Tobacco Documents Online

wORLD~OSiF1REL*I~JON SMOKINCLAND H Ar'rw ~ lammng Committee mnuttee on Invntations PREFACE ............................................ CHAPTER I--SETTIhTG THE THEME.................. A TIME FOR ACTION-Luther L. Terry, M.D., Chair- man, World Conference on Smoking and Health, and Vice President for Medical'Affairs, University of Pennsylvania.. Objectives: Exchange information-Stimulate action / Public awareness at peak ADDRESS by The Honorable Robert F. Kennedy ...... Advertising industry self-regulation / Legislation re tar and nicotine / Strong warning in all advertising /"Fair- ness" doctrine /'Local monitoring /'Sliding tax scale. aFTR.R n m 14 ©RLD C DIS- 6,-- bAJL, U1JAliJ1.11. Y ANU UEA1'ri-L?. C hyter tiam- -' mond,, Sc.D., Vice President for EpidemioTogy and Sta- tistics, American Cancer Society...................... 15 Increased illness and disability / Cessation of smoking effect / Statistics on lung cancer internationally / Chronic respiratory disease in children / Higher cancer death rates other sites / Recommendations. CIGARETTES AND CARDIOVASCULAR DISEASE- Jeremiah Stamler, M.D., Associate Professor of Medicine,. Northwestern University Medical School'.............. 44 One million cardiovascular deaths Ui S. 1965 / Cig- arettes major factor-cor pulmonale deaths / Mortality ratios higher for younger age groups / Cigarette smoking problem of 20th century--can be changed. CIGARETTES AND CANCER-George E. Moore, M.D., Director, Public Health Research, New York State Health Department........................................ 74 Clinical research needed re specific sites /Government's task force for lung cancer / Development less harmful cigarette / Need large-scale cooperative clinical trials. CIGARETTES AND RESPIRATORY DISEASE-C. M.__ '" Fletcher, C.B.E., M.D., Reader in Clinical Epidemiology, Kdysl'Postgraduate Medical School, London, Eng. .. 78 England's Respiratory Mortality Rate World's Highest Bronchitis-emphysema syndrome / Smokers chief air pollution victims / Cessation only preventive for ventila- tory incapacity. PROBLEMS IN' CONDUCTING SMOKING RE- SEARCH-Sir Austin Bradford Hill, M.D., Professor Emeritus of Medical Statistics,, University of London.... 92 V THE S IC BACKGROUND

vi Contents Ethics of smoking research dominant problems I Con- ~ sideration of whole population / Profile of smoker un- likely to be answer / Randomization of clinics. CHAPTER III THE BRITISH AND NORWEGIAN EX- PERIENCES .......................................... 97 THE BRITISH EXPERIENCE--Sir George E: Godber, IC.C.B., M.D:, D.P.H., F.R.C.P., Chief Medical Officer, British Ministry of Health, London.................... 97 Differences between cigarette and other preventive medi- cal problems / Three notable reports since 1950 /'British restrictions on advertising / Government Social Survey on habits and attitudes / Suggestions for reaching uncon- vinced. INFLUENCING SMOKING HABITS: A NORWEGIAN CONTRIBUTION-Kar1 Evang, M.D., F:R.SM., F.R.S.L, Director-General, The Health Services of Norway ....... 108 Anti-cigarette campaign based on Committee analysis by multi-disciplinary scientific approach / Smoking levels of various age groups / Choice of media / Restrictive and therapeutic measures / Economic aspect. CHAPTER 1V-II*TFLUENCING SMOKING BEHAVIOR..118 KEYNOTE ADDRESS-William H. Stewart, M.D., Sur- geon General, United States Public Health Service...... 118 People want knowledge transformed into action / Con- cept of reduced tar-nicotine / Specific groups vulnerable to smoking risks / Total abolition impracticable / Future guidelines. HOW DID SOCIETY GET INT'O THE CIGARETTE MESS? WHY IS 1T SO HARD TO FIND A WAY OUT? -Daniel Horn, Ph.D., Director, National, Clearinghouse for Smoking and Health, U. S. Public Health Service ...... 126 60,000 puffs annually for pack-a-day smoker / Resolu- tion of earlier forms of gratification behavior / Factors involved in initiation and cessation / Solution through behavioral approach. PANEL DISCUSSION OF PSYCHOLOGICAL AS- PECTS OF SMOKING-A. C. MeKennell, Ph.D., Pro- fessor of Psychology, The University of Southampton, Southampton, England .............................. 133 Smoker vs. non-smoker pattern set by age 20 / One-half smokers want to stop / Light smokers best target Richard L. Foster, Ph.D., District Superintendent, San R'amon Valley Unified School District, Calif. ..........135 Programs require strong public relations pitch / Lesson learned best by teaching people what they have to teach, Silvan S. Tomkins, Ph.D., Director, Center for Research in Cognition and Affect, City Univeisity of New York, N. Y.. 136 The habitual smoker / The positive affect smoker / The negative affect smoker / The addict. Godfrey M. Hochbaum, Ph.D., Chief, Behavioral Science Section, Bureau of Health Services, USPHS ............ 1Ci8 Smoking has values for most smokers / Reduction may be only effective approach to certain smoketa

Contents vii CHAPTER V-WORK GROUPS: POSTTTON PAPERS ....140 Work Group 1 Addiction, Habituation, Pharmacology Work Group 2A Behavioral Problems and Progress Work Group 2B Behavioral Problems andpro ess Work Group 8 Role of Physician and Other Exemplars Work Group 10 Commuuicstions-The Media TOBACCO-HABIT AND ADDICTION-William A. Hunt, Ph.D., Chairman, Department of Psychology, Loy- ola University, Chicago, Illinois and Joseph D. Matarazzo, • Ph.D., Professor and Chairman, Department of Medical Psychology, University of Oregon Medical School, Port- land .............................................. 140 Paradigm of tobacco habittiation / Need A.A, type of supportive therapy / Smoking over-learned behavior. THE PHARMACOLOGICAL BASIS OF ADDICTION TO TOBACCO: NICOTTNE--Murray E. Jarvik, M.D., Ph.D., Albert Einstein College of Medicine, New York.... 142 Chemical cause underlies addiction / Effect of nicotine on brain not fully known / Drug to mimic or antagonize nicotine action / Research needed' to help addicted. A. C. McKennell, Ph.D. The University of Southampton, England (see Chapter I~, p. 133) Silvan S. Tomkins, Ph.D., City University of New York..143 Smoking's function control of affective information / Face primary site of human feelings / Smoking a tech- nique to help relieve suppressed emotions I Important toxic consequences of smoking. Work Group 3 Towards A Less Harmful Cigarette (Proceed- ings published separately by the National Cancer Instttute, Monograph No. 28, FVashington, D. C., June 1968) Work Groups 4A & 4B School Programs: Program Content, Materials, Ages to Reach Work Group 6 Teacher Edncation SMOKING EDUCATION: WHEN, WHERE, AND HOW-Ira Gordon, Ph.D., Institute of Human Develop- ment, University of Florida, Gainesville ...... .145 Problem is teaching non-behavior / Strong self-value concept important l' Child should see non-smoker re- warded' / Teacher behavior persuasive. THE SCHOOL AND SMOKING-AN EXERCISE IN FRUSTRATION (?)-Louise E: Hock, Ph.D., New York University, New York, N. Y. ........................ 161 Modifying habits of an entire people / Guiding principles / / Specific suggestions / Methodology and materials Greater promise lies with models. CIGAREfiI'E SMOKING, RESPIRATORY SYMPTOMS AND ANTI-SMOKING PROPAGANDA, AN EXPERI- MENT-W. W. Holland, M.D., B.Sc., Reader in Clinical Epidemiology and Social Medicine, St. Thomas's Hospital Medical School, London and A. ELLIOTT, MD., D.P.H. County Medical Officer, Kent County Council, England.. 169 Smoking habits children under age 13, over 14. Work Group 5 College Programs THE PRESENT SITUATION IN COLLEGES AND UNIVERSITIES AND A LOOK AT THE FUTURE-

