McSharry et al. 1984; Andersen and Christensen 1983; Boyd et al. 1977). Similar results have been found in poultry workers (Andersen and Schonheyder 1984) and processing workers (McSharry and Wilkinson 1986) in relation to IgG responses to'hen 306 TIMN 438701

References ABBOTT, R.D.. YIN, Y., REED, D.M.. YANO, K. Risk of stroke in male ciaarette smokers. New England Journal of Medicine 315(12):717-720, September 18, 1986. ABERG, A., BERGSTRAND, R.. JOHANSSON. S., ULVENSTAM, G., VEDIN. A., WEDEL, H., WILHELMSSON, C., WILHELMSEN, L. Cessation of smoking after myocardial infarction. Effects on mortality after 10 years. British Heart Journal 49(5):416-422, May 1983. ALLRED. E.N.. BLEECKER, E.R.. CHAITMAN, B.R., DAHMS. T.E., GOTTLIEB, S.O., HACKNEY, J.D., PAGANO, M., SELVESTER, R.H., WALDEN, S.M., WARREN, J. Short-term effects of carbon monoxide exposure on the exercise performance of subjects with coronary artery disease. New England Journal of Medicine 321(21):1426-1432, November 23, 1989. AMELI, F.M., STEIN, M., PROVAN, J.L., PROSSER, R. The effect of postoperative smoking on femoropopliteal bypass grafts. Annals of Vascular Surgery 3(1):20-25, January 1989. AMERICAN CANCER SOCIETY. Cancer Prevention Study II. Unpublished tabulations. ASMUSSEN, I., KJELDSEN, K. Intimal ultrastructure of human umbilical arteries. Observa- tions on arteries from newborn children of smoking and nonsmoking mothers. Circulation Research 36(5):579-589, May 1975. ASSMANN, G.. SCHULTE, H., SCHRIEWER, H. The effects of cigarette smoking on serum levels of HDL cholesterol and HDL apolipoprotein A-I. Journal of Clinical Chemistiy and Clinical Biochemistry 22(6):397-402, June 1984. AUERBACH. 0., CARTER, H.W., GARFINKEL, L., HAMMOND. E.C. Cigarette smokinQ and coronary artery disease. A macroscopic and microscopic study. Chest 70(6):697-705, December 1976. BALLEISEN, J., BAILEY, J., EPPING. P.H., SCHULTE, H.. VAN DE LOO, J. Epidemiologi- cal study on Factor VII, Factor VIII and fibrinogen in an industrial population: I. Baseline data on the relation to age, gender, body-weight, smoking, alcohol. pill-using, and menopause. Thrombosis and Haemostasis 54(2):475-479, August 1985. BARRY, J., MEAD, K., 4ABEL. E.G., ROCCO, M.B., CAMPBELL, S., FENTON, T., MUDGE, G.H. JR., SELWYN. A.P. Effect of smoking on the activity of ischemic heart disease. Journal of the American Medical Association 261(3):398-102. January 20, 1989. BELCH. J.J.F., MCARDLE. B.M., BURNS, P., LOWE, G.D.O., FORBES. C.D. The effects of acute smoking on platelet behaviour, fibrinolysis and haemorheology in habitual smokers. Thrombosis and Haemostasis 51(1):6-8, February 28, 1984. BELL, B.A.. AMBROSE, J. Smoking and the risk of a stroke. Acra Neurochirurglca 64(1- 2):1-7, 1982. BELL, B.A., SYMON, L. Smokin- and subarachnoid haemorrhage. British Medical Journal 1(6163):577-578, March 3, 1979. BLANKENHORN, D.H., NESSIM, S.A.. JOHNSON, R.L., SANMARCO, M.E., AZEN, S.P., CASHILL-HEMPHILL, L. Beneficial effects of combined colestipol-niacin therapy on coronary atherosclerosis and coronary venous bypass grafts. Journal of theAmerican Medical Association 257(23):3233-3240. June 19. 1987. BONITA, R. Cigarette smoking, hypertension and the risk of subarachnoid hemorrhage: A population-based case-control study. Stroke 17(5): 83 1-835, September-0ctober 1986. BONITA, R.. SCRAGG, R., STEWART, A., JACKSON, R., BEAGLEHOLE. R. Cigarette smoking and risk of premature stroke in men and women. British Medical Journal 293(6538): 6-8, July 1986. CARSTENSEN, J.M., PERSHAGEN, G., EKLUND, G. Mortality in relation to cigarette and pipe smoking: 16 years' observation of 25,000 Swedish men. Journal of Epidemiology and Community Health 41:166-172, 1987. TIMIS 438659 261

TABLE 1.--Continuesi Symptoms' Age (number of Current smokers Former smokers Never smokers Reference subjects) Male Female Male Female Male Female (%) (%) (%) (%) (%) (%) Wheeze Wilhelmsen and Tibblin (1966)l 12.6 - 6.9 - 4.8 Weiss et al. (1963)°1 8.0 - 6.0 - 3.0 Fletcher et al (1959)t 16.3 12.9 12.5 2.3 Mueller et al. ( 197 l)b 1 Manfreda. Neison. Cherniack (1978)" 18.0 10.0 12.0 5.0 4.0 1.0 25-54 (256)` 26.8 25.4 10.8 12.1 4.2 3.5 25-54 (246)d 31.5 30.2 14.3 20.0 8.0 8.0 Hawthorne and 21.8 19.2 9.8 10.6 6.1 6.0 Fry (1978)] Miller et al. (1988)" ~ 40.8 28.4 14.7 6.9 12.2 7.4 Schenker. Samet. 14.-1t - 8.3 - 6.0 Speizer(1982)bt 18.5F 28.0g 'Symptoms not mutually exclusive. °Age adjusted. `Urban residents. 'JRural residents. `1-14 cig/day. ` 15-24 ci e/day. ~?25 cig/day. ~ "Weiphted values to be representative ot'state as whole. 'Grade 2: dyspnea when walking with people of same age on level ground. grade 3: dyspnea when walking at one's own pace on level ground. JDyspnea not defined. "Shortness of breath compared with persons of same sex and age. 'Ever wheeze. "'Wheezine not defined. "Wheezin2 apart from colds. men reported wheeze, precluding meaningful analysis for this variable. The high symptom rates seen in this study may reflect the older aQes of the participants and the selection factors contributinQ to enrollment in the Philadelphia Pulmonary Neoplasm Research Project. ~ Three other early investigations confirmed a lower prevalence of specific respiratory symptoms among former smokers (Table 1). Fletcher and coworkers (1959) reported the respiratory symptoms of 244 British post office workers, aged 40 to 59, as part of the study of the relationship between symptoms and tests of lung function. Former smokers of both sexes reported wheezing on most days or nights less often than current smokers, but former smokers also complained of grade 2 dyspnea (i.e., stopping for breath when walking at one's own pace on level ground) as often as current smokers. Fletcher and Tinker (1961) studied respiratory symptoms in 363 London male transport workers. Former smokers had lower prevalence rates for cough, phlegm production, and grade 3 dyspnea (i.e., stopping for breath after walking about 100 yards on level ground) than current smokers of 15 cigarettes or more per day. In a large community- TIMN 438686 291

(subjects who continued to smoke at the same level). The three groups were of comparable ages (35 to 39 years) and had a cumulative cigarette consumption of 20 to 26 pack-years. A symptoms ratio was calculated at 1, 3, 6, and 12 months by taking the number of symptoms (e.g., cough, expectoration, shortness of breath, and wheezing) observed and dividing by the total number of possible symptoms for that group. All groups started with ratio values of approximately 0.55. The ratios for quitters declined within 1 month of cessation and continued to decline over the course of the study from 0.52 to 0.08. In contrast, the ratios for modifiers decreased less than quitters, and nonmodifiers had no change in their ratios over 12 months (Figure 4). Data on individual symptoms were not presented, and smoking abstinence was not verified by biologic markers. In a followup study of more than 30 months, Buist, Nagy, and Sexton and colleagues (1979) again showed that among 15 quitters, respiratory symptoms disappeared by the third or fourth month of followup and did not return during the remainder of the study. However, after a small initial decrease in symptoms among 45 continuing smokers, further decreases were not recorded. The small sample sizes and a 41-percent loss to followup must be considered in interpreting the latter findings. Three studies reported different results for the effect of smoking cessation on respiratory symptoms in asthmatics. Higenbottam, Feyeraband, and Clark (1980) conducted a cross-sectional study of 106 consecutive asthmatic clinic patients and concluded that symptoms decreased after stopping smoking. Age-standardized prevalence rates for chronic cough, chronic cough and phlegm, and wheezing among asthmatics were lower for the 27 former smokers than for the 27 current smokers and the 52 never smokers. Only breathlessness was found more often in former smokers than in the other smoking groups, possibly reflecting irreversible smoking-induced changes. Quantification of smoking history and time since cessation among former smokers was not reported. In contrast. Fennerty and colleagues (1987) as well as Hillerdahl and Rylander (1984) reported increased respiratory symptoms in asthmatics who stopped smoking. Fennerty and coworkers (1987) found that 2 of 14 asthmatics (14.3 percent) who stopped smoking for 24 hours complained that asthmatic symptoms were worsenine. Neither of these two subjects showed a decrease in specific airway conductance or peak, flow, but one had an increase in airway responsiveness to methacholine. However, four of seven asthmatics who abstained from smoking for 7 days recorded a reduction in symptoms. Hillerdahl and Rylander (1984) studied 59 asthmatics who were recruited from an office practice and who had stopped smoking "permanently or for short periods of time." Using questionnaires, these reseachers found that symptoms worsened in 18 asthmatics (30.5 percent) who had stopped smoking. Three subjects claimed onset of new asthmatic symptoms within months of cessation. Asthmatics younger than 40 years of age were more likely to complain of worsening of their asthma than those subjects older than 40 years of age. Hillerdahi and Rylander (1984) concluded that among asthmatics who smoke, psychological reasons, improved secretion clearance, or both could explain.the findings. The uncon- trolled nature of these studies, the small numbers of subjects, the potential for selection and information bias, and the noncomparability of treatment regimens among, study participants limit the usefulness of these findings. 286 TIMN 438681

I tiD1.L. G.-Pi.UllLlllUrLL Current smokers Former smokers Never smokers Symptomsa Mean age Occupationally Occupationally Occupationally Reference (Total) exposed Control exposed Control exposed Control Babbott et al. Dairy farmers 39.0 - 19.Os - 16.0 (1980)`f (198) Industry workers - 30.0 - 9.09 - 10.0 (516) Chan-Yeung 26.1 21.8 14.1 8.2 10.0 7.5 et al. ( 1984) Kilbum, Warshaw, Thornton (1986) Dyspnea > grade 2 55.0 28.0 39.0 15.0 38.0 7.0 Broder et al. (1979)h 23.0 21.0 12.0 11.0 15.0 5.0 15.0 - 16.0 - 5.0 - Dopico et al. (1984)' 72.0 3.0 58.0 6.0 57.0 2.0 Babbott et al. (1980)r 45.0 36.0 51.09 34.0e 27.0 19.0 Chan-Yeung et al. (1984) 34.9 21.1 26.4 10.4 18.1 . 6.-1 Kilbum, Warshaw, Thornton (1986)~ Wheeze 65.0 7.0 59.0 6.0 54.0 2.0 Broder et al. (1979)k 5.0 4.0 8.0 6.0 4.0 8.0 3.0 - 7.0 - 3.0 - Dopico et al. (198=t)' 22.0 50.0 17.0 41.0 17.0 30.0 Babbott et al. (1980)f t 47.0 45.0 -11.09 39.0~ 31.0 22.0 Chan-YeunQ et al. (1984)`° 23.4 24.8 12.3 7.5 9.2 7.5 Kiibum. Warshaw, Thornton (1986)I 68.0 13.0 43.0 8.0 32.0 1.0 'Vlales only: symptoms not mutually exclusive. °Cough for more than a few days/wk. `Phlesm for more than a few days/wk. ''momine expectoration. 3`Chronic sputum production: sputum most days persisting for at least 3 mo/yr. 'Matched on aee and ciearette smoking (current, farm: industry 35.59 vs. 35.4 I: former. 43.?0 vs. 43.24: never. 34.01 vs. 33.88). `'vtatched on years since cessation, farmers 7.95 vs. industry 8.43. hShortness of breath. 'Grade 2 dyspnea. 'Dyspnea at two fliehts of stairs. k Wheeze in attacks. 'Ever wheeze. fOPersistant wheeze: wheeze with colds or wheeze on most davs or nights. 298 TIMN 438693

TABLE 6.-Continued Reference Yearot'study Location Woolf and Suero (1971) Toronto, Canada Schlesinger et al. 1968 Israel (1972) Fletcher et al. (1976) 1961 London, England Itiggins, Keller, Metzner (1977 ) 1962-65 Tecumseh, M1 Populati'on Findings 298 female volunteers, aged 25-54. employed at commercial firms Adjusted mean levels FEVt FEVt/FVCratio Never smokers 2.7 86.7 Former smokers 2.6 85.0 Current smokers 2.5 84.6 4,131 'male civil servants, aged 45 and older 1, 136 men, aged 30-59, employed at bank pr in maintenance of transportation eytupntettt 4,669 men and women, aged 20-74 . Mean value of the FEV t/FVC ratio Never smokers 76.0 Former smokers 74.3 Current smokers 73.6 Adjusted FEVt (L) Never smokers 3.3 Former smokers 3.2 Current smokers 3.0 Mean normalized FEV t score Men Women Never smokers 10.2 10.1 Former smokers 9.9 10.0 Current smokers 9.6 9.8

TABLE 5.-Age-standardized mortality ratios for influenza and pneumonia for current and former smokers compared with never smokers Reference Population Followup Cause of deallt Standardized mortality ratios by smoking status Gender, age group (yr) Never smokers HistorY of smoking f lammond (I965) 1,045,087 US men and women (ACS CPS-I) 4 yr Influenza and pneumonia Men 45-64 Men 65-79 Women 45-64 1.0 1.0 1.0 1.9 1.7 1.3 Never Former Current smokers smokers smokers by amount (g/day) Doll and Peto 34,440 male 20 yr Pneumonia 1.0 I.I 1-14 0.9 (1976) British doctors 15-24 1.1 >25 1.7 Smoking amount (cig/day) Former smokers" Current smokers Rogot and Murray 293,958 US 16 yr Inl7uenza and pneumonia <10 0.8 1.2 ~ (1980) veterans 1(}-20 1.0 1.7 ~ 21-39 1.0 2.2 >40 1.3 2.4 ~ ~ Never Former Current smokers ~ smokers smokers by amount (g/day) ~ ® Carstensen, 25.000 Swedish 16 yr Pneunroniu 1.0 0.6 1-7 1.3 a~ Pershagen, Eklund men 8-I S 1.0 (1987) > 15 1.7 w C ~

TA13LE 5.-Continued Reference Pupulation Followup Cnuse of death Standardized mortality ratios by smoking status American Cancer 1,080,555 US 4 yr Inlluenza and pneumonia Total former Former smokers by years of abstinence urrent Society (unpublished men and women smokers <I 1-2 3-5 6-1O 11-15 16 smokers tabulutions) (ACS CI'S-II) _ Men, total 1.3 -b - - - - - 1.8 Men <21 cig/duy 1.3 3.4 2.1 1.8 1.8 1.1 1.1 2.0 Men >21 cig/tlay 1.3 2.4 - 2.2 2.1 2.1 0.9 1.2 -Women total 1.2 1i - - - - - 2.7 Women <20 cig/day 1.0 - - 1.3 0.6 0.3 1.2 3.4 Women ?20 cig/day 1.1 1.3 2.4 0.6 2.4 1.3 0.2 2.0 NOTE: ACS ('PS-1 and -11=American Cancer Stx:icty Caneer Prevent ion Studies I and It. "Furmer cigarcue smokers whu stupped smoking rur reason+ uther Ihan a physician's urders- 1. Nol calcutaieJ.

physema is present, destruction of peribronchiolar alveoli can be found in the lungs of smokers (Saetta et al. 1985; Wright 1989); the loss of alveolar attachments may result in loss of elastic recoil (Wright 1989). The protease-antiprotease hypothesis proposes that the destruction of lung tissue resulting in emphysema occurs as a consequence of genetic or acquired imbalance of proteolytic and antiproteolytic enzymes in the'lung. As noted in the 1984 Surgeon General's Report (US DHHS 1984), this theory derives from two principal observa- tions: (1) a-l-antitrypsin, a major anti-elastolytic enzyme of the lower respiratory tract, is absent in persons genetically deficient in a-l-antitrypsin; these persons often develop emphysema at an early age (Laurell and Eriksson 1963), and (2) administration of proteolytic enzymes in animal models produces emphysema (Gross et al. 1965). Cigarette smoking is associated with increased numbers of neutrophils and activated macrophages in the lungs of smokers, and neutrophil elastase can cause emphysema in animal models (Harris et al. 1975; Galdston et al. 1984). In addition, the a-1-anti- protease of cigarette smokers has reduced functional activity (Gadek, Fells, Crystal 1979; Gadek et al. 1981). However, although damage to the airways and parenchyma of the lung by cigarette smoke underlies excess lung function loss and COPD in smokers, the factors determin- inQ the development of disease in individual smokers have been only partially charac- terized. A minority of cigarette smokers develop COPD. and cigarette smokina only partially explains the variability in FEV 1 decline (Burrows et al. 1977: US DHHS 1984). Data suQgest that cigarette smoking may influence airway as well as parenchymal inflammation. Thus, host factors determining the response of the airways and parenchyma to cigarette smoking, as well as the intensity of smokinQ, are likely to determine the -development of disease. ~ Cigarette smokinQ has a variety of effects on the immune system; those effects may be important in determining the risks of COPD and other respiratory diseases. Cigarette smokina is associated with elevated total serum IaE. This total IQE does not exhibit seasonal variability, as seen in atopic individuals, and the antigens responsible for this increase have not been identified. Cigarette smoking may influence the development of an atopic diathesis via effects on T-cell helper and suppressor activity (Ginns et al. 1982; Miller et al. 1982), epithelial permeability (Jones et al. 1980; Simani, Inoue. Hogg 1974), or functional alterations of antigen-presenting cells (Warr and Martin 1977). Cigarette smoking is associated with skin test positivity among children exposed to maternal cigarette smoking (Weiss et al. 1985; Martinez et al. 1988); however, this association is not seen in studies of active adult smokers (Burrows, Lebowitz. Barbee 1976). In adult subjects, skin test positivity is most prevalent among former smokers (Taylor, Gross et al. 1985). These data are consistent with the hypothesis that atopic individuals may not become or remain regular smokers because of airway inflammation secondary to inflammatory effects of cigarette smoking. Thus, cigarette smoking may interact with atopy in a complex manner, inducing atopy in less susceptible or initially nonatopic subjects and discouraging highly atopic subjects from taking up smoking. Eosinophils are primary effector cells for allergic inflammation (DeMonchy et al. 1985). Increases in eosinophils are associated with the severity and exacerbations of asthma (Horn et al. 1975). Increased eosinophils are also associated with the occurrence 283 TIMN 438678

TABLE 6.-Continued Reference Year of study Location Population Findings NOTE: FEVt=I-sec forced expiratory volume; FVC=furced vital capacity. °At initial examinaion. which includes continuing smokers and those who subseyuendy quit. w Camilli et al. (1987) Tucson, AZ 654 men and 893 women aged 20 and older, who had FEVt at baseline and followup exams Initial FEV t as percentage of predicted Men Women Nonsmokers 99.8 97.8 Former smokers 93.7 95.6 Current smokers" 91.8 91.6 Dockery et al. (1988) 1974-77 6 US communities 8,191 men and women Deficit of FEVi (L) compared with expected aged 25-74 Men Women Nonsmokers -0.03 -0.02 Former smokers -0.26 -0.05 Current smokers -0.51 -0.23 Lh

TAIILL S.-COntinued Reference Location Population T•ollowup Measure CV/VC % CC/TLC % SQN?/L "Lamel, Leroux, 12 men, 14 women 62±6 days % change -4.1 % -1.9% -10.3% Ramcltaran (1979) Pride et al. 8 male smokers who 4 yr No Sibnificant (1980) thought easy to stop improvement decline NOTE: CV=closing volume; VC=vitul capucity; "11.C=total Iung capacity. Average percentage change recaIculateJ rrom individual values.

RENAUD, S., DUMONT, E., BAUDIER, F., ORTCHANIAN, E., SYMINGTON, I.S. Effect of smoking and dietary saturated fats on platelet functions in Scottish farmers. Cardiovascular Research 19(3):155-159, March 1985. RIVAL, J., RIDDLE, J.M., STEIN, P.D. Effects of chronic smoking on platelet function. Thrombosis Research 45(1):75-85, January 1, 1987. ROBINSON, D., FERNS, G.A., BEVAN, E.A., STOCKS, J., WILLIAMS, P.T., GALTON, D.J. High density lipoprotein subfractions and coronary risk factors in normal men. Arteriosclerosis 7(4):341-346, July-August 1987. ROGERS, R.L., MEYER, J.S., JUDD, B.W., MORTEL, K.F. Abstention from cigarette smoking improves cerebral perfusion among elderly chronic smokers. Journal of the American Medical Association 253(20):2970-2974, May 24-31, 1985. ROGERS, R.L., MEYER, J.S., SHAW, T.G., MORTEL, K.F., HARDENBERG, J.P., ZAID, R.R. Cigarette smoking decreases cerebral blood flow suggesting increased risk for stroke. Journal of the American Medical Association 250(20):2796-2800, November 25, 1983. ROGOT, E. Smoking and General Mortality Among U.S. Veterans,1954-19b9. National Heart and Lung Institute, Publication No. (NIH) 74-544, 1974, p. 65. ROGOT, E., MURRAY, J.L. Smoking and causes of death among U.S. veterans: 16 years of observation. Public Health Reports 95(3):213 222, May-June 1980. RONNEVIK, P.K., GUNDERSEN, T., ABRAHAMSEN, A.M. Effect of smoking habits and timolol treatment on mortality and reinfarction in patients surviving acute myocardial infarc- tion. British Heart Journal 54(2):134-139, August 1985. ROSE, G., HAMILTON, P.J.S. A randomized controlled trial of the effect on middle-aged men of advice to stop smoking. Journal of Epidemiology and Community Health 32(4):275-28 1, December 1978. ROSE, G., HAMILTON, P.J.S., COLWELL, L., SHIPLEY, M.J. A randomised controlled trial of anti-smoking advice: 10-year results. Journal of Epidemiology and Community Health 36(2):102-108, June 1982. ROSE, G., HAMILTON, P.J.S., KEEN, H., REID, D.D., MCCARTNEY, P., JARRETT, R.J. Myocardial ischaemia, risk factors and death from coronary heart-disease. Lancet 1(8003): 105-109, January 15, 1977. ROSE, G., TUNSTALL-PEDOE, H.D., HELLER, R.F. UK Heart Disease Prevention Project: Incidence and mortality results. Lancet 1(8333):I062-1065, May 14, 1983. ROSENBERG. L., KAUFMAN, D.W., HELMRICH, S.P., MILLER. D.R., STOLLEY, P.D.. SHAPIRO, S. Myocardial infarction and cigarette smoking in women younger than 50 years of age. Journal of the American Medical Association 253(20):2965-2969, May 24-31, 1985. ROSENBERG, L., KAUFMAN, D.W., HELMRICH, S.P., SHAPIRO, S. The risk of myocar- dial infarction after quitting smoking in men under 55 years of age. New England Journal of Medicine 313(24):1511-1514, December 12, 1985. ROSENBERG, L., PALMER, J.R., SHAPIRO, S. Decline in the risk of myocardial infarction among women who stop smoking. New England Journal of Medicine 322(4):213-217, January 25, 1990. ROSENMAN, R.H., BRAND, R.J., JENKINS, C.D., FRIEDMAN, M., STRAUS, R., WURM, M. Coronary heart disease in the Western Collaborative Group Study. Final follow-up experience of 8.5 years. Journal of the American Medical Association 233(8):872-877, August 25, 1975. SALONEN, J.T. Stopping smoking and long-term mortality after acute myocardial infarction. British Heart Journal 43(4):463-469, April 1980. SALONEN, J.T., PUSKA, P., MUSTANIEMI, H. Changes in morbidity and mortality during comprehensive community programme to control cardiovascular diseases during 1972-7 in North Karelia. British Medical Journal 2(6199):1178-1183, November 10, 1979. 270 TIMN 438668

MARASINI, B.. BIONDI, M.L.. BARBESTI, S., ZATTA, G.. AGOSTONI, A. Cigarette smoking and platelet function. Thrombosis Research 44(1):85-94, October 1, 1986. MARSHALL, M. Ultrastructural findings on platelet depositions in initial atherogenesis. Wiener Klinische Wochenschrift 98(7):212-214, Apri14, 1986. MARTIN, J.L., WILSON, J.R., FERRARO, N., LASKEY, W.K., KLEAVELAND, J.P., HIRSHFELD, J.W. JR. Acute coronary vasoconstrictive effects of cigarette smoking in coronary heart disease. American Journal of Cardiology 54(1):56-60, July 1, 1984. MEADE, T.W., IMESON, J., STIRLING, Y. Effects of changes in smoking and other charac- teristics on clotting factors and the risk of ischaemic heart disease. Lancet 2(8566):986-988, October 31, 1987. MEADE, T.W., MELLOWS, S., BROZOVIC, M., MILLER, G.J., CHAKRABARTI, R.R., NORTH, W.R., HAINES, A.P., STIRLING. Y., IMESON, J.D., THOMPSON, S.G. Haemos- tatic function and ischaemic heart disease: Principal results of the Northwick Park Heart Study. Lancet 2(8506):533-537, September 6, 1986. MEADE, T.W., VICKERS, M.V., THOMPSON, S.G., STIRLING, Y., HAINES. A.P., MILLER. G.J. Epidemiological characteristics of platelet aQgregability. British Medical Journal 290:428-432, February 9. 1985. MULCAHY, R. Influence of ciQarette smoking on morbidity and mortality after myocardial infarction. British Heart Journal 49(5):410-=115, May 1983. MULCAHY, R., HICKEY, N., GRAHAM. I.M.. MACAIRT, J. Factors affecting the 5 year survival rate of men following acute coronary heart disease. American HeartJournal 93(5): 556-559, May 1977. MULTIPLE RISK FACTOR INTERVENTION TRIAL RESEARCH GROUP. Multiple Risk Factor Intervention Trial. Risk factor changes and mortality results. Journal of the American Medical Association 248(12):1465-1477. September 24. 1982. MULTIPLE RISK FACTOR INTERVENTION TRIAL RESEARCH GROUP. Coronary heart disease death, nonfatal acute myocardial ihfarction and other clinical outcomes in the Multiple Risk Factor Intervention Trial. American Journal of Cardiology 58(l):1-13, July 1, 1986. MULTIPLE RISK FACTOR INTERVENTION TRIAL RESEARCH GROUP. Mortality rates after 10.5 years for participants in the Multiple Risk Factor Intervention Trial. Journal of the American Medical Association 263(13):1795-1801, Apri14, 1990. MUNRO. J.M.. COTRAN, R.S. The pathogenesis of antherosclerosis: Atherogenesis and intlammation. Lahoraroiti Investigation 58(3):249-261, March 1988. NADIGER. H.A.. MATHEW, C.A.. SADASIVUDU. B. Serum malanodialdehyde (TBA reactive substance) levels in cigarette smokers. Atherosclerosis 64(1):71-73, March 1987. NETTERSTROM, B.. JUEL, K. Impact of work-related and psychosocial factors on the development of ischemic heart disease among urban bus drivers in Denmark. Scandinavian Journal of Work and Environmental Health 14(4):231-238, August 1988. NOMURA. A., COMSTOCK, G.W.. KULLER, L., TONASCIA, J.A. Ciaarette smoking and strokes. Stroke 5(4):483-486, July-August 1974. NOWAK, J., MURRAY, J.J., OATES, J.A., FITZGERALD, G.A. Biochemical evidence of a chronic abnormality in platelet and vascular function in healthy individuals who smoke cigarettes. Circulation 76( I):6-14, July 1987. NUSSEY, S.S., ANG, V.T., BEVAN. D.H., JENKINS, J.S. Human platelet arginine vasopres- sin. Clinical Endocrinology 24(4):427-433, April 1986. OCKENE, J.K., KULLER, L.H., SVENDSEN. K.H., MEILAHN, E. The relationship of smoking cessation to coronary heart disease and lung cancer in the Multiple Risk Factor Intervention Trial (MRFIT). American Journal of Public Health 80(8):954-958, August 1990. 268 TIMN 438666

TABLE 4.-Percentage of subjects with respiratory symptoms by smoking status, 1961 and 1971, in a cohort of middle-aged, rural Finns Smoking groups' I II III IV Never smokers 1961 Ex-smokers 1961 Smokers 1961 Smokers 1961 Never smokers 1971 Ex-smokers 1971 Ex-smokers 1971 Smokers 1971 Symptoms Males (89) Females (573) Males (102) Females (26) Males (75) Females (19) Males (21l) Females (47) Phlegm all day- 1961 winter 4 2 7 - 9 11 18 4 1971 6 4 7 4 7 - 27 13 Wheezing most 1961 days - 3 - 3 - 4 2 1971 2 6 4 - 1 - 9 ll Weather affects 1961 chest 6 14 ]0 15 13 11 13 6 1971 19 27 25 23 24 16 39 19 Breathlessness a 1961 rades 3-1 4 20 l0 12 15 16 11 9 1971 10 24 17 12 16 21 21 6 Chronic bronchit 1961 is 9 5 14 15 29 16 36 21 1971 11 8 15 12 9 5 41 21 Mean age (yr) 50 51 50 49 50 47 49 46 in 1961 'Fieures in parentheses are number of subjects. SOURCE: Huhti and Ikkala (1980). Respiratory Infections Numerous clinical studies have shown alterations in immune and inflammatory function among cigarette smokers compared with never smokers. Studies of peripheral blood have shown that current smokers have as much as 30 percent higher leukocyte counts than never smokers (Corre, Lellouch, Schwartz 1971; Friedman et al. 1973). Increases have been reported in polymorphonuclear leukocytes (Brid~es, Wyatt, Rehm 1985), which appear to have normal chemotactic, microbicidal, and secretory functions (Nobel and Penny 1975; Abboud et al. 1983), and monocytes (Nielsen 1985), which may partially lack the ability to kill intracellular Candida (Nielsen 1985). Total TIMN 438700 305

TABLE 6.-Continued Rcference Year of study Location Population Findings Anderson (1979) Lufa, Papua 733 men and women aged 25 and older Age and height-adjusted mean FEVi (L) G i N ew u nea Men Women Never smokers 2.6 2.4 Former smokers 2.6 2.3 Current smokers 2.6 2.4 Ifigenbouam et al. London, Engtand 18,403 mate civil servants, aged 4(>_-64 Age and height-adjusted mean FEVi (L) (198O) Former smokers 3.2 <7 yr abstinent 3.2 7-12 yr abstinent 3.2 _ 13 yr abstinent 3.1 Current smokers 3.1 lfuhti and Ikka)a (1980) 1961 Rural commune, l Fi d 473 men and 569 women. ti ll d f 10 FEVt at initial examination n an owe yr i or Men Women Never smokers 3.5 2.5 Former smokers 3.5 2.5 Current smokers° 3.3 2.8 13osscr et al. (1980) 1963 f3oston, MA 703 heallhy male veterans followed Initial FEV t adjusted for ag e for 10 r y Neversmokers 3.6 Formersmokers 3.6 Current smokers 3.3