viii Contents W. T. Robbins, M.D., Director, Student Health Service, University of California, Santa Barbara, Calif.......... 171 Behavior-changing activities / Projected study among 50,000 students / Survey of college policies re campus smoking. Work Groups 7A & 7B Giving Up Cigarette Smoking NEW YORK CITY SMOKING WITHDRAWAL CLINIC -Donald T. Fredrickson, M.D., Director, Smoking Con- trol Program, New York City Department of I3eaith..... 187 Purpose and design / Staff of volunteer ex-smokers / Orientation / Withdrawal / Reinforcement / Results. REPORT ON WITHDRAWAL CLINICS-Borje E. V. Ejrup, M.D., Clinical Associate Professor of Medicine, New York Hospital, Cornell Medical Center........... 198 Program for hard core smokers / Approach for use by general practitioner / Lobeline hydrochloride injection / Preventing relapses. SMOKING WITHD RAWAL IN MEDICAL PRACTICE -SOME EPIDEMIOLOGICAL ATTRIBUTES OF SMOKING-George Christakis, M.D., Assistant Dean and Associate Professor, Community Medicine, Mt. Sinai School of Medicine, New York........................ 207 Results of study of smoking on Island of Crete. RESULTS OF AN ANTI-SMOKIPIG CLII+lIiG-Keith P. Ball, MD., F.R.C.P„ and Miller Mair, M.A., Dip. Psych., Ph.D., Central Middlesex Hospital, London ............ 208 Method of' operation / Four implications for treatmentl AN' ATTEMPT TO DISCUSS' THE COST-BENEFIT PROBLEM RELATED TO SMOKING WIITFIDRAWAL CLINICS AND COURSES-Kje11 Bjartveit, M.D., Senior Medicai'OfEoer, National Mass Radiography Service, Oslo. 210 Death rates of 100,000 smokers vs. 100,000 non-smokers. THE ROLE OF THE PFIYSICIAN IN THE CONTROL OF SMOKING-Judith S. Mausner, M.D., Assistant Pro- fessor of Epidemiology, Woman's Medical College, Phila- delphia, Pennsylvania and Bernard Mausner, Ph.D., Pro- fessor of Psychology, Beaver College, Glenside, Pennsyl- vania .............................................211 Physician influence potentially greater than mass media or clinics / How to increase physic~an participstion. IMPLICATIONS FOR FUTURE TREATMENT PRO- GRAMS:-Jerome L. Schwartz, D:P.H~., Project Director, and Mildred Dubitzky, Ph.D., Research Psychologist, Smoking Control Research Project, Berkeley, California..216 Control methods / Selecting ants / Evaluating k Gro~ap / Fo Role of PhysiWhat dan ~a d«Othf ~~ efit7 Wor THE PHYSICIAN AS EXEMPLAR-Richard H. Over- holt, M.D., Director, Overholt Thoracic Clinic, Boston..226 Four constructive steps for physicians. CIGARETTE SMOKING: MAGNTITJDE OF THE PROBLEM-R. T. Ravenholt, M.D., M.P.H., Director, Population Service, Office of the War on Hunger, Agency for International Development ....................... 227 Smoking by medical;: dental, and nursing faculty.

Contents AN ANTISMOKING PROGRAM FOR TEACHERS- Eva J. Salber, M.D. and Theodore Abelin, M.D., Depart- ment of Epidemiology, School of' Public Health, Harvard University ........................................ 228 Clinic attendance by high school teachers related to su- perior knowled~e re smok~ng nsks. Work Group 9 Government Actioo and LegiShBon RESULTS OF GOVERNMENT ACITON AND LEGIS- - LATION-Leo Noro, M.D., Director, Institute of Occu- pational Health, Helsinki ............................ 230 No advertising / Eight governmental activities. REMARKS-David '1~ Carr, M.D., Professor of Clinical Medicine, Mayo Clinic, Rochester, Minnesota.......... 231 Smoking by sports figures and TV actors possibly more inflnentialthan commercial advertising. THE PRESENT STT4JATION IN ITALY-Professor Carlo Vetere, Director, Division of Health Education, National Ministry of Health, Rome................... 232 No advertising / General increase in sales / Ban on cigarette smoking m movie theatres opposed. WHAT OTHER CiOVERNMENT ACTION IS NEED- ED? Lester Breslow, M.D., Director, California State Department of Public H'ealth, Berkeley................. 232 Problem intensified by availability of cigarettes and longevity / Must solve economic problems of industry. REMARKS ON F.C.C. "FAIRNESS DOCTRINE" RUIr ING-John F. Banzhaf; III, Attorney, New York, N. Y.. .236 Attack on health agencies /'Request for support of Con- ference delegates in enforcement of ruling. RESULTS OF GOVERNMENT ACTION' AND LEGIS- LATION-Gyorgy Karpatf, M.D., Leader of Fight Against Cancer, Minister of Health, Budapest, Hungary.......... 237 Anti-alcoholism program strengthens anti-smoking cam- aign / Influence of tobaozo on male genital organs. T OTHER GOVERNMENT ACTION IS NEED- ED?: Vsevolod Bilyk, M.D., Inspector General, Minis- try of Health and Social Welfare, Bucharest, Ronmania..237 Smokers compnlsory checkup / Nicotinelesa cigarettes. Work Group 10 Comm~katlons-The Malla CHANGES IN ADVERTISING EXPENDITURE AND SMOKING BEHAVIOR AFTER THE BAN ON' TELE- VISION ADVERTISING IN THE UNTTED KINGDOM -John Wakefield, Head, Department of Social Research, Christie Hospital & Holt Radium Institute, England.....238 Increased cigar, pipe advertising /'TV revenue loss made up by other products / Coupon schemes popular / 3%'o increased smoking in 16-19 year-old agegrou TALK TO THEM IN THEIR OWN LANGUAG~Tony Schwartz, New Sounds, Inc., New York, N. Y............ 239 Special techniques to reach young people. THE NORWEGLAN EXPERIENCFf-Ottar S. Jacobsen, Secretary-General, Norwegian Cancer Society.......... 240 No radio and TV advertising / Restrictions on cigarette advertising in other media. WHAT TO DO ABOUT CIGARETTE ADVERTISING

a Contents -Emerson Foote, New York, former Chairman, National Interagency Council on Smoking and Health...........242 $300 million annual advertising expenditure /' F.C.C. Fairness Doctrine / Advertising Council sponsorship ant9=smoking campaign? CHAPTER VI-WHAT SHOULD SOCIETY DO TO CON- TROL CIGARBTI'E SMOKING? PANEL DISCUSSION OF WORK GROUP RECOMMENDATIONS The Honorable Frank E. Moss, U. S. Senator from Uteh.250 Education, not prohibition / Pending legislation. Sir George E. Godber, K.C.B., M.D., D.P.H., F.R.C.P.....251 Conference results in new and definite message for public. Karl Evang, M.D., F.R.S.M., F:R.S.L,, F.A.P.FLA....... 252 "Our platformm is strong" / In Norway, fear not a deter- rent to youth not moved by statistical evidence / Involve WHO, UNESCO, etc. William H. Stewart, M.D., U. S. Surgeon Generai....... 254 Recommendations point to three-part program. Ashbel C. Williams, M.D., President, American Cancer Society (1967) .................................... ,255 Establish research institute to get new answers. Arthur T. Roth, Board Chairmani Franklin National Bank.255 Employees smoking forbidden on basis of profits. Ernest L. Wynder„M.D., Memorial Sloan-Kettering Cancer Center .......................................... 256 Impractical to remove all. tar and' nicotine. Luther L. Terry, M.D., Conference Chairman........... 256 The challenge-to save worldwide loss of life. CHAPTER VII-WORK GROUP RECOMMENDATIONS,.258 Work Group 1 Addktlon, Habituation, Pharmacology of Tobacco Work Group 2A Behavioral Problems and Progress Work Group 2B Behavioral Problems and Progress Work Group 3 Towards A Less Harmful Cigarette Work Group 4A School Programs: Program Content; Materi'- a1s, Ages to Reach Work Group 4B School Programs: Program Content, MaterL als, Work Groap 5 College Work Group 6 Teacher Education Work Group 7A Giving Up Cigarette Smoking Work Group 7B Giving Up Cigarette Smoking Work Group 8 Role of Physician and Other Exemplars Work Group 9 Government Action and Legislation Work Group 10 Commmntcations-Tbe Media LIST OF PARTI'CIPANTS AND ADDRESSES............285 l

Robert F. Kennedy 13 enacted right away. For the industry we seek to reg- ulate is powerfuli and resourceful. Each new effort to regulate will bring new ways to evade, just as the television advertising ban in Britain brought forth an intensified coupon war to promote smoking. Still, we must be equal to the task. For the stakes involved are nothing less than the lives an& health of millions all over the world. But this is a battle which can be won-and with the commitment that is dem- onstrated by this conference; with the commitment that all of you show in being here and in your work at home-I know it is a battle which will be won.