CONCLUSIONS 1. Smoking cessation reduces rates of respiratory symptoms such as cough, sputum production, and wheezing, and respiratory infections such as bronchitis and pneumonia, compared with continued smoking. 2. For persons without overt chronic obstructive pulmonary disease (COPD), smoking cessation improves pulmonary function about 5 percent within a few months after cessation. 3. Cigarette smoking accelerates the age-related decline in lung function that occurs among never smokers. With sustained abstinence from smoking, the rate of decline in pulmonary function among former smokers returns to that of never smokers. 4. With sustained abstinence, the COPD mortality rates among former smokers decline in comparison with continuing smokers. TIMN 438744 349

before the start of followup had a higher survival rate than did continuing smokers (Figure 13). Within each stratum of reversibility, former smokers had lower mortality than current smokers. In contrast, mortality in the 3-year followup period of the Intermittent Positive Pressure Breathing Trial was not significantly related to smoking status. The followup period was relatively brief, however. Patient age and the level of FEV i at enrollment were the strongest predictors of mortality. In those prospective studies, smoking was evaluated on entry into the study. Sub- sequent changes in smoking status (i.e., smokers ceasing to smoke or former smokers reverting back to smoking) would reduce the estimated effects of smoking cessation compared with continued smoking. Overall, the extent of the evidence is limited, and a conclusion cannot yet be reached on the effect of smoking on mortality following diagnosis of COPD. 348 TIMN 438743

small improvements in CV and CC were observed, although slope of phase III improved by 10 percent. ~ Martin and coworkers (1975) stated that "CV did not improve with cessation" among 12 participants in a smoking cessation program tracked for I to 3 months. In a 4-year followup of eight men who successfully gave up smoking, Pride and colleagues (1980) reported no improvement in CV, but a significant decline in the slope of phase III within the first few months of cessation. Further improvement did not occur over subsequent years. In summary, abnormalities in the small airways, as measured by CV, CC, and slope of phase III, are substantially reversible among smokers who have not developed significant airflow obstruction. Recovery occurs rapidly and appears to be complete for these measures between 6 months and 1 year after cessation, although the implica- tions of these changes for morbidity and mortality are uncertain. Abnormal frequency dependence of lung compliance (an increased reduction of lung compliance as respiratory frequency increases) also indicates abnormal function of the small airways. Ingram and O'Cain (1971) examined six smokers with abnormal frequency dependence of compliance who quit smoking. At I to 8 weeks after cessation, values in all six had returned to normal. Martin and coworkers (1975) studied 12 participants in a smoking cessation program. At 1 to 3 months after cessation, dynamic compliance was less frequency dependent among 8 of the 12 subjects. Zamel. Leroux, and Ramcharan (1979) also reported less frequency dependence of dynamic compliance among 26 healthy smokers at 2 months after cessation. Diffusing Capacity Among Former Smokers Numerous studies, using a variety of inethods, have shown that pulmonary diffusing capacity is between 6 and 20 percent lower among smokers than among age-matched nonsmokers (Teculescu and Stanescu 1970: Van Ganse, Ferris, Cotes 1972: Krumholz and Hedrick 1973: Frans et al. 1975: Hyland et al. 1978: Enjeti et al. 1978; Bosisio et al. 1980; Miller et al. 1983; Knudson et al. 1984). Only a few studies, however, have assessed the effect of smoking cessation on diffusing capacity. Marcq and Minette (1976) measured single breath carbon monoxide (CO) diffusing capacity (DLcoSB) in male subjects with normal values of FEV 1 and FEV [ divided by FVC. Diffusing capacity was below normal in 13 of 54 (24 percent) of the current smokers compared with 1 of 17 (6 percent) of the former smokers of at least 6 months abstinence. Miller and colleagues (1983) examined D1..coSB in a survey of 511 randomly selected subjects from a population in Michigan. Among never smokers, the mean DLcoSB was 32.5 mL CO per mm Hg per minute for males and 23.0 mL CO per mm HQ per minute for females. Compared with never smokers and adjusted for age and heiQht, male current smokers had 17 percent lower (5.4 mL CO/mm Hg per minute), and female current smokers had 16 percent lower (3.6 mL CO/mm Hg per minute) DLcoSB . Male former smokers abstinent for at least 2 years were lower by 7 percent (2.3 mL CO/mm Hg per minute) compared with never smokers, whereas no difference was found between female current and former smokers. 327 TIMN 438722

OKADA, H., HORIBE, H., YOSHIYUKI, 0., HAYAKAWA, N., AOKI, N. A prospective study of cerebrovascular disease in Japanese rural communities, Akabane and Asahi. Part I: Evaluation of risk factors in the occurrence of cerebral hemorrhage and thrombosis. Stroke 7(6):599-607, November-December 1976. OSTFELD, A.M., SHEKELLE, R.B., KLAWANS, H., TUFO, H.M. Epidemiology of stroke in an elderly welfare population. American Journal of Public Health 64(5):450-458, May 1974. PACKHAM, M.S., MUSTARD, J.F. The role of platelets in the development and complications of atherosclerosis. Seminars in Hematology 23(1):8-26, January 1986. PASSERO. S., ROSSI, G., NARDINI, M., BONELLI, G., D'ETTORRE, M., MARTINI, A., BATTISTINI, N., ALBANESE, V., BONO, G., BRAMBILLA, G.L., ET AL. Italian multi- center study of reversible cerebral ischemic attacks. Part 5. Risk factors and cerebral atherosclerosis. Atherosclerosis 63(2-3):211 224, February 1987. PELLETIER, D.L., BAKER, P.T. Physical activity and plasma total- and HDL-cholesterol levels in weslem Samoan men. American Journal of Clinical Nutrition 46(4):577-585, October 1987. PERKINS, J.. DICK, T.B. Smoking and myocardial infarction: Secondary prevention. Postgraduate Medical Journal 6I (714):295-300, April 1985. PETITTI, D.A., WINGERD, J. Use of oral contraceptives, cigarette smokinQ, and risk of subarachnoid haemorrhage. Lancet 2(8083):234-236, July 29, 1978. ~ PHILLIPS, A.N.. SHAPER, A.G., POCOCK, S.J., WALKER.1i1., MACFARLANE, P.W. The role of risk factors in heart attacks occurring in men with pre-existing ischaemic heart disease. British HeartJourna160(5):404-410, November 1988. PITTILO, R.M., CLARKE, J.M., HARRIS, D.. MACKIE, I.J., ROWLES, P.M., MACHIN, S.J., WOOLF, N. Cigarette smokina and platelet adhesion. British Journal of Haematology 58(4):627-632, December 1984.- PITTILO, R.M., MACKIE, I.J., ROWLES, P.M., MACHIN, S.J., WOOLFF, N. Effects of cigarette smoking on the ultrastructure of rat thoracic aorta and its ability to produce prostacyclin. Thrombosis and Haemostasis 48:173-176, October 29, 1982. QUICK. C.R.G., COTTON, L.T. Measured effect of stopping smoking on intermittent claudica- tion. British Journal of Surgery 69(Supplement):S24-S26, June 1982. RABKIN, S.W. Effect of cigarette smoking cessation on risk factors for coronary athero- sclerosis: A control clinical trial. Atherosclerosis 52:173-184. November 1984. RAICHLEN, J.S., HEALY, B., ACHUFF. S.C., PEARSON, T.A. Importance of risk factors in the angiographic progression of coronary artery disease. American Journal of Cardiology 57(l) :66-70, January 1, 1986. RAMSDALE, D.R., FARAGHER, E.B., BRAY, C.L., BENNETT, D.H., WARD, C., BETON, D.C. Smoking and coronary artery disease assessed by routine coronary arteriography. British Medical Journal 290(6463):197-200, January 19, 1985. REED, D.M., STRONG. J.P., HAYASHI, T., NEWMAN, W.P. III, TRACY, R.E., GUZMAN, M.A.. STEMMERMANN, G.N. Comparison of two measures of atherosclerosis in a prospec- tive epidemiology study. Arteriosclerosis 8(6):782-787. November-December 1988. REICHLEY, K.B., MUELLER, W.H., HANIS, C.L. Centralized obesity and cardiovascular disease risk in Mexican Americans. American Journal of Epidemiology 125(3):373-386, March 1987. REINDERS, J.H., BRINKMAN, H.J.M., VAN MOURIK, J.A.. DE GROOT, P.G. Ciaarette smoke impairs endothelial cell prostacyclin production. Arteriosclerosis 6(l):15-23, January-February 1986. RENAUD, S., BLACHE, D., DUMONT, E., THEVENON, C., WISSENDANGER. T. Platelet function after cigarette smoking in relation to nicotine and carbon monoxide. Clinical Pharmacology and Therapeutics 36(3):389-395, September 1984. TIMN 438667 269

0 -10 -20 -30 .a NN SQ SS a -50 'Computing o starting from last smoking value FIGURE 12.-Effects of quitting smoking during followup among men aged 50-69 NOTE: Subjects in the SQ'` group are included in the SQ group. SOURCE: Camilli et al. (1987). In the Copenhagen City Heart Study, spirometry was performed on 2 occasions separated by 5 years for 12,698 adult residents of the city selected at random (Lange et al. 1989). In general, persons who stopped smoking during this interval experienced less decline of FEV 1 than those who continued to smoke (Table l0); the effect of cessation varied with subject age and amount smoked at the time of quitting. In 1986, the National Heart. Lung, and Blood Institute (NHLBI) initiated a multi- center investigation, the Lung Health Study, to determine whether smoking cessation and bronchodilator therapy can influence the course of subjects without clinical illness who are at high risk for the development of COPD (Anthonisen 1989). Six thousand smokers, aged 35 to 59 years, with evidence of airways obstruction were recruited. They were randomly assigned to one of three groups: a group that received no intervention or usual care group; a group that received, an intensive state-of-the-art 336 'TININ 438731

TABLE 9.--Continued Rate csfdec:line by timoking status Reference i'opulation Foltowup Gender Measure Never smokers Former smokers Quitters Smokers l3osse et al. Ileahhy US veterans 5 yr Male FEVt (tnL/yr)h h 61 49 78 (1981) 2 exanis FVC (mL/yr) 68 64 91 Van der Random santple in the 9-13 yr Malc anit FGV t(mL/yr)` 16.6 13.4 24.5 Lende et al. Netherlands, aged 4 exams Female VC (mLlyr)` 13.7 13.2 15.7 (1981) 15-39 Tashkin Population sample in 5 yr Ma(e FGVt (mL/yr) `t t 56 52 62 70 ci al. (i984) soutiienn California 2 exams FVC (mL/yr)` 61) 60 68 64 Femalz FEV t(mL/yr) 42 38 38 54 f=VC imL/yr3 44 42 44 54 Tayloret al. Volunleer population in 7.5 yr Male FEVt/li;(mL/yr/m'} 6.6 10.9 (1985) the United Kingdom 2 exams . .

<10 10-24 25-49 _S0 PACK-YEARS SMOKING ® Smokers 0 Ex-Smokers FIGURE 10.-Percent-predicted diffusing capacity (%pDL) by pack-years of smoking, current smokers and former smokers, in a study of adults in Tucson, AZ NOTE: Numbers above bars represent sample sizes. SOURCE: Knudson. Kaltenbom. Burrows (1989). factor for excessive loss of FEV( (US DHHS 1984), and smokers have much faster rates of loss of FEV 1 than never smokers (Table 9). Table 9 describes rates of chanse in lung function in selected major longitudinal studies. In each, former smokers or quitters have less decline than current smokers during the followup period. In many investigations, dose-response relationships have been found between the amount smoked during the followup interval and the rate of the FEV 1 decline (US DHHS 1984). For example, Fletcher and colleagues (1976) conducted a study of 792 employed men and performed pulmonary function measurements semiannually for 8 years. They reported that the annual loss of FEV l was 36 mL per year for never smokers. The rate of decline amona cigarette smokers increased with amount smoked per day (44 mL/year for <-4 cigarettes/day; 46 mL/year for 5 to 15 cigarettes/day; 54 mL/year for 15 to 25 cigarettes/day; and 54 mL/year for >25 cigarettes/day). The rate 329 TIMN 438724

TABLE 11.-Mortality attributable to COPD, United States, 1986 Smoking status Crude prevalence (%) Relative risk Population attributable risk (%) Estimated attributable deathsJ Current smokers Male 32.0 9.6 42.7 45,678 Female 24.0 10.5 54.3 31.049 Former smokers Male 34.9 8.7 41.7 44.604 Female 15.3 7.0 21.9 12,501 TOTAL 81.5 133.832 NOTE: COPD=chronic obstructive pulmonary disease. Jlnc(udes deaths for which COPD was listed ati either the underlyine or a contributing cause of death. SOURCE: CDC (1989). A study of mortality among female British physicians has also been reported (Doll et al. 1980). A cohort of 6,194 female doctors who had responded to the 1951 questionnaire was studied for 22 years. The age-adjusted mortality ratio for chronic bronchitis and emphysema among continuing smokers increased with reported ciga- rettes smoked per day (Table 12). Former smokers had a mortality ratio of 5.0 compared with never smokers, which represented a reduction in mortality ratios of 52 percent (1 to 14 cigarettes/day) when compared with light smokers and of 84 percent when compared with heavy smokers (~:25 cigarettes/day). Peto and coworkers (1983) reported COPD mortality based on a 20- to 25-year followup of 2,718 British men who had been enrolled in 5 different respiratory studies in the 1950s. There were no deaths attributed to COPD amonQ never smokers. The ratio of observed to expected COPD deaths was 1.20 and 0.65 for current and former smokers, respectively, with expected deaths based on the entire cohort including smokers and nonsmokers. Thus, the mortality ratio for former smokers was 46 percent lower than that of continuing smokers (Peto et al. 1983). Ebi-Kryston (1989) recently reported on chronic bronchitis mortality in a 15-year. followup of 17,717 male British civil servants. Compared with never smokers, former smokers had a mortality ratio of 5.57 and continuing smokers had a ratio of 8.21. Thus, former smokers had a mortality ratio reduced by 32 percent compared with continuing smokers. Although the data were not presented for COPD, the author reported that the results were similar (Ebi-Kryston 1989). In the United States. Rogot and Murray (1980) reported data on emphysema and bronchitis mortality among 293,958 U.S. veterans studied for 16 years. Former smokers were restricted to those who stopped smoking cigarettes for reasons other than a physician's orders. Current smokers had a mortality ratio.of 12.07 compared with 342 TIMN 438737

30 {- ~ (14) 20 — FEMALES 10 0 (91 44 (64) (33) (9) (50) (154) (72) (29) (96) (156) (34) (12) (19) -10 -20 NN XX SQ SS .. SS NN .. SS NN XX SS NN XX XX <35 35<50 50<70 >_70 AGE GROUPS (YR) FIGURE 11. (Continued)-Mean AFEVI values in never smokers (NN), consistent ex-smokers (XX), subjects who quit smoking during followup (SQ), and consistent smokers (SS) in several age groups NOTE: Numbers of subjects in each category are shown in parentheses. FEV i=1-sec forced expiratory volume. ~ SOURCE: Camilli et ai. (1987) mediate. The findings in the group quitting smoking during the first 12 months may underestimate the benefits of cessation because of subsequent relapse within this group; 16 percent of the quitters had an elevated serum thiocyanate level (> 100 um/dL) indicative of smoking at the first examination compared with 6 percent of never smokers and 7 percent of former smokers. 335 TIMN 438730

TABLE 7.-Spirometric studies of participants in smoking cessation programs Reference Population I)irksen, 31 men born in 1914, Jaozon, Malmo, Sweden Lindell (1974) Dude et al. 3 men and 7 women, (1975) aged 29-61, smoking clinic McCarthy, 15 subjects, smoking Craig, clinic Cherniack (1976) Bake et a!. 9 men and 8 women, (1977) aged 24-69, smoking clinic Buist et al. 6 men and 7 women, (1976) aged 24-53, smoking clinic Buist, Nagy, 3 men und 12 women, Sexton aged 24-52, (I979) smoking clinic Followup Measure TLC FVC or VC FEV i FEV t/FVC MMEF Vm,,O5 8-10 days Change from 110011, 0.7% 52-60 days initial 20 mL -1.3% 6-14 wk % change" -(l.8% -0.2% -0.7% -2.7% -2.0% -10.6% 25-48 wk % change~' 3.1 % 0.3% -9.6% 5 mo Change in % 4.4 4.8 -2.5 -7.3 2 yr predicted 2.2 -1.6 0.7 -1 I.1 11110 Change from +10 mL -4O ml. -40 mL , -60 mL/sec 3 mo initial values -Il)0 mL -310 mL -70 mL -I 10 mL/sec 6 mo -240 mL -12(l ntL +30 mL +40 niL/sec 12 mo -50 ntL -70 mL +60 mL -160 mL/sec 3-4 mo Change in % 2.4 1.5 6-8 mo predicted 6.5 4.6 30 mo 6.5 3.3

TABLE 10.- Decline of FEV I(mLlyr) in subjects in the Copenhagen City Heart Study Women Men Smoking ~roup <55 yr ?55 yr <55 yr >_55 yr Never smokers 13(722) 32(754) 21 (302) 34(151) Former smokers 18(321) 32(307) 27(306) 36(430) Continuina light smokers 17(641) 39(439) 22(279) 52(227) Quitting light smokers 15(80) 28(77) 17(51) 11 (31) Continuing heavy smokers 30(624) 48096) 42(634) 56(248) Quitting heavy smokers 9(17) -(8) 36(32) 43 (t4) NOTE: Numbers of subjects given in parentheses. Light smokers consumed <I5 ci_eJday: heavy smokers consumed?15 cig/day. FEVi=(-sec forcedenpiratory volume. SOURCE: Abstracted from table 2 in Lanee et al. ( 1989). smoking cessation program and regular therapy with an inhaled bronchodilator (ipratropium bromide); and a third group that received the smoking cessation program and a placebo bronchodilator. Placebo/bronchodilator therapy was administered in double-blind fashion. All groups were studied at yearly intervals for 5 years, with rate of change of FEV 1 as the primary end point and respiratory morbidity as a secondary end-point. In this investioation, a larce number of smokers with early airways obstruction were characterized and will be studied closely for 5 years. An extensive data base will be created to test numerous hypotheses reaardina smoking cessation. The question of airways reactivity as a risk factor for rapid lung function loss will be tested definitively in that methacholine sensitivity will have been measured both at the beginnina and at the followup period. ~ The findin2s of the lonaitudinal studies on smoking cessation and decline of FEV 1 have important implications. Persons losing FEV 1 at a greater rate are at risk of developing COPD. After cessation, the return of the rate of decline of FEV 1 to that of never smokers implies that the process leading to COPD can be arrested by cessation. PART III. AIRWAY RESPONSIVENESS, CIGARETTE SMOKING, AND SMOKING CESSATION Population-based studies support a role for smoking as a cause of heightened airway responsiveness (Woolcock et al. 1987; Sparrow et aI 1987; Burney et al. 1987). 'VIost cross-sectional studies that have evaluated this relationship have not adjusted for baseline airway caliber. which may be reduced among smokers (Woolcock et al. 1987; Burney et al. 1987; Welty et al. 1984; Van der Lende et al. 1981; Pham et al. 1984; Buczko et aI. 1984), so that it is difficult to determine how much of the increase in airway responsiveness is accounted for by a direct smoking effect or by a reduction in 337 TIMN 438732

Of the population, 14 percent were histamine-responsive as defined by PD20 (the dose of histamine resulting in a 20-percent decline in FEVO. Responsiveness was related to atopy in younger subjects (aged <40 years) and smoking in older participants (aged >40 years). Former smokers (N= 116) had bronchial reactivity similar to never smokers but lower than current smokers across all age strata (12 vs. 10 vs. 24 percent, respectively). The increase in threshold dose of histamine with age for former smokers was 0.053 per year compared with 0.086 per year among current smokers and 0.027 per year among never smokers. However, for those aged 35 to 44 years, former smokers were more responsive than the other smoking groups (14 vs. 13 and 7 percent for current and never smokers, respectively). The criteria forclassification of former smokers were not provided. Cerveri and colleagues (1989) found similar results in their study of 295 normal never smokers, 70 normal current smokers, and 50 former smokers randomly selected from the general population of a small town in Lombardy, Italy. The daily amount smoked was a stroncer predictor of airway responsiveness than the duration of cigarette use. Further, amona ex-smokers, duration of abstinence did not significantly influence airway responsiveness; however, former smokers with longer abstinence tended to have less bronchial reactivity. Longitudinal Studies Longitudinal population-based studies have not been conducted specifically to evaluate temporal changes in.airway responsiveness among former smokers. Several cohort studies designed to measure declines in spirometric function have included single measurements of airway reactivity. These studies generally confirmed lower responsiveness among former smokers than current smokers and suggested an associa- tion between bronchial reactivity and a more -rapid decline in ventilatory function. Vollmer, Johnson, and Buist (1985) examined bronchodilator responsiveness among subjects from 2 cohorts, 351 members of the Portland Cohort, which included a random sample of 507 Multnomah County employees, and 444 adults from the Screening Center Cohort, consisting of 1,024 subjects screened for emphysema. Individuals were classified as responsive if they showed a 7.72-percent increase in FEV i after two puffs of an isoproterenol metered-dose inhaler. Although no data were presented, former smokers were reported to have a distribution of responsiveness similar to that of current smokers and skewed toward higher values. In case-control analysis conducted within the cohort, responsiveness in both current and former smokers was associated with lower baseline pulmonary function and more rapid ventilatory decline over 9 to ll years. Former smokers in both cohorts had rates of decline that approximated or exceeded those for current smokers, especially among those subjects who were responsive. In a 6-year study of 267 white male grain elevator workers, Tabona and coworkers (1984) found that the percentage of former smokers who were methacholine responsive, defined as a PC20 <_8 mg/mL, was similar to that of never smokers (19.6 vs. 16.7 vs. 25.8 percent for former, never, and current smokers, respectively). In contrast to the Vollmer, Johnson, and Buist study (1985), former smokers showed the lowest ven- tiiatory decline of all smoking groups across all age categories (Tabona et al. 1984). 339 TIMN 438734

of loss among former smokers (i.e., smokers who stopped before the first examination) was 31 mL per year, not significantly different from that of never smokers. In addition, smokers who stopped in the first 2 years of the followup had an annual decline of 38 mL per year. The authors• concluded that smokers who stopped before or early in the study had FEV 1 declines similar to never smokers. In spite of FEV 1 levels having., been reduced by previous smoking, further damage to FEV i due to smoking ceases within a few years of cessation. However, recovery of function was not documented in the study of Fletcher and colleagues (1976). These results have been confirmed in multiple population-based longitudinal studies of FEVt and other pulmonary function parameters (Table 9). Camilli and associates (1987) examined longitudinal decline of FEV 1 in a population sample of 1,705 adults in Tucson, AZ. Mean followup was 9.4 years with an average of 5.2 examinations. Former smokers were defined as having stopped before enroll- ment and continuing to abstain at their last two followup examinations. Quitters smoked on entry into the program but stopped before their last two followup examina- tions. Rates of loss for former smokers and quitters were comparable with those for never smokers and less than those for smokers (Table 9). The age-specific rates of loss (Figure 11) suggest that the benefits of cessation may be greatest among the youngest smokers, that is those with the shortest smoking history. FEV i increased in the youngest group, a finding that the authors interpreted as indicating that the earliest effects of smoking are relatively reversible and could represent, in part, a bronchoconstrictive effect. AmonQ the males in the 50- to 69-year-old age group (Figure 12), 10 of the 24 subjects who quit did so before their second followup examination. For these 10 subjects. the revised annual loss of FEV i from the time Qf cessation returned to that of never smokers, and was much less than that among smokers. In several years, reduced luncr function due to previous smoking was not recovered, except possibly among former smokers who had only been smoking a short time. Taylor, Joyce, and coworkers (1985) examined the annual decline of hei~ht-corrected FEV i(FEV i divided by height3) over 7.5 years in 227 men who were free of a clinical diagnosis of asthma and had not received bronchodilator treatment. Former smokers had an annual decline of FEV 1 divided by height3 (8.0 ± 0.8 mL/year/m3) that was not statistically different from that of never smokers (6.6 ± 0.6 mL/year/m3) but was significantly less than that of continuing smokers (10.9 ± 0.7 mL/year/m3). The 71 former smokers included 50 smokers who had stopped during the followup period. Smokers with bronchial reactivity to inhaled histamine had significantly accelerated annual decline of FEV 1, but an effect of bronchial reactivity was not found among former smokers or never smokers. The reactive former smokers had a lower level percent-predicted FEV 1 at the end of the followup (96.4 vs. 111.4 percent predicted). Because their annual rate of loss was not accelerated, the low level of former smokers must be attributed to either steeper decline while they were smokinQ, low level of FEV 1 before they started smoking, or both. ~ Townsend and colleagues (in press) have recently reported on FEV1 decline in participants in the Multiple Risk Factor Intervention Trial. The analysis was limited to 4,926 subjects who had not used P-blocking agents or smoked cigars, cigarillos, or pipes 333 TIMN 438728

Buist and coworkers (1976) observed a group of six men and seven women who stopped smoking for at least 1 year after a smoking cessation program. Small changes were noted in spirometric parameters. The authors reported that MMEF distinguished between smokers and quitters in that over a 1-year period MMEF declined significantly among smokers but not among quitters. Buist, Nagy, and Sexton (1979) supplemented this sample with participants from another smoking cessation program and extended followup to 30 months for both groups. Significant improvements were observed in VC, FEV l, and MMEF among the quitters during the first 6 to 8 months (Figure 8). No further improvement was observed up to 30 months. 120 r 110 100 90 1 FVC FEV1 ` ------• . -!- - - - F i 0 10 20 30 80 L MO AFTER CLINIC 0 ~ 10 - 20 30 Quiiters - - - - - - Smokers FIGURE 8.-Mean values for FVC and FEV i, expressed as a percentage of predicted values, in 15 quitters and 42 smokers during 30 months after 2 smoking cessation clinics NOTE: Asterisks (*) denote a significant difference from the initial value at p<0.05. FVC=forced vital capacity: FEV i=1-sec forced expiratory volume. SOURCE: Buist. Nagy. Sexton (1979). Zamel, Leroux, and Ramcharan (1979) studied 26 healthy smokers for 2 months after cessation. They reported significant increases in VC and FEV 1 of 3.0 and 4.0 percent change, respectively. In contrast, Pride and coworkers (1980) in a 4-year study of eight male smokers "who thought they would find it easy to give up smoking," reported no improvement in spirometric tests of MMEF. 322 TIMS 438717

W W d TABLE 9.-Popnlati()n-based longitudinal studies of annual decline in pulmonary function Rate of decline by smoking status Rc:ference Poptd;tlion ro11owup Gender Me;tsure Never smokers Former smokers Quitters Smokers Wilhelntsen, SSwedish men born in 1913 4 yr Male VC (nil./yr) 63 58 58 94 Orlta, F1?Vt (uil../yr) 43 33 40 70 libblin I'EF (1./niin/yr) 128 140 100 155 (1969) Ashley et at. Framingham Study IO yr Male FVC (ml./yr) 39 46 58 (1975) 2 ex.utts FEV t/FVC ("/,/yr) 0.3 -1).1 0.5 Fenmale FVC (utl./yr) 33 30 39 . FEV i/FVC (')i',/yr) 4.2 0.2 3.0 Fletcher British workers 8 yr Male r•ryt (mL/yr) 36 31 38 50 et al. (1976) Setnittnnual Kau{finunn French workers 12 yr Male . F6V t(ml./yr)" 42 44 49 etal. (1979) I(uGti and Midelle-aged rural Finns IUyr Male FGVi (mt/yr) 33 45 44 51 lkkalu Femali FEVt (u)L/yr) 27 27 39 35 (1'I80) Wotylf and Canadian valttnteers aged 5 yr FGniale FGVt ("/,/yr) 0.3 0.2 0.7 "Lantel 25-54 2 exurns FEV t/FVC (r%./yr) 1.3 1.4 1.7 (198O) 13osse et al. I lealilty US veterans l0 yr Male FrV1 (mL/yr) 52 57 62 (19H0) 3 ex.uns • FVC (i)il./yr) 69 72 73

In contrast, Bolin, Dahms, and Slavin (1980) and Fennerty and coworkers (1987) found increases in airway responsiveness after cessation. Bolin, Dahms, and Slavin (1980) evaluated the effect of discontinuing smoking on methacholine sensitivity in seven asthmatic subjects. PC20 was measured before and I day after stopping smoking and was found to be 5.62 mg/mL and 1.56 ma/mL, respectively. This increase in airway responsiveness was seen among four of the seven subjects. Finally, Fennerty and colleagues (1987) recorded PD20 to histamine in 14 asthmatics before and 24 hours after smoking cessation. PD20 did not increase significantly. In seven subjects who abstained for 7 days, however, PD20 dose increased significantly (0.67 ± 0.43 mg/mL vs. 2.28 ± 2.03 mg/mL). These studies are limited by short followup, small numbers of subjects, and a lack of adjustment for baseline airway caliber or pulmonary function. Additionally, the analyses did not control for seasonal variation in testing, and the latter three studies did not include a control group. In summary, former smokers appear to have bronchial reactivity comparable with that of never smokers. The comparability of bronchial reactivity among former smokers and never smokers implies that smoking-induced changes iu airway respon- siveness may resolve with abstinence. Available data, however, are limited and not definitive. More research is needed to determine the interaction of smoking cessation with nonspecific airway responsiveness in alterin;• rates of decline in ventilatory function. ~ PART IV. EFFECTS OF SMOKIVG CESSATION ON COPD WIORTALITY The Centers for Disease Control reported that 71,099 persons in the United States died in 1986 with COPD (ICD-9-CM 491-2, 496) as the underlyinQ cause, and 164.049 persons died with COPD as-the underlying cause or as a contributing cause (CDC 1989). It was estimated that 81.5 percent of COPD mortality was attributable to smoking (Table 11). Data from both prospective and retrospective studies have consistently indicated an increased mortality from COPD in cigarette smokers compared with never smokers. In addition, the dearee of tobacco exposure, as measured by the number of cigarettes smoked daily or duration of smoking, strongly affects the risk of death from COPD. This literature was reviewed in the 1984 Report of the Surgeon General (US DHHS 1984), in which cigarette smoking was identified as the major cause of COPD mortality for men and women in the United States. The proceedings of a recent workshop sponsored by NHLBI address the rise in mortality from COPD (Speizer et al. 1989). Several prospective studies have shown that cessation of smoking leads to a decreased risk of mortality compared with that of continuing smokers (Table 12). In the British Physicians Study, Doll and Peto (1976) reported on a 20-year followup of 34,440 male British doctors who completed a questionnaire about their smoking behavior in 195 l. Compared with never smokers, age-adjusted death rates for chronic bronchitis or emphysema were elevated for current smokers and for former smokers (mortality ratio=16.7 and 14.7, respectively). TIMN 438736 341