CHAPTER II The Scientific Background The first session of the Conference then heard a sum- mary of the most recent research findings in a number of fields: epidemiology, cardiovascular disease, cancer, res- piratory disease, and problems of conducting smoking re- search. The first of these scientific talks, by E. Cuyler Hammond, SC.D., Vice President f or Epidemiology and Statistics of the American Cancer Society, was delivered on behalf of the following international Committee on Epidemiology: Johannnes Clemmesen, D.M.Sc., Director, Danish Cancer Registry, Copenhagen, Denmark H. N. Colburn, M.D., M.P.H., Medical Consultant, Smoking and Health Program, Department of National Health and Welfare, Ottawa, Canada Charles M. Fletcher, C.B.E., M.D., F.R.C.P., Postgrad- uate Medical School, London, England William M. Haenszel, M.A.,, Chief, Biometry Branch, National Cancer Institute, Bethesda, Maryland E. Cuyler Hammond, Sc.D., Vice President for Epide- miology and Statistics, American Cancer Society, New York, N.Y. Takeshi Hirayama, M.D., Chief, Epidemiology Divi sion, National Cancer Center, Tokyo, Japan W. W. Holland, M.D., B.Sc., Department of Clinical Epidemiology & Social Medicine, St. Thomas' Hos- pital Medical School, London, England S. Koller, M.D., Ph.D., Johannes Gutenberg Univer- sity, Mainz, Germany Naum Marchevsky, M.D., Office of Secretary of Public 14

E. Cuyler Hammond 15 Health, flipolito Irigoyen 370, Buenos Aires, Argen- tina,, South America Torbjorn Mork, M.D., Ph.D., Deputy Director, Cancer Registry of Norway, Oslo, Norway Jerzy Staszewski, M.D., Institute of Oncology, Gliwice, Poland C. B. Walker, Principal Research Officer, Bio-Statistics Division, Department of National Health & Welfare,, Ottawa, Canada. Dr. Hammond's paper, on worl& costs of cigarette smoking in disease, disability and death, follows: Since early this summer our Committee has been working on the preparation of material for this ses- sion. During July, small preliminary meetings were held in Oslo, Copenhagen, London and Mainz. Members of the Committee have worked diligently in reviewing a tremendous volume of material pub- lished in many different languages. In addition, sev- eral of the members analyzed original data from new studies as well as more extensive data resulting from several more years of follow-up of subjects in large prospective epidemiological studies. Some of this new material will be presented this morning. This last Saturday and Sunday all but two of us met in New York, discussed our joint' findings and prepared this report. We will first present findings on totali mortality and total morbidity and then present findings in relation to three important diseases: coronary heart disease, lung cancer and other lung diseases. We will men- tion other diseases more briefly, discuss the problem of extrapolation to countries where studies have not been made and end by presenting several reeommen- dationsforfuture research. Total Mortality In the past, the effects of smoking have usually been expressed in terms of mortality ratios. For example, it has been said that the death rate of heavy cigarette

16 The Scientific Background 1 smokers is two to three times as high as the death rate of non-smokers. Now we have data covering a sufficiently wide age range to be able to construct life tables in relation to smoking habits. We have chosen to use this form of presentation today. It will be noticed that slightly different groupings by age and by amount' of smoking have been used in dif- ferent countries. This prevents direct comparison. Life Table for 35 year old Men Estimate from British Doctor Study Non Current No. Cig.A Day Age Smokers 1-14 15-24 25+ $ ~ $ $ 00. 100.0 100.0 100.0 40 99.5 99.2 99.2 97.8 45 98.6 98.]. 97.9 95.1 50 96.3 '95.4 94.4 90.4 55 94.3 91.6 89.9 84.5 60 88.7 83.8 81.1 74.3 65 81.9 74.3 71.5 62.9 70 69.7 58.4 57.7 46.2 Slide 1 From Doll and Hill's study of British doctors. This slide (#1) shows life tables for 35-year-old men constructed' from findings during 12 years of' follow-up of 34,000 British physicians in the study carried out by Doll and Hill. 69.7% of male British physicians who never smoked regularly may be ex- pected to live to the age of 70. In contrast, only 58.4% of those who smoke 1 to 14 cigarettes a day, 57.71o of those who smoke 15 to 24 cigarettes a day, and 46.2 jo of those who smoke 25 or more cigarettes a day may be expected to live to that age. The next slide (#2) shows life tables for male Canadian war veterans. It is based upon findings among 78,000 men traced for six years in a study undertaken by Best, Walker and others of the Cana- dian Department of National Health and Welfare.

Life Table For 32.5 Year Old Men .EOSRestzoz Estimate For Canadian Study Never Cigarette Smokers By. Daily Amount Smoked Age Regularly 1-9 o~p 32.5 100.0 % 100.0 J 67.5 71.5 72.5 60.2 77.5 46.8 84.3 29.2 Life Expectancy: Age 72.1 Years 59.6 Years Lost 0 63.2 48.8 32.6 14.8 68.5 36.0 3.6 10-20 % 100.0 57.8 43.7 30.8 13.9 67.2 34.7 4.9 21+ % 100.0 56.3 42.2 28.1 13.3 66.4 33.9 5.7

18 The Scientific Backgroun& The figures in the body of the table show the chances such 32.5 year-old-men have of living to various ages depending upon their smoking habits. 71.5% of those who never smoked regularly may be expected' to live to the age of 67.5. In contrast, only 56.3% of those who smoke 21 or more ciga- rettes a day may be expected to live to that age. 29.2% of the non-smokers, but only 13.3% of those who smoke 21 or more cigarettes a day may be ex- pected to live to the age of 84.3. As shown at the bottom of the table, 32.5-year- ol& Canadian veterans who never smoked regularly have a remaining life expectancy of 39.6 years. In contrast, those who smoke 21 or more cigarettes a day have a remaining life expectancy of only 33.9 years. This is a difference of 5.7 years. Findings in independent' studies carried out in the United States are very close to findings in the study of Canadian veterans. This slide (# 3) shows life tables for 25-year-old American men. It is based upon findings from a study of 447,000 ~ men enroiled in an epidemiological study by volunteer workers of the American Cancer Society. 39,178 of these men died during the five year period July 1, 1960 through June 30, 1965. The figures shown here were adjusted to the 1959- 1961 United States life table for men. 77:7% of 25-year-old non-smokers may be ex- pected to live to age 65. Only 54.0 fa of 25-year-old men who smoke (and' continue to smoke) 40 or more cigarettes a day may be expected to live to that age. 119.2% of 25-year-old non-smokers may be ex- pected pected to live to the age of 85. Only one-third that many men who smoke 40 or more cigarettes a day may be expected to live to that age. Lighter smokers survive longer than heavy smok- ers, but even smoking 1 to 9 cigarettes a day short- ens life expectancy. As shown at the bottom of the table, 25-year-ol&

Age 80 85 Life Table For 25 Year Old Men Estimate For United States Never Cigarette Smokers By Daily Smoked Regularly 1-9 10-19 20-39 100.0 100.0 100.0 100.0 .... .... .... ... 77.7 67.3 63.4 61.1 66.7 52.4 47.7 45.9 52.3 36.2 33.3 30.3 35.6 20.6 18.6 18.1 19.2 7.3 8.5 7.2 Amount 40+ 100.0 . 54.0 40.0 25.7 14.3 6.5 ~ Life Expectancy: Er .~ Age 73.6 69.0 68.1 67.4 65.3 ° Years 48.6 44.0 43.1 42.4 40.3 0 'soseSsIVzoz ~ Years Last 0 4.6 5.5 6.2 8.3

Life Expectancy (Years) At Various Ages Estimate For United States Males Never d Cigarette Smokers By Daily Amount Age Smoke Regularly 1-9 10-19 20-39 40 + 25 48.6 44.0 43.1 42.4 40.3 30 43.9 39.3 38.4 37.8 35.8 35 39.2 34.7 33.8 33.2 31.3 40 34.5 30.2 29.3 28.7 26.9 45 30.0 25.9 25.0 24.4 23.0 50 25.6 21.8 21.0 20.5 19.3 55 21.4 17.9 17.4 17.0 16.0 60 17.6 14.5 14.1 13.7 13.2 65 14.1 11.3 11.2 11.0 10.7 N O 9o9RestVzoz

Loss` In Life _ Expectancy At Various Ages Estimate For United States Males ----- - --- Cigarette Smokers By Daily Amount 1-9 a day 10-1.9 a day 20-39 a day 40+ a day Age ---- Years % Years °a Years % Years % 25 4.6 9.5 5.5 11.3 6.2 12.8 - 8.3 17.1 30 4.6 10.5 5.5 12.5 6.1 13.9 8.1 18.5 35 4.5 11.5 5.4 13.8 6.0 15.3 7.9 20.2 40 4.3 12.5 5.2 15.1 5.8 16...8 7.6 22.0 45 4.1 13.7 5.0 16.7 5.6 18.7 7.0 23.3 50 3.8 14.8 4.6 18.0 5.1 19.9 6.3 24.6 55 3.5 16.4 4.0 18.7 4.4 20.6 5.4 25.2 60 3.1 17.6 3.5 19.9 3.9 22.2 4.4 25.0 65 2.8 19.9 2.9 20.6 3.1 22.0 3.4 24.1 ~ Life expectancy of inen who never smoked regularly taken as 0 standard.