never smokers. Former smokers had a mortality ratio of 5.20 compared with never smokers. The proceedings of the workshop sponsored by NHLBI on rising COPD mortality included several reports from population-based cohort studies (Speizer et al. 1989). Tockman and Comstock (1989) described mortality in more than 35,000 white residents of Washington County, MD, who were enrolled in 1963 and followed through 1975. Based on the 1963 smoking information, former smokers generally had lower mortality rates for COPD than did current smokers. Marcus and colleagues (1989) reported similar analyses for subjects in the Honolulu Heart Program cohort. Coding of death certificates for COPD differed substantially between the Honolulu Heart Program and the State Health Department. Mortality rates based on the Honolulu Heart Proaram coding showed a temporal pattern of declining mortality from COPD among former smokers with increasing mortality among the current smokers during the followup period 1965-1984. Recent data from ACS CPS-II provide new evidence on mortality from COPD (ACS, unpublished tabulations). The age-adjusted death rates for COPD for men and -women were approximately tenfold higher among current smokers compared with never smokers. The mortality ratios for male and female former smokers compared with never smokers were 8.5 and 7.0, lower than for current smokers (ACS, unpublished tabulations). Several studies have reported on variation in COPD mortality by duration of abstinence (Table 13). In these studies, COPD mortality for former smokers initially increases after cessation above the rates for continuina smokers. The maximum mortality ratio for former smokers was found within the first 5 years of abstinence for ACS CPS-II and between 5 and 9 years after cessation for the British Physicians Study (Doll and Peto 1976). As discussed in Chapter 2, this initial increase in mortality probably reflects cessation by persons with smoking-related illnesses or symptoms. However, even in the U.S. Veterans Study (Rogot and Murray 1980), in which only former smokers who stopped for reasons other than a physician's orders were con- sidered. death rates for emphysema and bronchitis among former smokers were hiQher than for those of current smokers after 5 to 9 years of abstinence. Following this initial rise in COPD mortality after cessation, the mortality ratios drop with increasing duration of abstinence (Table 13). However, even after 20 years or more of abstinence, the risk of COPD mortality among former smokers remains elevated in comparison with never smokers. PART V. FORMER SMOKERS WITH ESTABLISHED CHRONIC OBSTRUCTIVE PULMONARY DISEASE Effect of Smoking Cessation on FEV 1 Decline Among COPD Patients The beneficial effects of smoking cessation on reducing the annual loss of pulmonary function are clearly shown in population studies and followup of smoking cessation participants. These populations have been relatively young and lar~ely free of 345 TIMN 438740

TABLE 9.-Continued Rate of decline by smoking status Reference Poptdulion Fullowup Ciender Measure Never smokers Former smokers Quitters Smokers Cumilli et al. Population sample in -- Mean 9.4 yr Male FEVt (nil./yr)1 12.9 10.8 13.2 25.8 (1987) `I'ucson, AZ 5.2 exams Female FEV f(mL/yr)0 7.6 6.5 2.9 14.6 Burrows Population sample in 10.0 yr Male FEV t(ml./yr)f 11.8 26.6 et aI. (1987) Tucson. AZ 5.4 exams Townsend MRFIT 2-4 yr Male FEV t(nil./yr) 51 44 50 59 et al. (in press) NOTI_': Negative nrunbers indicate mi increase. Furmer .nwkets swppeJ smoking prior to start of swdy, quitters stopped smoking after start of study. Mean values for all smukers have been c:dculated weighted by numtn:r of subjects, where pu6lished data was ,tratified by amount of smuking. VC=vital capacity; FEV t=1-sec forced expiratory volume; PEF=pcak expirawry flow; FVC=forccd vital capacity; l I1=height cutxal; MRFIT=Multiple Risk Factor Intervention Trial. "AJjuxtcd for initial level. hAdjusteJ for age. , `Adjusted for initial level, height, sex, and area uf residence. Weighted tuean for smokers. d Adju.ted fur age, height, and area ol'rc+idence. 'Includex fonner smukers and quitters, tAdjusled to age 50, height 172 cnl. EAJjusteJ to age 50, height 161 cm. h Reealculaled frum FFV t/I~Wspel'II IL' values.

TABLE 8.-Studies of closinf; volume (CV/VC%), closing capacity (CCYI'LC°Io), and slope of alveolar plateau (SBNA) among participants in smoking cessation programs Itelerence Loculion Poprdation followup Measure CV/VC % CC/!'LC % SBN2/L Dirksen, Malmii, Sweden 31 men born in 1914 8-10 days Change from initial +l.l)% -1.0% Janzon, 52-60days -l).6% -1.6% Lindell (1974) [Bode et al. Smoking clinic 3 men. 7 women 6--I4 wk % changel' -35.7% 5.9% (1975) Martin et al. Smoking clinic 12 subjects 1-3 mo Plots, quantitative (1975) data unpublished McCarthy, 15 subjects 25-48 wk %, change 0.0% -13.2% -26.6% Craig. Cherniack (1976) Buist et al. Smoking clinic 6 men, 7 women , I mo Change from initial -1.6% -0.8% -0.3% (1976) 3 mo -19% +1.6% 0.0% 6 mo -4.1 % -5.7% -0.4% 12 nw -3.6% -2.6% -0.3% Bake et al. Smoking clinic 9 men, 8 women 5 mo Change in % predicted 2.8 -1.8 -33.2 (1977) 2 yr -2.5 0.3 -43.8 Buist, Nagy, Smoking clinic 3 men, 12 women 3-4 mo Change in % predicted -23.1 -1.6 -25.6 Sexton 6-8 nw -33.0 -19.5 -51.9 (1979) 30 mo -25.4 -15.4 -48.4

TAY3L[: 6.-Association between cigarette stnoking status and FEVi levels in selected cross-sectional studies of adult populations Reference Year of study Location Population Goldsmith et al. (1962) 1961 San Francisco, 3,311 longshoremen CA Findings Mean FEVt % of predicted value Never smokers 1(X) Former smokers 97 Current smokers 93 Edelman et rtl. (1966) 13.dtintorc,, MD 41() male volunteers, aged 20--103 By parti.t{ regression an:llysis, significant reduction of FEV t among current and former cigarette smokers tiiggins and Kjet1herg 1954-6() Tecumseh, MI 5,14() men and women, aged 16-79 Age-adjusted mean FEV t(L) (1967) 1 Men Women Never smokers 3.3 2.3 Former smokers 3.3 2.3 Current smokers 3,1 2.3 f Iigbins ct al. (1968) 1963 Marion County, 926 white men, aged 20-69 Mean FEVt (L) WV Never smokers 3.6 Former smokers 3.3 Current smokers 3.5 Wi}he)msen, nrhu, 1963 C)i)teburg. Sweden 331 men, age 50 Mean FEVt (L) Tibblin (1969) Never smokers 3.8 Former smokers 3.7 Current smokers 3.5

respiratory disease. The question arises whether the course of the disease can be influenced by smoking cessation once clinically overt COPD becomes apparent. Hughes and coworkers (1982) examined the annual change in lung function among 56 male patients with radiologic evidence of emphysema. Patients who had stopped smoking prior to entry into the study and who did not smoke subsequently had a lower initial level of FEV1 compared with patients who were smoking (45 vs. 55 percent predicted), but the annual rate of loss of FEV I for the former smokers was less (16.4±8.8 mL/year vs. 53.5±5.4 mL/year). Similar results were reported for annual decline of VC (14.9±18.6 mL/year vs. 53.1±11.3 mL/year). Diffusing capacity was lower at the initial assessment among smokers, 57 percent predicted, compared with former smokers, 75 percent, but diffusing capacity did not change significantly during followup. Postma and coworkers (1986) examined the change in lung function in a 2- to 21-,vear followup of 81 patients with chronic airflow obstruction. Fifty-nine of the patients smoked throughout the study, and 22 stopped at the start or some time during followup. Initial level of FEV i was lower among former smokers, but the annual loss of FEV 1 was smaller (49±7 mL/year) than for smokers (85±5 mL/year). In the National Institutes of Health Intermittent Positive Pressure Breathing Trial, 985 patients with COPD but without chronic hypoxemia were enrolled and studied for almost 3 years (Anthonisen et al. 1986). Spirometry was performed at entry and repeated every 3 months. The mean annual decline of FEV 1 was 44 mL per year, the investigators reported that neither past nor present smoking behavior affected the decline of FEV 1 although the data were not provided. In summary, two of the three studies suggested that cessation of smoking is followed by a reduction of the annual loss of pulmonary function, even among patients with advanced COPD or emphysema. However, a beneficial effect of smoking cessation was not found in the large Intermittent Positive Pressure BreathinQ Trial. Additional investigation of the effect of continuing to smoke on lung function decline in patients with COPD is warranted. ~ Effect of Smoking Cessation on Mortality Among COPD Patients The evidence for an effect of smoking cessation on survival of patients with COPD is limited. Traver, Cline, and Burrows (1979) found no association between the smoking status and the survival of 2 patient groups, 200 COPD patients in Chicago, IL, who were studied for 15 years and 100 patients in Tucson. AZ, evaluated for up to 7 years. In a followup of up to 13 years, Kanner and coworkers (1983) examined the survival of 100 patients with chronic airflow limitation, aged 32 to 55 at enrollment. Twelve- year survival probabilities were 86, 79, and 64 percent for never, former, and current smokers, respectively. Postma and colleagues (1985) studied survival of 129 patients with severe chronic airflow obstruction (FEV1 _1,000 mL) for up to 18 years. All nonrespiratory deaths were censored. Patients were classified by the degree of reversibility of airflow obstruction. For both smokers and former smokers, relative survival was highest among those with the ereatest reversibility of airflow obstruction. Smokers who quit smoking 347 TIMN 438742

reported that the level of FEV i had a highly significant quantitative relationship with pack-years in a general population sample of 2,369 subjects in Tucson. AZ, and that smokers and former smokers had comparable levels accounting for pack-years. Higenbottam and coworkers (1980) assessed lung function in the 18,000 males in the Whitehall Civil Servants Study. Mean FEV 1 values among former smokers, adjusted for age and height, were lower than those for never smokers, but greater than those for current smokers. FEV I among former smokers decreased with increasing total con- sumption of cigarettes, but length of abstinence had little effect on FEV 1 among former smokers, although the minimum period considered was less than 6 years. The authors suggested that the depression of lung function associated with cigarette smoking has two components-an irreversible component related to total consumption and a com- ponent rapidly reversible on cessation. Beck. Doyle, and Schachter (1981) analyzed FEV 1 data from 4,690 subjects, aged 7 years and older, in 3 separate U.S. communities. These investigators also found that the deficit in FEV I compared with that expected for never smokers increased with cumulative smoking as measured by pack-years and duration of smoking. After adjusting for cumulative smoking, FEV 1 was 147 mL lower among male smokers and 78 mL lower amonQ female smokers compared with former smokers. Dockery and coworkers (1988) studied 8,191 randomly selected adults in 6 U.S. communities. These researchers found that the deficit of observed FEV i compared with expected age-, height-. and sex-specific values increased linearly with cumulative pack-years among former smokers and current smokers (Figure 7) (Dockery et al. 1988). For the same pack-years, FEV 1 was 123 mL higher among male former smokers and 107 mL higher amona female former smokers compared with current smokers. In a followup study of 227 men, Taylor, Joyce, and coworkers (1985) reported that percent-predicted FEV i for former smokers (107.8 percent predicted) was between that of smokers (100.5) and never smokers (119.1). Within each smoking cateaory, men with increased bronchial reactivity to inhaled histamine had lower levels of percent- predicted FEV 1 than did nonreactors. These differences were statistically significant among smokers (84.6 vs. 108.5 percent predicted for reactors and nonreactors, respec- tively) and former smokers (96.4 vs. 121.5 percent predicted for reactors and nonreac- tors. respectively). The results of these studies suggest that permanent loss of FEV 1 occurs with smoking and that the extent of the loss is associated with the cumulative amount smoked. However, before the development of overt COPD, cessation is associated with an average improvement of 75 to 150 mL, implying that smokincr also causes reversible decrements of function. ~ Pulmonary Function Studies After Smoking Cessation Studies in which the lunQ function of smokers was measured before and after smoking cessation are reviewed in this Section; tests of pulmonary function included spirometry, nitrogen washout. and other techniques potentially sensitive to the effects of cessation. Inflammatory lesions of the small airways have been demonstrated to occur in young adult smokers before the appearance of clinically significant airflow obstruction 316 TIMN 438711

prechallenQe pulmonary function (Fanta and Ingram 1981). Atopy may modify the influence of smoking by further increasing nonspecific airway responsiveness. As noted by O'Connor, Sparrow, and Weiss (1989), this modification may be underes- timated in most studies because those with an allergic predisposition and heiahtened nonspecific responsiveness may not begin smoking--or if they do begin, they may soon quit. The importance of smoking-induced hei~htened airway responsiveness in the pathogenesis of asthma is unknown, and airway hyperresponsiveness is a suspected risk factor for COPD. Mechanisms of Heightened Airway Responsiveness Among Smokers and Former Smokers In both clinical and population-based studies, smoking has been associated with increased airway epithelial permeability (Jones et al. 1980: Minty, Jordan, Jones 1981; Mason et al. 1983), elevated levels of IgE (Burrows et al. 1981; Warren et al. 1982: Zetterstrom et al. 1981: Hallaren et al. 1982: Bonini et al. 1982: Stein et al. 1983), and greater numbers of peripheral eosinophils (Burrows et al. 1980: Taylor, Gross et al. 1985; Tollerud et al. 1989: Kauffmann et al. 1986). These physiologic and im- munologic alterations may partly explain the observed relationship between cigarette consumption and heightened airway responsiveness and/or asthma (Brown. McFadden. Ingram 1977: NIalo, Filiatrault, Martin 1982: Cockcroft et al. 1979: Buczko et al. 1984: Casale et al. i 987; Van der Lende et al. 1981; Gerrard, Cockcroft et al. 1980: Kabiraj et al. 1982; Pham et al. 1984: Enarson et al. 1985: Taylor, Joyce et al. 1985: Woolcock et al. 1987; Sparrow et al. 1987: Rijcken et al. 1987; Burney et al. 1987). Allergy to environmental antigens is known to modify this relationship (Burrows. Lebowitz, Barbee 1976: Welty et al. 198=1; Buczko et al. 1984; Schachter. Doyle, Beck 198-1; Kiviloog. Irnell. Eklund 1974; Dodse and Burrows 1980). The complexity of these interrelationships is only partially explained by published tindinss, and additional clarifying studies are needed. This Section reviews studies that have addressed the above associations with respect to ex-smokers which may explain why airway respon- siveness returns to normal with abstinence. Smoking increases. pulmonary epithelial permeability, which rapidly returns to normal amon- young smokers after cessation. Minty, Jordan. and Jones (1981) used a radiolabeled aerosol technique to study 10 younor asymptomatic male smokers who had stopped smoking for 1. 3. 7, 14. and 21 days. They found that recovery of the epithelial inteerity began within 24 hours and reached maximum at 7 days. Mason and col leagues (1983) later confirmed these findings in 10 young smokers. These studies included small numbers of subjects and had short followup periods after cessation, making interpretation and generalization of the findings difficult. Cross-Sectional Studies Cross-sectional population-based data have shown that former smokers have less airway responsiveness than current smokers. Burney and colleagues (1987) studied 511 randomly selected subjects aQed 18 to 64 years usina inhaled histamine challenee. 338 TIMN 438733

Taken together, these studies suggest that smoking cessation quickly results in small improvements in lung function, as assessed by spirometry. Although the changes were not uniformly statistically significant in the investigations reviewed in this Section, the number of subjects was small in most of the studies. Compared with baseline before cessation, FVC or VC and FEV 1 may improve by about 4 or 5 percent at 4 to 8 months after cessation. In absolute value, this improvement is comparable with the ap- proximately 100-mL improvement reported by Beck, Doyle, and Schachter (1981) and Dockery and coworkers (1988) based on cross-sectional comparison of former smokers to current smokers. Tests of Small Airways Function Several investigators have studied the effects of smoking cessation using measures of small airways function as determined by the single breath nitrogen test (Table 8) and other tests. In the sinale breath nitrogen test, the subject breathes one breath of 100 percent oxygen from residual volume to total lung capacity (TLC). A concentration gradient of nitrogen is thus established with the highest concentrations at the apex. Subsequently, the subject exhales, and the nitrogen concentration of the exhaled air is monitored. The indices of small airways function provided by this test include the closing volume (CV) expressed as a percentage of the vital capacity (CV/VC percent), the closing capacity (CC) expressed as a percentage of TLC (CCITLC percent), and the slope of the nitrogen concentration during the alveolarplateau (slope of phase III). Both CV and CC are increased by abnormalities of the small airways, whereas the slope of the nitrogen concentration reflects the evenness of the ventilation distribution. Buist and colleagues (1976) studied a group of 25 cigarette smokers who attended a smoking cessation clinic. Cessation resulted in significant improvements in CV, CC, and the slope of alveolar plateau at 6 and 12 months. after cessation. Participants in a second smoking cessation clinic were added, and the followup continued to 30 months (Buist, Nagy, Sexton 1979). At the 6- to 8-month followup. CV had improved by 33 percent predicted among those who quit, CC by 20 percent predicted. and the slope of the alveolar plateau by 52 percent. No further improvements were evident at the 30-month followup (Figure 9). Similar improvements have been reported by several other investiQators. Bode and coworkers (1975) found that CV improved by 20 percent 6 to 14 weeks after cessation compared with initial values among 10 subjects. These investigators reported that the slope of phase III was unchanged by cessation. McCarthy, Craia, and Cherniack (1976) observed 131 smokers aged 17 to 66 years who volunteered to attend a smoking cessation clinic. For 15 persons abstinent from 25 to 48 weeks, cessation resulted in a significant 13-percent reduction in CC and a 27-percent reduction in the slope of phase III. Bake and coworkers (1977) showed a 33-percent reduction in the percent-predicted slope of phase III among 17 subjects at 5 months after cessation. On the other hand, only small changes in CV and CC were observed. Zamel, Leroux, and Ramcharan (1979) investigated 26 smokers for an average of 62 days after cessation. Similarly, 323 TIMN 438718

CHAPTER S SMOKING CESSATION AND REPRODUCTION TIMN 438761

However, former smokers who were methacholine responsive had greater FEV i loss over the 6 years of the study than those who were not methacholine responsive. Atopy, presence of symptoms. and initial lung function were not predictive of decline in lung function. Finally, Taylor, Joyce, and coworkers (1985) conducted an investigation over a 7.5-year period of bronchial reactivity and FEV i annual rate of decline among 227 London men, aged 25 to 61 years. These investigators confirmed the results for current smokers of Vollmer, Johnson, and Buist (1985) and Tabona and coworkers (1984). Similarly, former smokers had intermediate levels of methacholine responsiveness compared with the other groups, and those former smokers who were responsive had lower rates of baseline ventilatory function. In contrast, however, former smokers had comparable rates of ventilatory decline, regardless of methacholine responsiveness. In all of these longitudinal studies, bronchodilator or methacholine responsiveness was measured near the end of the study period. Furthermore, precise definitions of former smokers with regard to amount smoked, duration of abstinence, and reasons for quitting were not provided. As discussed previously, the prevalence of airway respon- siveness may also lead to a decision to stop smoking. These limitations in study design must be considered in interpreting the associations amona smoking cessation, non- specific airway responsiveness, and annual decline in FEV t. Clinical Studies Four small clinical studies have addressed airway responsiveness before and after smokinQ cessation. Buczko and coworkers (1984) studied 18 aQe- and sex-matched pairs of healthy nonatopic asymptomatic smokers and nonsmokers. Niethacholine responsiveness was defined as the threshold dose causing a decrease in partial tlows. measured at a volume of 40 percent of the VC above residual volumes (V4op). below the 95-percent CI of CV. In the tirst part of the study. these researchers found that smokers had Qreater overall methacholine responsiveness than never smokers, but the difference was Sisnificant only for smokers with greater than 10 pack-years ofciQarette consumption (Buczko et al. 1984). In the second part of the study, 17 smokers were studied with methacholine testing before and 3 months after smokin- cessation. Threshold dose did not increase significantly for the group as a whole; however. airway responsiveness did decrease among a subset of five smokers with the greatest initial responsiveness. Similar results were found by Simonsson and Rolf (1982) who measured methacholine responsiveness in 10 heavy smokers without symptoms or abnormal pulmonary function tests. They studied each subject I week before cessation and 1. 1.6, and 12 months after smoking cessation. Carboxyhemoglobin was measured to verify smoking abstinence. At baseline, only two subjects were responsive as deter- mined by a 15-percent reduction in FEV 1 after inhalation of 0. 1 percent methacholine. Within 1 month of abstinence, airways responsiveness decreased among four subjects. By 12 months, however, no further significant improvement in airway responsiveness was found for the group. 340 TIMN 438735

(Niewoehner, Kleinerman, Rice 1974). Tests sensitive to abnormalities of the small airways (e.g., helium-oxygen flow volume curves, the single breath nitrogen test or other tests of closing volume, and frequency dependence of compliance) would be expected to be particularly sensitive for detecting changes in function after cessation. In most of the studies reviewed in this Section, participants were enrolled through smoking cessation clinics and subsequently monitored for pulmonary function and smoking status. The data from these studies can assess reversible effects of smokina through documentation of functional change coincident with cessation; irreversible effects can be estimated by comparison of lung function level with predicted values for normal function. f Changes in Spirometric Parameters After Cessation Studies of spirometric measurements of pulmonary function before and after smoking cessation are summarized in Table 7. Many of these studies suggested an improvement in pulmonary function following cessation, although the magnitude of the improvement was small in some of the studies. Dirksen. Janzon, and Lindell (1974) studied a randomly selected sample of men bom in 1914 in Malmo, Sweden. Fifty-eight heavy smokers were solicited to participate in a smoking cessation program, with 31 abstaining for 2 months. Vital capacity (VC) and FEV I /FVC improved 8 to 10 days after cessation. Bode and coworkers (1975) studied 10 healthy subjects who participated in a smoking cessation program and remained abstinent for 6 to 14 weeks. Small and nonsignificant improvements were found for VC (0.3 percent change) and FEV 1(0.9 percent chanQe). Maximum expiratory flow rates with helium at 50 and 25 percent of VC significantly increased. Martin and colleacrues (1975) observed 12 successful subjects from a smoking cessation clinic for I to 3 months. Changes of Vmax5o and Vmax25 after smokinQ cessation were variable and not statistically significant. Resi.dual volume and total pulmonary resistance were also unchanQed. McCarthy. CraiQ, and Cherniack (1976) studied a Qroup of smokers who volunteered to participate in a smoking cessation program. At 25 to 48 weeks after cessation, only 15 participants were still not smoking. AmonQ these subjects, FVC increased from 3.92 L to 4.04 L(3.1 percent change), but FEV 1 (-0.3 percent change) and mid-maximum expiratory flow (MMEF) (-9.6 percent change) decreased. Fifty-nine subjects were evaluated between 6 and 24 weeks following cessation. S ignificant improvements were noted for FVC (2.3 percent of initial value) and the peak expiratory flow rate (6.7 percent of initial value). The FEV I, Vmax5o, and Vmax25 did not change significantly. Bake and colleagues (1977) observed 17 subjects who were abstinent from cigarettes for at least 5 months. During this interval, VC and FEV t improved by 4.4 and 4.8 percent predicted, respectively, while VmaxSO and Vmax25 were reduced by 2.5 percent predicted and 7.3 percent predicted, respectively. At 2-year followup, only nine subjects were still smoking. No significant differences from baseline function were found in this ~roup. TIMN 438714 319

TABLE 12.---Continued Reference Population Tockman and 17,036 Washington County. MD. Cumstock (1989) men aged 35-85 at start of followup periods 19,074 Washington County, MD, women aged 35-85 at start of followup periods Marcus et al. I 1,1361apanese-Amrrican (1989) men in I lawaii, aged 45-65 at Cnrollrnent Standardized mortality ratio by smoking status f•ollowup Cause of death Never smokers Fonner smokers Current smokers 13 yr COPD 1963-68 1.00 2.5 2.5 1969-75 0.(} 1.5 3.6 13 yr COl'D 1963-68 1.00 i.6 3.1 1969-75 1.31 1.(3 7.5 20 yr COPD 1965-69 1.00 7.0 3.9 1970-74 1.4 4.3 1.8 1975-79 2.0 1.9 2.7 198(1-84 1.7 1.1 5.7 N(YTG: COPI)=chronic obstructive putmunary disease; ACS CPS-11-American C:uxer Society Cancer Prevantioa Study It. "Fornier smokers who mopped sntoking cigarenev for reations other than phyxiciari's orders. hDbserveJ dem!»/expectcr{ dcatbs.

Zamel. Leroux, and Ramcharan (1979) measured DLcoSB among 26 healthy smokers before and 2 months after cessation. Although DL,-oSB improved slightly following cessation (0.8 mL CO/mm Hg per minute), the difference was not statistically sig- nificant. Knudson, Kaltenborn, and Burrows (1989) measured DLcoSB in the seventh popula- tion survey conducted in the longitudinal study of a population-based sample in Tucson, AZ. Among current and former smokers, DLcoSB dropped as cumulative consumption of cigarettes increased (Figure 10). Current smokers had significantly lower DLcoSB than either former smokers or never smokers; in persons with normal spirometry, former and never smokers had comparable DLcoSB; former smokers in the group with abnormal spirometry had significantly lower DL.coSB. The DLcoSB quickly returned to normal as the duration of abstinence increased. Within 2 years of quitting, DLcoSB had reached 100 percent of that predicted for women; after 3 years of abstinence, mean DL.coSB was 100 percent of that predicted for men. These data suggest that the effects of cigarette smoking on pulmonary diffusing capacity, as on other measures of lung function. include both irreversible and reversible components. The extent of irreversible change is predicted by cumulative consump- tion; the reversible component improves quickly after cessation. Other Measures Among 19 heavy smokers studied by Dirksen. Janzon. and Lindell (1974), ventilation distribution measured by open-circuit nitrogen clearance improved 1 week after smok- ing cessation. Regional lung function measured with 13'Xe showed improvement I to 3 months after cessation in the study by Martin and colleagues (1975). Zamel and Webster (1984) performed detailed studies of f ve men and five women before and 60 days after cessation. Although Vmax6opercent TLC with helium and air and the maximum flow-static recoil curve did not change, static recoil pressure at 60 percent TLC did decrease significantly 2 months after cessation in 18 of 22 smokers. Michaels and coworkers (1979) also observed a decrease in static recoil pressure at any lung volume after smoking cessation. These authors concluded that a decrease in small airway muscle tone miaht have accounted for the'se findings. Longitudinal Population-Based Studies The natural history of COPD has been described in longitudinal studies of up to two decades. Although a population has not been studied from childhood to the develop- ment of COPD during adulthood, the available data from existing separate investiga- tions encompass the entire course of the disease and support the conceptual model presented earlier (Figure 2). Measures of pulmonary function begin to decline after 25 to 30 years of age. For FEV 1, the annual rate of decline, as estimated from cross-sectional studies. is about 20 to 30 mL annually (US DHHS 1984). Faster loss of function over a sufficient period of time can lead to the development of clinically significant airflow obstruction (Figure 2). The available longitudinal data indicate that cigarette smoking is the primary risk 328 TIMN 438723

1000. 100-I 10- 500- 1000- 1500- 749 1249 1749 BIRTHWEIGHT (g) Perinatal mortality= fetal deaths and neonatal deaths/total births Neonatal mortality= death through 28 days in liveborn infants/live births O --b Fetal mortality= stillbirths/total births k••\ -r- 2000- 2249 --T- r 2500- 3000- 2749 3249 1 -1 3500- 4000- 3749 4249 FIGURE 1.-Perinatal, neonatal, and fetal mortality rates by birthweight in singleton white males, 1980 SOURCE: Williams and Chen (1982). 380 TIMN 438772

TABLE 12.- Prospective studies of COPI) mortality in relation to cigarette smoking status Standardized mortality ratio by smoking status Reference Population Pollowup Cause of death * Never smokers Former smokers Current smokers Doll and Peto 34,440 Rrilish male physicians 20 yr Chronic bronchitis and emphysema 1.0 14.7 16.7 (1976) Doll et al. 6,194 t3ritish fenwle physicians 22 yr Chronic bronchitis and emphysema 1.0 5.0 1-14cig/day 10.5 (1980) 15-24 cig/day 28.5 Rogot and 293,958 US veterans aged 31-84 16 yr C3rouchitis and etuplrysema 1.0 5.2" _>25 c ib/day 32.0 12.1 Murray (1980) Peto et al. 2,718 British men (5 cohorts) 20-25 yr COPD Oh 0.7b 1.2 1) (1983) Carstensen, 25,129 Swedish men Chronic bronchitis and emphysema 1.0 1.8 1-7 cig/day 1.9 Pershagen, . 8-I5 cig/day 2.9 Eklund (1987) >15 cig/day 5.3 fibi-Kryston 17,717 British nwle civil 15 yr Chronic bronchitis 1.0 5.6 8.2 (1989) servants aged 40-64 ACS CNS-II COPD Men 1.0 8.5 10.1 (unpublished Women 1.0 7.0 10.5 tabulations)

30 20 10 0 -10 -20 .. <35 AGE GROUPS (YR) 35<50 50<70 Sa ?70 FIGURE 11.-Mean aFEV 1 values in never smokers (NN), consistent ex-smokers (XX), subjects who quit smoking during followup (SQ), and consistent smokers (SS) in several age groups NOTE: Numbers of subjects in each cateoory are shown in parentheses. FEV i=l-sec forced expiratory volume. ~ SOURCE: Camilli et al. (1987) during the trial and who were observed over 2 to 4 years during the latter half of the study. Subjects who quit smoking during the first 12 months of the study lost FEV 1 at a significantly lower rate than those reporting smoking throughout the trial. Cross-sec- tional analysis of data from the midpoint of the trial indicated the highest level of FEV t for never smokers and the lowest levels for continuing smokers at all ages; FEV 1 levels for former smokers at enrollment and those quitting during. the first year were inter- 334 TIMN 438729 MALES (53) (39) (28)(15) (47) (47) (82) (24) (72) (48) (70) (12) (12) SS