22 The Scientific Background men who never smoked regularly have a remaining life expectancy of 48.6 years. In contrast, those who smoke 40 or more cigarettes a day have a remaining life expectancy of only 40.3 years. This is a differ- ence of 8.3 years. This slide (#4) shows remaining years of life ex- pectancy for men of various ages in relation to their smoking habits. Of' course, as we grow older, we have less years of life remaining to us; but at alli ages, non-smokers may look forward to more years of life than cigarette smokers. This slide (#5) shows the difference between the life expectancy of non-smokers and the life expec- tancy of cigarette smokers of various ages. In age range 35 through 65, men who smoke, (and con- tinue to smoke) 40 or more cigarettes a day have 20% to 25of'o less years of life remaining to them than are remaining to their friends who never smoked r aaularly. Light smokers are not as badly off' in this respect as heavy smokers. Men who started to smoke cigarettes early in their youth tend to smoke more cigarettes per day and~ tend' to inhale the smoke more deeply than men who started to smoke later in life; and they have correspondingly shorter life expectancies. This is shown in this slide (# 6) where the cigarette smok- ers are clbssified' by the ages at which they began to smoke. Now let us consider (slide #7) the impact of cig- arette smoking upon the male population of the United States as a whole. The column at the left of this table shows the survivorship of 25-year-old men who never smoked regularly. That at the right shows the survivorship of all 25-year-old American men. The figures imply that' if it were not for cigarette smoking, life expectancy for 25-year-old' American men would be 3.4 years longer than it is today. If you are inclined to think that this is a small differ- ence, consider the following figures:

so9e9stzoz Ufe Ta . le For 25-Year-Old Men Estimate For United States ~a Never k d S Cigarette Smokers By Age Began Smoking ~ a~ mo e Age Regularly 25-34 20-24 15-19 <15 % % % % % ~~ 25 100.0 100.0 100.0 .100.0 *100.0 .. 65 77.7 67.2 65.0 60.2 55.5 ~~ch 70 66.7 52.5 50.8 44.7 39.7 a; 75 52.3 38.7 35.1 30.0 24.6 o a 80 35.6 24.6 20.4 16.8 14.0 p o. 85 19.2 10.5 9.1 6.3 5.8 o --- ~ Life Expectancy: ~ Age 73.6 69.9 68.8 67.1 65.4 0 o Years 48.6 44.9 43.8 42.1 40.4 < ~ Years Lost 0 3.7 4.8 6.5 82

OT9886tiZoz Survivorship Of 25-Year Old Men United States Males b. ~ W~ Never Smoked Reculariy A I I Men n % Surviving ~ ~ 25 100.0 100.0 ~ ,~ ., .... .... p ~ P., & ~ 65 77.7 67.8 ~ n&n p 70 66.7 A 55.2 75 52.3 ~ 41.2 ~ ~ ~ 80 35 6 26.7 . ~ ~ ~ 85 19.2 13.6 ~ p p ~ a~ Life Expectancy Age 73.6 a 70.2 ~ ~ Years 48.6 45.2 "° p~ Years Lost 0 3.4

E. Cuyler Hammond Life Expectancy At Age 25 United States 1919-1965 Year White Male 1919-1921, 41,6 1'929-1931 4 1.8 1939-1941 43.31 1949-1951 44.9 1959-1961 45.7 1963 45.5 1965 45.6 25 N'on-White Male 35.5 32.7 35.9 39.5 41.4 40.7 40.7 1960-1!965 Estimated Years Lost Due To Cigarette Smoking (total U.S. Males)---- 3.4 Slide 8 Vital Statistics of the United States and the American Cancer Society's epidemiology study of'over one million men and women, This slide (#8) shows trends in the life expec- tancy of' 25-year-old American men~ from 1919 to 1965. This has been a half century of tremendous ad- vances in medicine, public health and the American standard of living. The sulfa drugs and antibiotics were developed; infectious diseases, pneumonia, and tuberculbsis were brought under control, and new miracles of surgery were introduced. As a result, life expectancy increase& by 4.0 years for white males and 5.2 years for non-white males. During this same period, cigarette smoking in the United States became popular and increased by leaps

.26 The Scientific Background and bounds. It is estimated that approximately 3.4 ~years of life expectancy are now lost due to the habit'. a Thzs loss is not far short of the net gain~ from halff a century of scientific and sociali progress. Were it not for this, it appears that life expectancy of American men would have increased by about 7.4 years for white males and 8.6 years for non.white males. Most distressing is the fact that the life expectancy of American men has shown no improvement since 1959-1961. This flattening off has been observed in, other countries in recent years; and.in some countries life expectancy of males has actually started to de- crease. Morbidity Information on morbidity among smokers and non-smokers has recently been reported by the US. National Health Survey of Cigarette Smoking and Health Characteristics. Data on smoking habits were collected for all persons aged 17 and over in a sam- ple of 42,000 families interviewed during the 12 months ending June, 1965 about episodes of illness and disability. Three conventional indices of morbid- ity were used; days lost from work, days of re- stricted activity and bed days. All three yielde& essentially similar findings. This slide (#9) summarizes the findings on "days lost to work." The risks expressed as days lost per person-year were higher among cigarette smok- ers than among non-smokers. This was found in both sexes under age 64. No such difference was ob- served among males over 65, but this point could not' be investigated for females over 65 because of insufficient data. The information on bed-days is consistent with that on work days lost in indicating the excess morbidity reported' for smokers to be con- centrated at ages under 65 and to be negligible at the older ages. The contrast between cigarette smokers and non- smokers was further accentuated when the smokers

E. Cuyler Hammond DATA DERIVED FROM THE NATIONAL HEALTHISllRYEY (by sex, by age) 27 Male e ma e 17- 45- and 65 17- 45- and ~ 4'i 64 over 44 64 bver HQRK-Lf15 nA 8 . Estimatedtotal.days (millionsl----- I112 127 121 80 55 4 ~ Rate:------------------------ - ___ Never smoked ciqarettes---- ----- 3.4' 5.6 .8' 4.5 5.3 5.D , History.of cigarettes----------- 4.4 8.5 _8 6.5 6.9 I_ Morbidity ratio-------- ------------ 1.3 1.5 .O 1 .4 1.3 I Difference in morbidityrates------- 1.0 2.9 0 i2.0 1.6 I- 'I Exeess dayss as percentage of total-- . 20 28 0 ' 18 11 I- Slide 9 U.S. Public Health Service. The Health Consequences of Smoking: A Public Health Service Review: 1967. P.H.S. Publication No. 1696. were subdivided by rate of cigarette use. The excess morbidity among smokers increased directly with amount smoked. Recent preliminary findings from a new study in Japan show similar results. The experience of smokers and non-smokers can be compared in various other ways. For exampie, the excess days lost by smokers may be expressed as a percentage of the total days lost. During the study year an estimated 399 million work days were lost in the United States, of whicL a total of 77 million days or 19 percent;, represented the excess work-days lost attributable to the higher rates among cigarette smokers. The overall figure of 19 percent is a weighted~ average of the several sex-age groups. Attention is called to the fact that each of the three contrasts summarized in the table emphasizes a different aspect of morbidity relationships; but each of them leads to the same conclusions. The National Health Survey also reported on~ symptoms described by the respondents. Such data contain many diagnostic errors and uncertainties and' must be interprete& cautiously. Nevertheless, it is of

28 The Scientific Background interest to note that' the symptoms descriptive of chronic bronchitis and emphysema, heart disease and peptic ulcers accounted for a large proportion of the excess morbidity among cigarette smokers. These diseases have been identified in several pros- pective studies as displaying substantially higher mortality among cigarette smokers than among non-smokers. Findings in other studies are consistent with find- ings in the United States National Health Survey. For example, in the American Cancer Society Study previously described, it was found that the per cent' of men hospitalized within a period of two years was higher among cigarette smokers than among non- smokers and increased with amount of smoking. A study in southeast England has found'~ an excess morbidity of smokers to prevail even at young ages. School children between the ages of 11 and 18 who smoked had poorer school attendance than non- smokers; and the excess number of school-days lost was attributable to the greater frequency and longer duration of illness among the children who smoked than among children who did not smoke. Coronary Heart Disease In most industrialized countries coronary heart disease is one of the main causes of death. In the United'~ States, for example, more persons die from coronary heart disease than from any other single cause. Several prospective epidemiological studies have established that the incidence rates of this dis- ease are higher among cigarette smokers than~ among non-smokers. Figures shown in this slide (t 10) are based on 12 years experience among men in the for- ward study undertaken in Framinghamy Massachu- setts. They illustrate the dramatic increase of coro- nary disease with advancing age. When comparing the experience of cigarette smokers with that of non-smokers, we find that the relative risk of coro- nary attack among c barette smokers compared to t