BECKLAKE, M.R., PERML'TT, S. Evaluation of tests of lung function for "screening" for early detection of chronic obstructive lung disease. In: Macklem, P.T., Permutt, S. (eds.) The Lung in Transition Between Health and Disease. New York: Marcel Decker, 1979, pp. 345-387. BELL, D.Y., HASEMAN, J.A., SPOCK, A., MCLENNAN, G., HOOK, G.E.R. Plasma proteins of the bronchoalveolar surface of the lunas of smokers and nonsmokers. American Review of Respiratory Disease 124:72-79, 1981.V BEST, E.W.R. A Canadian Study of Smoking and Health. Department of National Health and Welfare, Epidemiology Division, Health Services Branch, Biostatistics Division, Research and Statistics Directorate, Ottawa, 1966, p. 137. BODE, F.R., DOSMAN, J., MARTIN, R.R., MACKLEM, P.T. Reversibility of pulmonary function abnormalities in smokers. A prospective study of early diagnostic tests of small airways disease. American Journal of Medicine 59(1):43-52, July 1975. BOLIN, J.F., DAHMS, T.E., SLAVIN, R.G. Effect of discontinuing smoking on Mecholyl sensitivity in asthmatic smokers. (Abstract.) Journal of Allergy and Clinical Immunology 64:207-208, 1980. BONINI, S. Smoking, IgE and occupational alleroy. (Letter.) British rl~ledicalJocu•na1284:512- 513, February 13. 1982. ~ BOSISIO, E., GRISETTI, G.C., PANZUTI, F., SERGI, M. Pulmonary diffusing capacity and its components (DM and VC) in young, healthy smokers. Respiration 40(6):307-310, 1980. BOSSE, R., SPARROW, D., GARVEY. A.J., COSTA, P.T. JR., WEISS, S.T., ROWE, J.W. Cigarette smoking, aging, and decline in pulmonary function: A longitudinal study. Archives of Environmental Health 35(4):247-252, July/August 1980. BOSSE, R., SPARROW, D., ROSE, C.L., WEISS, S.T. Longitudinal effect of age and smoking cessation on pulmonary function. American Review of Respiratory Disease 123(4):378-381, 1981. BOYD, G., MADKOUR, M., MIDDLETON, S.. LYNCH, P. Effect of smoking on circulating antibody leveis to avian protein in pigeon breeder's disease. Thorax 32:643-652, 1977. BRIDGES, R.B., WYATT, R.J., REHM. S.R. Effect of smoking on peripheral blood leukocytes and serum antiproteases. Furopean Journal ofRespiratory Diseases 66(Supplement 139):?-1•- 33, 1985. BRODER, I.. MINTZ. S.. HUTCHEON. M.. COREY, P., SILVERMAN. F.. DAVIES. G., LEZNOFF, A., PERESS, L., THOMAS, P. Comparison of respiratory variables in ~rain elevator workers and civic outside workers of Thunder Bay, Canada. American Review of Respiratory Disease 119(2):193-203, February 1979. BROWN, N.E., MCFADDEN, E.R. JR., INGRAM, R.H. JR. Airway responses to inhaled histamine in asymptomatic smokers and nonsmokers. Journal of Applied Physiology 42(4): 508-513, April 1977. BUCZKO, G.B.. DAY, A., VANDERDOELEN, J.L., BOUCHER, R., ZAMEL, N. Effects of cigarette smoking and short-term smoking cessation on airway responsiveness to inhaled methacholine. American Review of Respiratory Disease 129(1):12-14, January 1984. BUIST, A.S., GHEZZO. H., ANTHONISEN, N.R., CHERNIACK. R.M., DUCIC, S.. MACK- LEM, P.T., MANFREDA, J., MARTIN, R.R.. MCCARTHY, D., ROSS. B.B. Relationship between the sinsle-breath N2 test and age. sex, and smoking habit in three North American cities. American Review of Respiratory Disease 120(2):305-318, August 1979. BUIST, A.S.. NAGY, J.M., SEXTON, G.J. The effect of smokin- cessation on pulmonary function: A 30-month follow-up of two smoking cessation clinics. American Review of Respiratory Disease 120(4): 953-957, October 1979. BUIST, A.S., SEXTON, G.J., NAGY, J.M., ROSS, B.B. The effect of smoking cessation and modification on lung function. American Review of Respiratory Disease 114(1):115-122, July 1976. 352 TIMN 438746

EDELMAN, N.H., MITTMAN, C., NORRIS, A.H., COHEN, B.H., SHOCK, N.W. The effects of cigarette smoking upon spirometric performance of community dwelling men. American Review of Respiratay Disease 94(3):421-429, September 1966. ENARSON, D.A., CHAN-YEUNG, M., TABONA, M., KUS, J., VEDAL, S., LAM, S. Predictors of bronchial hyperexcitability in grainhandlers. Chest 87(4):452-455, April 1985. ENJETI, S., HAZELWOOD, B., PERMUTT, S., MENKES, H., TERRY, P. Pulmonary function in young smokers: Male-female differences. American Review of Respiratory Disease 118:667-676, 1978. FANTA, C.H., INGRAM, R.H. JR. Airway responsiveness and chronic airway obstruction. Medical Clinics of North America 65(3):473-487, May 1981. FEINLEIB, M., ROSENBERG, H.M., COLLINS, J.G., DELOZIER, J.E., POKRAS, R., CHEVARLEY, F.M. Trends in COPD morbidity and mortality in the United States. American Review of Respiratory Disease 140(Supplement):S9-S 18, 1989. FENNERTY, A.G., BANKS, J., EBDEN, P., BEVAN, C. The effect of cigarette withdrawal on asthmatics who smoke. European Journal of Respiratory Diseases 71(5):395-399, November 1987. FERRIS, B.G. JR., ANDERSON, D.O. The prevalence of chronic respiratory disease in a New Hampshire town. American Review of Respiratory Diseases 86:165-177, 1962. FERRIS, B.G. JR.. HIGGINS, I.T.T.. HIGGINS, M.W., PETERS, J.M., VAN GANSE, W.F., GOLDMAN. M.D. Chronic nonspecific respiratory disease, Berlin. New Hampshire, 196 1- 1967: A cross-sectional study. American Review of Respiratory Disease 104(2):232-244, 1971. FERSON, M.. EDWARDS, A., LIND. A.. MILTON. G.W., HERSEY, P. Low natural killer-cell activity and immunoglobulin levels associated with smoking in human subjects. International Journal of Cancer 23:603-609, 1979. FINKLEA, J.F., HASSELBLAD, V., RIGGAN, W.B., NELSON, W.C., HAMMER. D.I., NEWILL, V.A. Cigarette smoking and hemagglutination inhibition response to influenza after natural disease and immunizat ion. Ameri can Review of Respiratory Disease 104( 3):368- 376, September 1971. FINKLEA, J.F.. SANDIFER, S.H., SMITH, D.D. Cigarette smoking and epidemic influenza. American Journal of Epidemiology 90(5):390-399, November 1969. FISHER, G.L.. MCNEILL, K.L., FINCH, G.L.. WILSON, F.D.. GOLDE, D.W. Functional evaluation of lung macrophages from cigarette smokers and nonsmokers. Journal of the Reticuloendothelial Societv 32(4):311-321, October 1982. FLETCHER, C., PETO. R. The natural history of chronic airflow obstruction. British Medical Journal 1(9):1645-1648, June 1, 1977. FLETCHER, C.M., ELMES, P.C., FAIRBAIRN, A.S., WOOD. C.H. The significance of respiratory symptoms and the diagnosis of chronic bronchitis in a working population. British Medical Journa15147:258-266, August 29, 1959. ~ FLETCHER. C.M., PETO, R.. TINKER. C., SPEIZER, F.G. The Natural History of Chronic Bronchitis and Emphysema. An Eight-Year Study of Early Chronic Obstructive Lung Disease in Working Men in London. New York: Oxford University Press, 1976. FLETCHER, C.M., TINKER. C.M. Chronic bronchitis: A further study of simple diagnostic methods in a working population. British Medical Journa15238:1491-1498, May 27, 1961. FRANS, A.. STANESCU. D.C., VERITER, C., CLERBAUX, T., BRASSEUR. L. Smoking and pulmonary diffusing capacity. Scandinavian Journal of Respiratory Diseases 56(3):165- 183, September 1975. FRIEDMAN. G.D., SIEGELAUB, A.B. Changes after quitting cigarette smoking. Circulation 61(4):716-723, 1980. TIMN 438749 355

preterm deliveries is so small that the mean would not be affected unless the shift were very large (US DHEW 1979; US DHHS 1980). Cessation Before Conception Most studies of cigarette smoking and birthweight have failed to separate never smokers from women who quit smoking prior to conception. MacMahon. Alpert, and Salber (1966) first examined the association of pre-pre;nancy smoking with birthweight and found no significant difference in the mean birthweight of infants whose mothers smoked before but not during pregnancy compared with never smokers. Subsequent research has confirmed the absence of an association between smoking prior to conception and reduced birthweight (Table 6). In all of these studies, smokers who quit before conception had mean birthweight values that were equivalent or higher than those of never smokers. Other studies in which information on mean birthweiaht could not be derived (Kline, Stein, Hutzler 1987; Anderson et al. 1984; Wainright 1983), with the exception of Zabriskie (1963), have also consistently shown no association between birthweiaht and smoking that ceased prior to conception. Zabris- kie (1963) failed, however, to adjust for smoking during pregnancy, and these results are not directly pertinent in a comparison of birthweiQht in never smokers and smokers who quit before conception. ~ TABLE 6.-Summary of studies of mean birthweight, by smoking status Mean birthweight (_) Reference Never smoked Smoked before but not during pregnancy Smoked during pregnancy Cope. Lancaster. 3,376 3.395 3.200 Stevens (1973) Van den Bers ~ 3,463 3.457 3.355 (1977) Rush and Cassano 3.357 3,384 NR (1983) Visnjevac and Mikov 3,327 3.331 3.097 (1986) NOTE: NR=not reponed. In interpreting these data, misclassification of exposure needs to be considered. MacArthur and Knox (1988) reported that women who quit smoking during pregnancy, and possibly those who quit before pregnancy, were more often living with a partner who smoked. Passive smoke exposure may adversely affect the fetus (Martin and 382 ,rIMN 438774

FRIEDMAN, G.D., SIEGELAUB, A.B., SELTZER, C.C., FELDMAN, R., COLLEN, M.F. Smoking habits and the leukocyte count. Archives of Environmental Health 26:137-143, March 1973. GADEK, J.E., FELLS, G.A., CRYSTAL, R.G. Cigarette smoking induces functional anti- protease deficiency in the lower respiratory tract of humans. Science 206(4424):1315-1316, December 1979. GADEK, J.E., FELLS, G.A., ZIMMERMAN, R.L., RENNARD, S.I., CRYSTAL, R.G. Anti- elastases of the human alveolar structures: Implications for the protease-antiprotease theory of emphysema. Journal of Clinical Investigation 68(4):889-898, October 1981. GALDSTON, M., LEVYTSKA, V., SCHWARTZ, M.S., MAGNUSSON, B. Ceruloplasmin. Increased serum concentration and impaired antioxidant activity in cigarette smokers, and ability to prevent suppression of elastase inhibitory capacity of alphat-proteinase inhibitor. American Review of Respiratory Disease 129:258-263, 1984. GERRARD, J.W., COCKCROFT, D.W., MINK, J.T., COTTON, D.J., POONAWALA, R., DOSMAN, J.A. Increased nonspecific bronchial reactivity in cigarette smokers with normal lung function. American Review of Respiratory Disease 122(4):577-58 1, October 1980. GERRARD, J.W., HEINER, D.C., KO, C.G., MINK, J., MEYERS, A., DOSMAN, J.A. Imrnunoglobulin levels in smokers and non-smokers. Annals ofAllergy 44(5):261-262, May 1980. GINNS, L.C., GOLDENHEIM, P.D., MILLER, L.G., BURTON, R.C., GILLICK, L., COLVIN, R.B., GOLDSTEIN, G., KUNG, P.C., HURWITZ, C., KAZEMI, H. T-lymphocyte subsets in smoking and lung cancer. Analysis by monoclonal antibodies and flow cytometry. American Review of Respiratory Disease 126:265-269, 1982. GLEZEN, W.P., DECKER, M., PERROTTA, D.M. Survey of underlying conditions of persons hospitalized with acute respiratory disease during influenza epidemics in Houston, 1978- 198 1. American Review of Respiratory Disease 136:550-555, 1987. GOLDSMITH, J.R., HECHTER, H.H., PERKINS, N.M., BORHANI, N.O. Pulmonary func- tion and respiratory findings among longshoremen. American Review of Respiratory Disease 86(6):867-874, December 1962. GOODMAN, R.M., YERGIN, B.M., LANDA, J.F., GOLINVAUX, M.H., SACKNER, M.A. Relationship of smoking history and pulmonary function tests to trachael mucous velocity in nonsmokers, young smokers, ex-smokers, and patients with chronic bronchitis. American Review of Respiratory Disease 117(2):205-214, February 1978. GOTOH, T., UEDA, S., NAKAYAMA, T., TAKISHITA, Y., YASOUKA, S., TSUBURA, E. Protein components of bronchoalveolar lavage fluids from non-smokers and smokers. European Journal of Respiratory Diseases 64(5):369-377, July 1983. GREENING, A.P., LOWRIE, D.B. Extracellular release of hydrogen peroxide by human alveolar macrophages: The relationship to cigarette smoking and lower respiratory tract infections. Clinical Science 65:661-664, 1983. GROSS, P., PFTI'ZER, E.A., TOLKER, E., BABYAK, M.A., KASCHAK, M. Experimental emphysema. Its production with papain in normal and silicotic rats. Archives of Environmen- tal Health 11(1):54-58, July 1965. GRUCHOW, H.W., HOFFMANN, R.G., MARX, J.J. JR., EMANUEL, D.A., RIMM, A.A. Precipitating antibodies to farmer's lung antigens in a Wisconsin farming population. American Review of Respiratory Disease 124:411-415, 1981. GULSVIK, A., FAGERHOL, M.K. Smoking and immunoglobulin levels. (Letter.) Lancet 1:449, February 24, 1979. HALLGREN, R., NOU, E., ARRENDAL, H., HIESCHE, K. Smoking and circulating IgE in bronchial carcinoma. Acta Medica Scandinavica 211:269-273, 1982. HAMMOND, E.C. Evidence on the effects of giving up cigarette smoking. American Journal of Public Health 55(5)}:682-691, May 1965. 356 .. TIMN 438750

'1'ABLC 13.-Standardized mortality ratios for COPD among current and former smokers broken down by years of abstinence Former smokers" by yr of abstinence Study Current smokers° <I 1-2 3-S 6-10 11-15 ~16 ACS CNS-11 Men (unpublished tabulations) <2lcig/day 9.7 15.8 21.3 16.7 12.1 9.1 2.7 221 cig/day 13.5 22.6 28.5 25.9 20.2 12.6 4.5 Women <20 cig/day 6.1 11.5 10.0 12.6 3.5 3.4 2.6 _2(1 cig/day 17.1 25.8 32.8 21.3 9.8 8.3 3.9 US Veterans Study (Rogot and Murray 1980) British Physicians Study (men) (Doll and Peto 1976) C Former smokers by yr of abstinence urrent smokers <5 5-9 10-14 15-20 _15 12.1 11.7 14.4 10.2 5.7 7.6 Former smokers by yr of abstinence Currcnt smokers <5 5-9 11~--14 ?15 35.6 34.2 47.7 7.3 8.1 NOTE: COPD=chrouic obsu'uclive pulmonary disease; ACS CPS-11=Anterican Cancer Society Cancer Prevention Study II. "The reference ealegury, never suw6er%, has a standardized mortality ratiu ol 1.0 by delinitiun.