E. Cuyler Hammond 29 Coronary Artery Disease by Age and Smoking Framin ham Study I nci d. Rate per Excess Rate 100,000 Smokers/Morbility Age Non S. Smokers Non S. Ratio _ 35-44 1.4 4.1 2.7 2.9 45-54 4.6 11.1 6.5 2.4 55-64 16.2 25.4 9.5 1.6 Slide 10 From the Albany and Framingham Studies by Doyle, Dawber, Kannel, et al. non-smokers decreases as we go from the younger to the older age-groups. However, the absolute differ- ences im incidence rates between smokers and non- smokers increase with advancing age. In other words, the excess number of cases associated with smoking in ea& age group increases with age. The prospective epidemiological studies have also provided the opportunity to evaluate the effect of' smoking independently and in combination with other "risk factors" found to be associated with cor- onary heart disease. Based on the data from the stud- ies in Framingham, Massachusetts and Albany, N.Y., (slide # 11) it will be seen that the smoking of cigarettes increases the risk of developing coro- nary disease in both individuals with low and with high levels of serum cholesteroL A similar indepen- dent effect of smoking on coronary disease incidence can also be demonstrated in relation to a number of other "risk factors" (e.g. systolic blood pressure; pulmonary function, physicali actiyity, socio-environ- mental stress and personality type). In more com,- plex statistical analyses of the same data, it has been demonstrated that an independent effect of cigarette smoking prevails also when taking into account the simultaneous effect of a number of other "risk fac- tors:'

30 The Scientific Background Coronary Artery Disease Age-Adjusted Morbidity ratios by serum cholesterol Albany-Framingham Study Serum Choles. Non- Cigarette Level Smokers Smokers Low 1.0 1.8 High 2.0 4.5 Slide 11 From the Albany and Framingham Studies by Doyle, Dawber, Kannel, et al. A similar pattern of association between cigarette smoking and coronary heart' disease emerges from prospective studies on mortaIity. These studies have also clearly demonstrated that for both men and women the mortality ratios within each age and sex category generally, increase with the amount smoked daily. For example in age-group 45-54, coronary heart disease death rates among heavy smokers is three times the rate for non-smokers, this being true for both men and~ women. Some prospective studies have also provided the opportunity to assess the effect of discontinuance of smoking. The cessation of smoking is associated with a decrease in mortality zates. The decrease in risk of coronary heart disease mortality is substantial after a few years, and the rates after 10 years are equal to those of' persons who have never smoked. Most of the available data on the association of cigarette smoking and coronary heart disease morbid- ity and mortality have come from large scale epide- miological studies carried out in the United States, the United Kingdom and Canada. Data from more limited epid'emiological studies in other countries

E. Cuyler Hammond 31 such as F'm.land and Norway, suggest that the inde- pendent association of cigarette smoking to coronary heart disease is also found in populations other than in those from which most of' the hitherto available data have come. Due to the multifactorial etiology of coronary heart disease, it is, however, rather difficult to extra- polate directly from one country to another as to the impact of cigarette smoking on morbidity and mortality. Differences in the exposure of various pop- ulations to other "risk factors" (for example, dis- parities in dietary habits and constitutional factors) may determine the relative importance of cigarette smoking for the epidemiological pattern of coronary heart disease. Lung Cancer Up to 1912 medical workers were concerned only about cancer of the oral cavity in relation to smok- ing. For example, Abbe of New York in 1915 was relieved that it had taken him long to see a case of tongue cancer in a cigarette smoker. It was from these years that the first suggestions came that deaths from lung cancer were increasing in fre- quency among the male sex, and during the follow- ing decades it became clear that this was the case in most countries of Europe. At' this time tuberculosis was on the retreat, and~ it took up to about 1950 to prove finally that the increase was real and not due to better diagnostic methods. However, gradually the cigarettes came into suspicion as. a cause while other factors such as influenza epidemics and tarring of roads were excluded one after the other. During these years, there was no efficient treat- ment, so that statistics on deaths could serve as a measure for morbidity, but clinical workers in Aus- tria and~ Germany began inquiring among lung can- cer patients about their smoking habits, followed by a series of five more complete studies from the United' States and England appearing in 1950; and

32 The Scientific Background later also from Holland, Switzerland and other coun- tries. It appeared from all these studies, that' lung cancer patients had smoked' more than other people, and that heavy smokers stood'~ a higher risk of get- ting the disease. It was on this background of "retrospective" stud- ies, that it became possible to invest the very con- siderable effort necessary in the prospective studies, which follow a considerable number of men and' women, after having recorded their smoking habits. In this way it has been possible to evaluate the risk of various diseases following the smoking of tobacco -mainly cigarettes which dominate smoking habits in England and the Ut<ited States, Canada and Japan, where these "prospective" studies have been carried out. An overall international survey shows that lung cancer has become a major problem exactly in those industrialized countries where cigarettes are con- sumed in large quantities, and where good statistics are available, but when we attempt a quantitative es- timate of the problem~ we run into some fundamental difficulties. One difficulty is that, roughly taken, we must as- sume an average period of' about twenty years of smoking before lung cancer is diagnosed. Fortu- nately, however, it takes far less time to reduce the risk by giving up smoking; but this long period of in- duction means that we may find lung cancer limited to the male sex for a long time after the women have taken to smoking. Rates for women have been in- creasing in most countries, but they will be lagging behind the rates of men, until all age groups of' both sexes have been smoking the same amount of ciga- rettes for about twenty years. This is also the reason why most countries show far lower rates for lung cancer among men in rural areas, where cigarette smoking was adopted' later than in the towns. It is true that also non•smokers show slightly higher rates for lung cancer in the cit-

E. Cuyler Hammond 33 ies-as indeed most cancers do-but this very small difference is probably largely due to occupational risks in the towns. In spite of all these difficulties it has been possible to compare the situation for lung cancer deaths in-a number of countries at a time when women had not been smoking long enough to show higher mortality rates. The mortality rates for 1950 (slide #12) were plotted according to the overall cigarette consump- tion in the country concerned about 20 years earlier. There is a clear trend for countries of high consump- tion to show high rates. However, the situation has already changed. Rates have increased and are continuing to increase in all these countries, particularly in those where many men live in cities. Naturally, it also plays a very important part when smoking began in the var ious countries and how fast the use of cigarettes spread. This will decide the development during the coming years, because those birth groUps who were young enough to adopt heavy smoking from~ their earlier years w17l have the top level of risk all through their lifetime. Only when the youngest of the birth groups now alive have lived their lives through will we know the effects of the smoking habit of the young generation. The rate of increase in each country will therefore depend on the age at which people start smoking and on how much they smoke. (slide # 13) Chronic Respiratory Disease Chronic respiratory disease is a difficult field to ' cover because of the differences in definitions and ':. diagnostic terms in different countries. For this rea- ~' son we propose to avoid terms such as bronchitis ;' and emphysema and shall discuss the effects of ciga ~- rette smoking symptoms and abnormalities of lung i ! function which require no definition, Chronic cough~ and sputum, generally described as a smoker's cough, has been shown by sur-

34 CRUDE MALE DEATH RATE FOR LUNG CANCER IN 1950 AND PER CAPITA CONSUMPTION OF CIGARETTES IN 1930 IN VARIOUS COUNTRIES. soo Aoo iQp The Scientific Background - GREAT. BR ITAIN ~ F1 ~AND ~ SINTZERLAND Z+sQ73±~30 HOLLAND ~ U.S.A. uSTRALIA CANADA t ICELAND CUGARETTE CONSUMPTION Slide 12 Doll, R. Etiology of lung cancer. Advances Cancer Res. 3:1-50, 1955.

E. Cuyler Hammond 35 AGE-ADJUSTED MORTALITY RATES FOR CANCER OF THE LUNG AND BRONCHUS BY BIRTH COHORT AND AGE AT DEATH FOR MALES, UNITED STATES 1914, 1930-32 , 1939-41, 1949•50„ 1959-61. Pti ~o o 0 t 0 S ~ 3 -- -- - -r r ~ Y ! / / J I 1930-3T /1G8 Slide 13 Dom, H.F., and Cutler, S:J. Morbidity from cancer in the United States. Public Health •Monogram No. 56:1,207, 1959, an& un- published calculations of the Biometry Branch, National Cancer Institute, U.S. Public Health Service.

36 S The Scientific Background S 0 « 2 f Slide 14 Holland, W.W. The natural history of chronic bronchitis. I!. Co1= Iege of' General Practitioners 11:Supp.2:8-16, 1966.

E. Cuyler Hammond 37 veys in at least a score ofi countries to be directly re- lated to numbers of cigarettes smoked. Here (slide #14) is an example of a survey car- ried out by doctors among van-drivers in London and rural towns in England and in United States east coast towns. Less than one in ten of non*smokers admit to persistent cough and phlegm but about half of the heavier smokers admit to this regular cough- ing and spitting.. Many studies in whole communities have shown that this is a universal finding and that women who smoke the same amounts are affected similarly. Respiratory Tract Symptoms (%) reported by school children (11+ yrs) according to smoking habits . day Total Smokers 1 5+ No. 9907 16451 396 i Cough 4.0 6.9 16.1 Phlegm 4.6 9.1 22.6 Slide 15 Source: Holland, W.W. and Elliott, A., Unpublished data. Recent studies in England (slide # 15 ) have' demonstrated that cigarette smoking produces this effect in early life. Here you can see that in school children aged 11-18 who smoke five or more ciga- rettes daily, the prevalence of cough is five times as great as in children who do not smoke, and ap- proaches that found in adults in the same areas. The next feature of' chronic respiratory disease, and one which is disabling, is a liability to recurrent respiratory illnesses of all kinds. All the surveys re- ferred to in twenty countries have shown that smok- ers are much more frequently affected by such ill- nesses than non-smokers. Even nurses and college

38 The Scientific Background Mean Forced Expiratory Volume (1•0 Sec.). ® LONDON .~., COUNTRY TOWNS 91U:S.A. 3.8r 3.6 0 e N w . +N E- . M. =- w a b e d e Wnon-smokern, (b)ez-smokere, (c)smokers, I-14q. per day, (d)smoJters 15-24g. per day, (e) smokers, 25g. or mor• per day. Mean value and numbex of subjects- (a) 3-0(13); 3130), 3T(891; (b).2-8(36), 3•0(77), 3•7(89); (c) 2?(74), 3'0U421 ~6(60);, (d)2•7(98), 29U'34), 3•4q69h W2•5(291. 2'3(4(N, 3ACZ18). Slide 16 Idolland, W.W. The natural history of chronic bronchitis. 7.. College of General Practitioners 11':Supp.2:8-16, 1966.

E. Cuyler Hammond 39 students have been shown to lose significantly more time from their work if they smoke cigarettes th= if they do not. We have already pointed out how se- verely illnesses of this kind affect the general popula- tion of this country. In the United Kingdom, 27,000,000 working days are lost from bronchitic iilnesses each year and we know that a big propor- tion of these illnesses would not occur in a non- smoking population. The thir& main consequence of cigarette smoking is that it impairs the function of the lungs in a way which ultimately becomes irreversible. In nearly all the surveys, we have referred to simple tests of' breathing capacity that have been carried out. Here (slide #16) is an example from the van- drivers' survey. The important thing to note is the impairment of breathing capacity as cigarette con- sumption increases. The better performance of American men may be an effect of' relative freedom from air pollution and physique. . This test "reveals the beginning bf injury to the lung but most of these working men are unaware of it at this stage. In smokers, the deterioration contin- ues faster in the heavier than in the lighter smokers and before long, wheezing and shortness of breath begin to prevent all too many of them from under- taking anything but the slightest exertion, and it is this inability to breathe properly which finally kills them. This is shown by considering statistics from five countries in whi& death from chronic bronchitis and emphysema have been related to smoking habits. Here we have the American Cancer Society's study (slide #17 ) and you can see that' mortality from emphysema is 6.6 to 11.4 times as great in male cig- arette smokers as in male non-smokers and 4.9 to 7.5 greater in female smokers than in female non-smokers. This sorry story of sputum, illness, dyspnea and ~ death brought about by cigarette smoking is perhaps its most unpleasant and distressing effect:

40 The Scientific Background Age specific death rates per 100,000 persons by age, sex, and smoking status Emphysema Age Smoker Non- and of Cig. Smoke Exc. Rate Mort. Ratio Sex only Smoke/ Non.S. Males 45-64 24 4 20 6.6 65-79 153 13 140 11.4 Females 45-64 3 1 2 4.9 65-79 45 6 39 7.5 Slide 17 From the American Cancer Society's epidemiology study of over, one million men and women. Other Diseases Analyzing the excess death rate for all sites of cancer, ( siide $k 18 ) we can see that only half of it comes from lung cancer, and the other half from the rest of the localizations. Therefore, looking at lung cancer, we see only half of the picture. We want to emphasize then that the importance of lung cancer must not make us forget that, for other sites, ciga- rette smoking is a relevant associated factor. Most of the prospective studies conducted in various coun- tries show that such sites of cancer as buccal and pharynx, larynx, esophagus, bladder, and pancreas have high mortality ratios when comparing cigarette smokers versus non-smokers. It seems that for other vascular diseases a consist- ent association with cigarette smoking was noted in non-syphilitic aortic aneurysm, considering all' ages, (slide #19 ) but we have to point out that this is not one of the main causes of death. In cerebrovas- cular diseases, it seems that the possible tobacco in-

Cancer Death Rates*- Cigarette Smokers Vs. Non-Smokers Age 40-69 ~ Death Rates Site of Never Cigarette Mortality Cancer Smoked Smoker Ratio All Sites Lung Buccal, Pharynx Larynx Esophagus Bladder Kidney Prostate Pancreas Liver Stomach Colon, Rectum 137 275 II I I I 3 5 14 6 4 11 33 1.98 88 7.86 7 5.54 4 3.38 4 3.67 10 3.31 7 1.43 14 0.97 19 2.95 7 1.84 16 1.39 33 0.98 Age 70- 89 Death Rates Never Cigarette Mortality Smoked Smoker Ratio 701 1,222 1.75 14 257 19.07 9 21 2.33 0 15 ao 3 21 8.32 25 79 3.16 15 22 1.49 185 207 1.11 45 86 1.90 18 19 1.04 90 123 1.38 138 158 1.14

Death Rates*- Cigarette Smokers Vs. Non-Smokers Cause of Death Coronary Artery Other Heart Aortic Aneurysm Cerebral Vascular Other Circulatory Emphysema Pneumonia, Influ. Gastric Ulcer Duodenal Ulcer Cirrhosis of Liver Accidents, Suicide Age 40 - 69 Death Rates Age 70 - 89 Death Rates Never Cigarette Mortality Never Cigarette Mortality Smoked Smoker Ratio Smoked Smoker Ratio 345 627 1.82 2,337 2,910 1.25 56 78 1.39 478 659 1.38 8 24 2.96 26 110 4.25 67 86 1.28 838 784 0.94 14 21 1.51 232 262 1.13 3 26 8.83 12 189 15.52 7 14 1.97 145 224 1.54 1 7 5.38 12 36 2.97 3 6 1.97 37 55 1.47 10 18 1.70 18 18 0.98 58 70 1.20 108 175 t.62

E. Cliyler Hammond 43 fluence we see in younger groups (40-69), disap- pears in older groups. In practically all the studies carried out, a strong correlation between cigarette smoking and peptic ulcer was observed, both in mortality and in morbid- ity. The ratio of smokers versus non-smokers varies in different countries but it is higher than 2.0 in practically. all of them. There is a remaining group of diseases and conditions which need' further studiess to confirm the hazardous effects of tobacco con- sumptfon. Recommendations The evidence we have summarized shows how vast' and complex are the effects of cigarette smoking on health in different communities and countries. It' illustrates the difficulties of' extrapolating findings to countries for which dataa are not available. Our general position is that while there is no doubt about the universal hazards of cigarette smok- ing, these issues deserve further study and in many instances, more data must be collected. International collaboration in epidemiological studies of the effects of cigarette smoking should be encouraged and means secured to provide continuous long-term sup- port to such an effort. The following specific recommendations are made: 1) Countiies should periodically collect data on tobacco consumption by age and sex for representative samples of the general popula- tion. 2) Further data should be collected' on symptoms and diseases by standardized methods in rela- tion to tobacco consumption for representative samples of various populations. 3) A series of case-control studies of coronary heart disease and other diseases should be un- dertaken in several countries with widely

44 The Scientific Background differing risks to explore the range and varia- tion of the association with smoking history and other factors. 4) Countries with the requisite technical re- sources and where the possibilities of defining and monitoring populations are favorable and promise to yield important substantive find- ings, should undertake prospective studies. Such studies are already in process in Japan and some other countries. 5) More studies of the effects' of stopping smok- ing are required. In connection with anti- smoking campaigns, studies should always be made of their effects upon smoking habits and of trends of morbidity and mortality in groups of people who do and do not change their smoking habits. 6) Because of the overwhelming effect of ciga* rette smoking in the causation of lung cancer and' chronic respiratory diseases, studies of the relationships between other environmental factors such as air pollution, and these dis- eases should make allowance for smoking habits. This applies also to other diseases as- sociated with cigarette smoking. The next speaker was Dr. Jeremiah Stamler, Associate Professor of Medicine, Northwestern University Medical Schooi' and Director of the Heart Disease Control Pro- gram o f the Chicago Board o f Health. He was one of the first heart specialists in the U.S. to recognize the impor- tance of cigarette smoking in cardiovascular disease. Dr. Stamler's talk follows: It is appropriate to note first that in the Uhited States,, as in the other economicaliy developed coun- tries of the world, the cardiovascular disease prob- Iem is the number one health problem, both in terms