U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Consequences of Smoking for Women. A Report of the Surgeon General. U.S. Departrnent of Health and Human Services, Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health, 1980. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Coirsequences of Smoking: Cardiovascular Disease. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Office on Smoking and Health. DHHS Publication No. (PHS) 84-50204,1983. U.S. DEPARTMENT*OF HEALTH AND HUMAN SERVICES. Reducing the Health Conse- quences of Smoking: 25 Years of Progress. A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. DHHS Publication No. (CDC) 89-8411, 1989. U.S DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE. The Health Consequen- ces of Srnoking. A Report to the Surgeon General: 1971. U.S. Department of Health, Education, and Welfare, Public Health Service, Health Services and Mental Health Ad- ministration. DHEW Publication No. (HSM) 71-7513, 1971. U.S DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE. Smoking and Health. AReportof the Surgeon General. U.S. Department of Health, Education, and Welfare, Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health. DHEW Publication No. (PHS) 79-50066,1979. U.S. PUBLIC HEALTH SERVICE. Smoking and Health. Report of the Advisory Committee to the Surgeon General of the Public Health Service. U.S. Department of Health, Education, and Welfare, Public Health Service, Center for Disease Control. PHS Publication No. 1103, 1964. VANDORMAEL, M.G., DELIGONUL, U., KERN, MJ., HARPER, M., PRESANT, S.. GIBSON, P., GALAN, K., CHAITMAN, B.R. Multilesion coronary angioplasty: Clinical and angiographic follow-up. Journal of the American College of Cardiology 10(2):246-252, August 1987. VESSEY, M.P., LAWLESS, M., YEATES, D. Oral contraceptives and stroke: Findings in a large prospective study. British Medical Journal 289(6444):530-531. September 1, 1984. 'VLIETSTRA, R.E., KRONMAL, R.A., OBERMAN, A., FRYE, R.L.. KILLIP, T. III. Effect of cigarette smoking on survival ofpatients with angiograpitically documented coronary artery disease: Report from the CASS Registry. Journal of the American Medical Association 255(8):1023-1027, February 28, 1986. VON DER LIPPE, G., LUND-JOHANSEN, P. Reduction of sudden deaths after myocardial infarction by treatment with beta-blocking drugs. In: Zanchetti, A. (ed.) Advances in Beta-Blocker Therapy 11. Proceedings of the Second International Bayer Bera-Blocker Symposium, October 16-17,1981. Venice, 1982, pp. 100-105. WEIN'I'RAUB, W.S.. KLEIN, L.W., SEELAUS, P.A., AGARWAL, J.B., HELFANT, R.H. Importance of total life consumption of cigarettes as a risk factor for coronary artery disease. American Journal of Cardiology 55(6):669-672, March 1, 1985. WEIR, J.M., DUNN, J.E. JR. Smoking and mortality: A prospective study. Cancer 25(1):105- 112, January 1970. WELIN, L., SVARDSUDD, K., WILHELMSEN, L., LARSSON, B.. TIBBLIN, G. Analysis of risk factors for stroke in a cohort of men bom in 1913. New England Journal of Medicine 317(9):521-526, August 27, 1987. WILHELMSEN, L., BERGLUND, G.. ELMFELDT, D., TIBBLIN, G., WEDEL, H., PEN- NERT, K., VEDIN, A., WILHELMSSON, C., WERKO, L. The Multifactor Primary Preven- tion Trial in Goteborg, Sweden. European HeartJournal7(4):279-288, April 1986.. 272 TIMN 438670

BUIST, A.S., VOLLMER, W.M. Prospective investigations in asthma. What have we learned from Ionaitudinal studies about lung growth and senescence in asthma? Chest 91(Supple- ment):119S-126S. June 1987. BURNEY. P.G.J.. BRITTON, J.R., CHINN, S., TATTERSFIELD, A.E., PAPACOSTA, A.O., KELSON, M.C., ANDERSON. F., CORFIELD, D.R. Descriptive epidemiology of bronchial reactivity in an adult population: Results from a community study. Thoras 42(1):38-44, January 1987. BURROWS, B. An overview of obstructive lung diseases. Medical Clinics of North America 65(3):45 5-471, May 198 1. BURROWS, B., HALONEN, M.. BARBEE. R.A.. LEBOWITZ, M.D. The relationship of serum immunoglobulin E to cigarette smoking. American Review of Respiratnry Disease 124 (5):523-525. November 1981. BURROWS, B., HASAN, F.M., BARBEE. R.A.. HALONEN, M., LEBOWITZ, M.D. Epidemiologic observations on eosinophilia and its relation to respiratory disorders. American Review of Respirator;y Disease 122(5):709-719, November 1980. BURROWS. B., KNUDSON, R.J., CAMILLI, A.E., LYLE. S.K.. LEBOWITZ. M.D. The "horse-racing effect" and predicting decline in forced expiratory volume in one second from screening spirometry. American Review nf Respirator _y Disease 135(4):788-793, April 1987. BURROWS. B., KNUDSON, R.J., CLINE, M.G., LEBOWITZ, M.D. Quantitative relation- ships between cigarette smoking and ventilatory function. American Reti•iex. of Respiratory Disease l 15(2):195-205. February 1977. BURROWS, B.. LEBOWITZ, A.D.. BARBEE, R.A. Respiratory disorders and allergy skin-test reactions. Arrnals of Internal Medicine 84(2):134-139, February 1976. BURTON. R.C., FERGUSON, P.. GRAY. M., HALL, J.. HAYES, M.. SMART. Y.C. Effects of aQe, eender, and cigarette smoking on human immunoreaulatory T-cell subsets: Estab- lishment of normal ranges and comparison with patients with colorectal cancer and multiple sclerosis. Diagnostic Immunology 1(3 ):216-223. 1983. CAMILLI. A.E.. BURROWS, B., KNUDSON, R.J., LYLE, S.K.. LEBOWITZ. M.D. LonQi- tudinal changes in forced expiratory volume in one second in adults. Effects of smoking and smoking cessation. American Review of Respiratory Disease 135(4):794-799, April 1987. CAMNER. P.. PHILIPSON, K., ARVIDSSON. T. Withdrawal of cigarette smoking: A study on tracheobronchial clearance. Archives of Environmental Health 26:90-92. February 1973. CARSTENSEN, J.M., PERSHAGEN, G., EKLUND, G. Mortality in relation to cigarette and pipe smoking: 16 years' observation of 25,000 Swedish men. Journal of Epidemiology and Cnmmcrnin• Health 41: l 66-172. 1987. CASALE, T.B., RHODES, B.J.. DONNELLY, A.L., WEILER, J.M. Airway responses to methacholine in asymptomatic nonatopic cigarette smokers. Journal of Applied Ph.ysiology 62(5):1888-1892, 1987. CENTERS FOR DISEASE CONTROL. Chronic disease reports: Chronic obstructive pul- monary disease mortality-United States, 1986. Morbidity and Mortality Weeklv Report 38(32):549-552, August 18. 1989. CERVERI, L. BRUSCHI, C., ZOIA. M.C., MACCARINI, L., GRASSI, M.. LEBOWITZ. M.D., RAMPULLA, C., GRASSI, C. Smoking habit and bronchial reactivity in normal subjects. A population-based study. American Reriew of Respiraton• Disease 140:191-196, 1989. CHAN-YEUNG. M.. VEDAL, S.. KUS. J., MACLEAN, L., ENARSON. D.. TSE, K.S. Symptoms, pulmonary function, and bronchial hyperreactivity in western red cedar workers compared with those in office workers. American Review of Respiratoq Disease 130:1038- 1041, 1984. TITVIN 438747 353

HAMMOND, E.C., HORN, D. Smoking and death rates-Report on forty-four months of followup of 187,783 men. Journal of the American Medical Association 166(11):1294-1308, March 15, 1958. HARRIS, J.O., OLSEN, G.N., CASTLE. J.R., MALONEY, A.S. Comparison of proteolytic enzyme activity in pulmonary alveolar macrophages and blood leukocytes in smokers and nonsmokers. American Review of Respit•atoty Disease 111(6):579-586, May 1975. HARRIS, J.O., SWENSON, E.L.W., JOHNSON, J.E. Human alveolar macrophages: Com- parison of phagocytic ability, glucose utilization, and ultrastructure in smokers and non- smokers. Journal of Clinical Investigations 49(11):2086-2096, November 1970. HAWTHORNE. V.M., FRY, J.S. Smoking and health: The association between smoking behaviour, total mortality, and cardiorespiratory disease in west central Scotland. Journal of Epidemiology and Community Health 32:260-266, 1978. HAYNES, W.F. JR., KRSTULOVIC, V.J., BELL, A.L. JR. Smoking habit and incidence of respiratory tract infections in a group of adolescent males. American Review of Respiratory Disease 93(5):780-785, May 1966. HEISKELL, C.L., MILLER, J.N., ALDRICH, H.J., CARPENTER, C.M. Smoking and serolo8ic abnormalities. Journal of the American Medical Association 181:88-91, 1962. HERSEY, P., PRENDERGAST, D., EDWARDS. A. Effects of cigarette smoking on the immune system. Follow-up studies in norma subjects after cessation of smokinQ. Medical Journal of Australia 2(9):425-429, October 29, 1983. ~ HIGENBOTTAVi. T.. SHIPLEY, M.J.. CLARK. T.J.H., ROSE. G. Lung function and symptoms of cigarette smokers related to tar yield and number of cigarettes smoked. Lancer 1(8165):409-1• 12, February 23, 1980. ~ HIGENBOTTAM, T.W., FEYERABAND, C., CLARK, T.J.H. Ciaarette smoking in asthma. British Journal of Diseases of the Chest 74:279-284, 1980. HIGGINS, I.T.T., HIGGINS, M.W.. LOCKSHIN. M.D.. CANALE, N. Chronic respiratory diseases in minina communities in Marion County, West Virginia. British Journal of Industrial Medicine 25:165. 1968. HIGGINS, M.W., KELLER. J.B.. METZNER, H.L. Smoking, socioeconomic status and chronic respiratory disease. American Review of Respiratory Disease 116(3):-103--110. September 1977. HIGGINS, M.W., KJELSBERG. M. Characteristics of smokers and nonsmokers in Tecumseh. Michigan. II. The distribution of selected physical measurements and physiologic variables and the prevalence of certain diseases in smokers and nonsmokers. American Journal of Epidemiology 86(1):60-77, 1967. HILLERDAHL. G., RYLANDER. R. Asthma and cessation of smoking. Clinical Alleroy 14:45-47, 1984. HOIDAL, J.R.. FOX. R.B., LEMARBRE, P.A., TAKIFF, H.E., REPINE, J.E. Oxidative metabolism of alveolar macrophages from young asymptomatic cigarette smokers. Increased superoxide anion release and its potential consequences. Chest 77(2, Supplement):270-271, February 1980. HOIDAL, J.R.. FOX, R.B., TAKIFF, H.E., REPINE. J.E. Alveolar macrophacres (AM) from young, asymptomatic cigarette smokers (CS) and nonsmokers (NS) use equal amounts of oxygen (02-) and glucose (1-14C), but smoker AM make more superoxide anion. (Abstract.) American Review of Respiratory Disease 119(4, Supplement):222A, April 1979. HOIDAL, J.R., NIEWOEHNER, D.E. Lung phagocyte recruitment and metabolic alterations induced by cigarette smoke in humans and in hamsters. American Review of Respiratory Disease 126:548-552, 1982. HOLT, P.G. Immune and inflammatory function in cigarette smokers. Thora.r 42:241-249. 1987. TIMN 438751 357

CONTENTS Partl. Female .....................................................371 Pregnancy and Pregnancy Outcome .................................. 371 Introduction ....................................................371 Pathophysiologic Framework ..................................... 371 Nonexperimental Studies ......................................... 374 Fertility and Infertility ......................................... 374 Ectopic Pregnancy and Spontaneous Abortion ...................... 375 Fetal, Neonatal, and Perinatal Mortality ........................... 376 Birthweight and Gestational Duration ............................ 379 Introduction ................................................ 379 Continued Smoking .............. :........................... 381 Cessation Before Conception .................................. 382 Cessation After Conception ................................... 383 Birthweight ............................................. 383 Preterm Delivery ........................................ 386 Complications of Pregnancy ................................... 387 Randomized Trials of Smoking Cessation During Pregnancy ............ 387 Prevalence of Smoking and Smoking Cessation During Pregnancy and Time Trends in Prevalence and Cessation ............................... 390 Introduction .................................................390 Prevalence of Smoking and Smoking Cessation ................... . . 390 Time Trends in Smoking and Smoking Cessation ................... 393 Estimates of Attributable Risk Percent .............................. 393 Age at Natural Menopause .......................................... 396 Introduction ...................................................396 Pathophysiologic Framework ..................................... 396 Studies of Former Smokers ......................................... 98 Part1l. Male ....................................................... 400 Introduction ..................................................... 400 Pathophysiologic Framework ........................................ 400 Sexual Activity and Performance .................................... 401 Sperm Density and Quality ......................................... 404 Conclusions ........................................................ 410 References ........................................................ 411 369 TIMN 438762

TABLE 3.-Summary of studies of perinatal and neonatal mortality in smokers and nonsmokers during pregnancy N b f Perinatal mortality Neonatal mortality' Reference um er o births Category Smokers Nonsmokers Smokers Nonsmokers Yerushalmy 6.800 Whites 13.9 12.4 (1964) Blacks 22.0 23.4 Comstock and 12,287 23.e 15.6' Lundin (1967) Amount smoked Meyer and 51,490 <1 ppd 28.0 23.0 Tonascia ? t ppd 33.4 (1977) Social class` Rantakallio 12.068 1+11 28.1d 22.4' (1978) III+IV 25.1 19.6 Farmers 25.5d 39.0d Unknown 29.4`t 36.8d Amount smoked Rush and <5 cig/day 15.9 18.7 Cassano 5-14 cig/day 26.1 (1983) >15 cig/day 28.3 Butler. 21,788 41.1 32.0 17.6 13.7 Goldstein, Ross 1972) Amount smoked Andrews and 18,631 1-4 cig/day 25 24 McGarry 5-9 cig/day 20 (1972) 10-19 ciglday 32 _20 cig/day 36 Race and Amount smoked Niswander 37,912 White 31.4 and Gordon 1-10 cig/day 31.5 (1972) -11 cig/day 38.2 Black 38.5 1-10 cig/day 41.5 ?11 cig/day 57.4 377 TIMN 43100769

STEIN, R., EVANS, S., MILNER, R., RAND, C., DOLOVICH, J. Isotopic and enzymatic IgE assays in non-allergic subjects. Allergy 38(6):389-398, August 1983. SUCIU-FOCA, N., MOLONARO, A., BUDA, J., REEMTSMA. K. Cellular immune respon- siveness in cigarette smokers. (Letter.) Lancet 1:1062, May 25, 1974. TABONA,M., CHAN-YEUNG, M., ENARSON, D., MACLEAN, L., DORKEN, E., SCHUL- ZER, M. Host factors affecting longitudinal decline in lung spirometry among grain elevator workers. Chest 85(6):782-786, June 1984. TAGER, I.B., SEGAL, M.R., SPEIZER, F.E., WEISS, S.T. The natural history of forced expiratory volumes-Effect of cigarette smoking and respiratory symptoms. American Review of Respiratory Disease 138:837-838, 1988. TAGER, I.B., WEISS, S.T., MUNOZ, A.. ROSNER, B., SPEIZER, F.E. Longitudinal study of the effects of maternal smoking on pulmonary function in children. New England Journal of Medicine 309:699-703. September 22. 1983. TASHKIN, D.P., CLARK, V.A., COULSON, A.H., SIMMONS, M., BOURQUE, L.B., REEMS, C., DETELS, R., SAYRE, J.W., ROKAW, S.N. The UCLA population studies of chronic obstructive respiratory disease. VIII. Effects of smoking cessation on lung function: A prospective study of a free-living population. American Review of Respiratory Disease 130(5):707-715. November 1984. TAYLOR. R.G.. GROSS. E., JOYCE. H.. HOLLAND, F.. PRIDE, N.B. Smokine,