Jeremiah Staniler 45 of'mortality and morbidity: A few statistics`are suf& cient }o illustrate the scope of' the problem: The ` ` United},States in 1965 broke a-record. For the first time, we broke the one-million naark in total number ` of cardiovascular deaths. These deaths constituted kt ~X. deaths. ..; approaimately 55 per cent-a clear majority-of all What specific diseases exact this huge toll? Cor erosclerohcR,:~cardiovascular disease--severe atherci= ' But' far and away the most important offender is ath= ," Hypertensive disease also is significantly responsible." -: sible<for the rising death rate 'from cor` pulmonale: cigarette smoking is certainly a major factor respon-` "'chronic lung disease-is one important entity.. And puhnonale-pulmonary heart disease secondary ao sclerosis, involving coronary - and l cerebral arteries, _ = ing heart attacks, 'strokes,'` aneurysms, add serious ._ aorta, arteries of the lower extremities,, -and produc- thromboemboii'c. ;occlusive disease _ thro'ughout..;,the - Severe atherosclerosis of the corona arteries is Y' the liiggest_ single: problem. It was responsible in 1965.;for approximately 600,000. of the -more than .. white males, the U:S. death rate from coronary heart greater ,among -men prior to age 65 than among women. -The available data indicate that for middle-aged were persons under the age of 65, with the toll much Most important.of'all, about 175,000 of these deaths _. one million deaths from all cardiovascular diseases. disease rose steadily from about 1920 to 1950.ia-*----,` _ ...: .. . . ~ The curve then 1'evelled off, with little or no increase from 1950 to:-1960. -p'rom 1969 to 1965 there may : have even been a slight decline; it is too early to be certain about the trend during the current decade. parallels the increase in ciganette consumption from ` the rise in death rate from coronary heart disease -Is it an accident--0r no mere coincidence-that * References listed pp 72-73

46 The Scientific Background ~ 1920 on? Are the two phenomena causally related -as is undoubtedly the case for hmg cancer and chronic bronchopul'monary disease? The cohorts of middle-aged men hit by the epidemic of heart at- tacks in the 1940s, 1950s and 1960s are cohorts whose members in the majority have been habitual long-term cigarette smokers-for the first time in U.S. history. It' has been known for years that cigarette smok- ing has acute effects on the cardiovascular system, chiefly secondary to absorption of nicotine.'8• 19 Sig- nificantly more nicotine enters the human organism with cigarette than with pipe or cigar smoking.u+l' The effects of nicotine are both direct and indirect, the latter being a resultant of nicotine-induced cate- cholamine secretion. These acute effects include an increase in heart rate, blood pressure, cardiac output and also coro- nary blood flow (in people without severe coronary artery disease) even though the direct effect' of nic- otine is to decrease coronary blood fiow."•" Seram- free fatty acids rise.- _ In persons with coronary disease, there is an im- paired ability, or no ability, to increase coronary blood flow in the face of the increased energy de- mand upon the heart with cigarette smoking.13 In some coronary patients this is paralleled by a wors- ening of the electrocardiogram with cigarette smok- ing, and/or a dissolution of' the normal contour of the ballistocardiograph. It has also been shown re- cently that there is a significant rise in blood carbon- monoxide concentration, as carboxyhemoglobin, fol- lowing the smoking of a cigarette, accompanied by a shift in the oxygen-dissociation curve of oxyhemo- globin, leading to potential impairment of tissue (e.g. myocardial) oxygenation mechanisms.18 While reserves are large in the healthy individual, this can be a serious problem in the patient with severe dis- turbance of the coronary circulation due to coronary sclerosis, leading to aggravation of angina, abnor-

~ Jeremiah Stamler 47 mality of the ECG and baliistocardiogram, etc. As long as data on the cardiovascular effects of smoking were confined to acute manifestations, on blood pressure, pulse, etc., they commanded only limited attention. As Dr. Hammond indicated, it has been the amassing of voluminous and impressive epi- demiological evidence-indicating a relationship be- tween habitual cigarette smoking and coronary inci- - dence, morbidity and mortality- that has compelled serious attention in the last few years. It , is appropriate to review the epidemiological findings briefly, because they are fundamental to un- derstanding where we are now and where we have to go from here. Figure 1 summarizes eight year fol- low-up data from the U.S. Veterans study.10. 18 Cor- onary heart disease mortality rates are presented by number of cigarettes currently smoked per day for men at two age groups, 35-44 and 45-54. Compared to nonsmokers, cigarette smokers experienced signif- icantiy increased mortality rates. The rates were five to ten times higher in the smokers of one or more packs per day. Figure 2 presents data from the same study for the next two older age groups, 55-64 and 65- 74.110.18 Here the coronary mortality rates of heavy smokers were about double those of' nonsmokers. For all age groups, a classical staircase dosage effect is manifest. Note that the increased risk of death is much greater for younger than for older age groups (Figs. 1 and 2). Tlierefore age-adjusted mortality ratios tend to give an incomplete and even mislead- ing picture, with small ratios in the order of 1.7 or 1.8, whereas for younger men the ratios are indeed much higher. The age-specific data are particularly significant since millions are already addicted to smoking by their teens or twenties. The next figure (#3) from Dr. Hammond's data-~ 18-deals with CHB mortality ratios. Unfor; tunately, prior to this meeting I did not have access to the impressive data on life expectancy, and on

i i 0 ~ MORTALITY RATE ® NON - SMOKERS ~ n o~ ~. 'FJ- Q0- Q ® < 10 CIGARETTES 10-20 ~ ~ 40Q_ . Q 21-.39 tna ~ ~ 40+ ~~ ~M 300- tA NtA ~ Yae 200-~ I.ILIL.LLJ 0 AGE 35-44 RATES ARE PER 100,000 PER YEAR AGE 45=54 502 v

~ ® E2 m ~ NON -SMOKERS E 10 CIGARETTES 10-20 21-39 40t bi 4 1 1,000- 830 912 Nu~N ..~~ ° a~ AGE 5S -64 RATES ARE PER 100.000 PER YEAR AGE 65-74

9v9lqsGvzoz .~ ~ MORTALITY a RATIO 4.0- 3.0 - 2.4 m NQN-SMOKER6 < 10 CIGARETTES 8 10-19 20 - 39 . ® 40* 0 2( 2 I•9 . .' 1.5 . . . 0 1 -- ~. . . ~ . : ~ ~. AGE 45-54 AGE 55-64 EXpEGTED PEATHS WERE LESS TMAh! 10 . 3..4 AGE 65-74

4e9e8svzoz I ~ MORTAI{TY 1.1 a RATI O 3.0 - 0.9 u 2.0 m 0 0 m 0 © AGE 45-54 2.7 ~ NON= SMOKERS ~] E 10 CIfiARETTES p . 10-19 - 20-39 2.0 0 w PE# © AGE 55 -64 1.4 m 0 0 0 AGE 65 -74 M

52 The Scientific Background number of years lost as a result of cigarette smoking.. The data in Figure 3 again illustrate that the cost of cigarette smoking is much greater for younger than for older men. Figure 4 shows the same data for women.°- 1" Again, coronary heart disease mortality ratios for heavy cigarette smokers are significantly above one for all ages, markedly so for the younger women. The next set' of data deals not with mortality but with incidence (Fig. 5). These are data from Dr. Jo- 5 10 1'S - 20 YEARS AFTER AGE 40 Figure 5 Probability' of developing coronary heart disease in men after age 40, Albany Civil Servants Study (17). seph T. Doyle's prospective epidemiol'ogic study of Albany civil servants.e The graph gives probabilities for men of developing clinical coronary heart disease during the years from age 40 to 60, for cigarette smokers and non-smokers. All these men were on initial examination free of clinical coronary heart disease. Risk was a little more than 10 per cent for the non-smokers, whereas it was in the order of 35

Table 1- Smoking and Relative Risk of Developing Clinical Coronary Heart Dieeene - Middle-Aged American Men (15) GCgaF3sIVZoz Study .na }.Lr.nc. No. Ap at Onset Duration of pollov-up Years Rd.tlv. Rtak ef CHO for Specified Smoking Status Prolyha" and (11) Atu.sy* 70-62 6 79-SS D4• N.v.r Srnk.d Porarr Ctsar.tt. S.ck.r. -20 Ct6ar.tt.s per day 20 Cigarettes p.r day 20- clprntt.. p.r day 100 1fl7 17~ 1S5 274 N.a.re tl.otrlc ~19~ 40-SS 4.4 Nno-..ek.r oE I1 CtS.rac.. per d.y ll CIS.r.t4.s per day l00 177 Never Did Swok. S.uked Cigarettes Cigarettes = North (20) Dakota* 33 Cigarettes - not nav nw _ 100 107 221 Never Spmk.d or 10-19 20 Smoked Oaly Cigarettes ctS.rett.. P.epl.s C.s (1) 40-59 3 3 in Past par d.y p.r day . 100 . 710 . 725 . +Coroo.ry h..rt dl..e.. limited to .yourdlal CnfLrctioq. CncludinS CND de.th..

54 The Scientific Background per cent-three and a half times as high--4or those smoking more than a pack a day. These data speak for themselves, and several sim- ilar studies have yielded similar findings (Table 1) 16 The relationship noted between cigarette smokingg and mortality holds for morbidity. Figure 6 presents data from the Framingham Heart Attacks I 152 124 IOQ AII FnaW" Y 82 Yu ~ I 59 _ ~ 0 Ul >E ® a ~ N N 121 ua NEVER EX- CIGAR & ALL HEAVY SkHIKER SMOKER PIPE CIGARETTE CICARETTE SMOKERS SMOKERS SMOKING STATUS INITIAL EXAM Heart Attack •iD10 Dft Tkm AP: MI. CI„CHD Deatk Heavy CiEaette Smakc-0rer Pack/Day Figure 6 Smoking status at initial exRR+i++At+on and morbidity ratios for coronary heart disease (exclusive of angina pectoris), 12 year follow-up data, men originally age 30-62, Framingham Study (17). study, again on incidence of heart attacks, here de- fined as coronary heart disease other than angina pectoris (i.e. including myocardial infarction, acute coronary insu.fficieney and coronary heart disease death): ' Note the relatively higher rates for aIl cigarette smokers and for heavy cigarette smokers (a pack or . C.

Jeremiah Stamler 55 more a day). Note also the more favorable situation for the ex-smokers, and~ the very favorable situation for the cigar and pipe smokers. (It is as a result' of such data that in our Coronary Prevention Evalua- tion Program in Chicago, we try to get cigarette smokers either to give up tobacco altogether or to switch from cigarettes to cigar or pipe in moderation without inhaling. )13 The association between cigarette smoking and CHID risk persisted after rates were ad- justed to take account of the possibly confounding effect of other coronary risk factors (e.g. hypercho- lesterolemia, hypertension and overweight).°-e These data underscore the point that our problem in smoking apparently. does not go back to Sir Wal- ter Raleigh. The problem has not in the main been with us for three hundred years. Rather, it is prima- rily a problem of the 20th century, related to the shift in the 1920s to mass cigarette smoking, a by- product in turn: of World War I and the prior inven- tion of cigarette manufacturing machines in the 1890's, which made a cheap package of cigarettes possible. These facts need to be remembered and re- iterated, especially for those who say, "Human hab- its can't be changed." Smoking habits changed in one direction in the 20th century; they can be changed in the other directionl One other very important' aspect of the cigarette smoking and coronary incidence situation is illus- trated by Fig. 7, again from the Framingham study 1z A high proportion of those who smoke ciga- rettes experience their coronary disease in the form of sudden death. Overall, about 35 to 40 per cent' of all first myocardial infarctions in middle age termi nate fatally in the acute period. About half these fa- talities are sudden d'eaths-defined as death within 60 minutes of onset' of symptoms. Nonsmokers of cigarettes have a much lower risk-about one-fifth the risk-of dying suddenly from a first coronary episode compared with smokers of a pack or more a

56 The Scientific Background Sudden Death 217 NON-SMOKER AlL CI6ARETTE HEAVY OF CIGARETTES SMOKERS CIGARETTE SMOKERS (OVER PACKLDAY) Figure 7 Smoking status at initial examination and mortality ratios for sudden death, 12 year follow-up data, men originally age_30-62, Framingfiam Study (17). day. This is a very important aspect of the cigarette smoking problem. - The next figures deal with a key point mentioned by Dr. Hammond. Severe atherosclerotic disease- the underlying pathology in coronary heart disease -is multifactorial in etiology. The contemporary ep- idemic of premature (middle-aged) CHD in the U.S. and other economically developed countries, like all epidemics, has resulted from a confluence of multiple etiologic factors. It is a consequence of at least three major habit patterns of modern life: eat- ing habit and its effect on serum lipids, as well as on weight, blood pressure and risk of diabetes; seden- tary living habit; and cigarette smoking habit: The precise role of each of these etiologic factors

Jeremiah Stamler 57 needs to be pinpointed.ls Nonsmokers frequently suffer from premature CHD in the U.S. and other countries where the populations consume diets high in saturated fats and cholesterol, and consequently have high serum lipid levels. Furthermore, among populations consuming diets low in saturated fats and cholesterol and consequently having low serum lipid levels (e.g. the mass of the Japanese popula- tion), CHD in middle age is rare, despite heavy cig- arette smoking. These facts lead to the conclusion that cigarette smoking is not a primary, essential or sufficient cause of atherosclerotic disease. However, it is a most significant contributory cause in popula- tions like our own, wherein the nutritional-metabolic prerequisites for premature severe atherosclerosis are frequently present. This basic conclusion is doc- umented by Figures 8, 9 and 10. Figure 8 presents data from our group's prospec- tive epidemiologic study of male employees of the Peoples Gas Light and Coke Company in Chicago: 1,329 men age 40-59 on initial examination in 1958, free of evidence of frank CHD at that time, and fol- lowed since without systematic interventiona° These are eight-year mortality data, from coronary disease and from all causes, with the men stratified by serum cholesterol level and smoking status in 1958. Note the contrasts in mortality rates. The non-hypercho- lesterolemic nonsmokers had a CHD mortality rate less than one-sixth that of the hypercholesterolenuc smokers. (Non-hypercholesterolemia was here de- fined as less than 250 mg./dl. Actually, optimal levels are under 200-present too infrequently in this typical U.S. middle-aged male population to permit analysis of this type. ) Note also that the difference in mortali 'ty from all causes is almost threefold. Figure 9 presents similar data from the Framing- ham and Albany studies.e-'s Again note the marked difference between cigarette smokers with high serum cholesterol and nonsmokers of cigarettes with a "low choiesterol" (in quotation marks, because the

:~~se~s~zoz tiy 8 YEAR MORTALITY RATE PER 1,000 MEN ., 0 12O- o o =I ~wrv ~ .~no 100- 80- CORONARY h1ORTALITy .. old 4 MlVIR, PAST. fMOKIN STA7U5 ALL OCCASIONAL.UGHT f18 MORTALITY-ALL CAUSES 100 76 79 72 73 • 59 ~ t::::A r/1 ; 26 CIGARETTtS (10)/OAY), HEAVY CIGRR,PIPE SERUM <250. 250- 215? <250 250- 275> :_ ~ CHOLEST6ROL ALL - 274 274 ~ ~ NO.OF 40 9 4 2 16 4 9 ti v EVENTS p~ AfO:OF MEN 1 329 314 106 ' 71 515 138 151 p n AT RISK I o ALL ALL 109 1.329 NEVER, PAST. C16ARETTES (IO)/OAY), OCCASIONAl,L16NT HEAVY CIGAR, PIPE S 250 250- 275><350 250- 275> 274 274 15 9 5 43 i4 i9 314 106 71 515 138 151 U 00

MORBIDITY RATIO 5.0 - 4.0- 3.0 - i.0- © NON-SMOKER OF CIGARETTES, "LOW CHOLESTEROL" R7 NON-SMOKER OF CIGARET-T-ES, "HIGH CHOLESTEROL" QID CIGARETTE SMOKER, "LOW CHOLESTEROL" = CIGARETTE SMOKER, 'HIGH CHOLESTEROL' 0 ~t~9BS6#~ZOZ "LOw" IS gELQW MEDIAN, 'HIGM" IS ABOVE MEDIAN VALUE 2.0 1.9 OF SERUM CHOLESTEROL 4.5 pr

60 The Scientific Background median was used to divide the population, and its value was in the 230's, whereas a truly low choles- terol is one under 200). Note the effect of each risk factor when present alone, and the additive or com- pounded effect-the "insult added to injury"-when both were present. Thus there is a cumulative effect: These findings underscore the validity of Dr. Hammond's recom- mendation on the need for further study of this phe- nomenon of interplay among risk factors, particu- larly international research in those populations with truly low serum cholesterol levels because of dif- ferent habitual diet patterns, e.g. the Japanese. - Figures _10 and 11 illustrate another such inter- play-between~ blood pressure and cigarette smok- ing.. These are data from the Albany and Framing- ham, and from the San Francisco longshoremen studies.2, 8,18 Figure 12 illustrates the interplay be- tween overweight and cigarette smoking." Finally, Figures 13 , and 14 demonstrate the contributory effect of cigarette smoking to coronary risk when multiple abnormalities are presenta, lg These data are not only important from the stand- point of biomedicaI theory, i.e. for the clarification of concepts concerning the interrelationships among several habits and traits in the etiology of athero- sclerotic disease. They are also of great practical sig- nificance, for they clearly point the way to effective approaches for the control of this most common dis- ease related to cigarette smoking.1s Figure 15 deals with a different aspect of the ciga- rette smoking-coronary heart disease interrela- tionship-the matter of mechanism. Until recently, it was not known whether cigarette smokers had more of the underlying, disease, atherosclerosis, than non- smokers. The speculation was that cigarette smoking accelerated thrombogenesis rather than atherogene- sis, or that it interfered with the development of col- lateral vessels in the ischemic heart. These mecha- nisms may indeed be operative. But now data are

MORTALITY RATIO L...~i 'NON-SMOKER,* NORMOTENSIVE IM 'NON- SMOKER,` HYPERTENSIVE QIA HEAVY SMOKER, NORMOTENSIVE = HEAVY SMOKER, NYPERTENS)VE 9.6 5,8 I 5.9 9.4 AGE 45-54 A6E 55-6+ "NON-SMOKER": THOSE NOT SMOKING ANY CIGARETTES OR LESS THAN 20/DAY.

stgsesvzoz M 4RBIDITY TI A O R ® , CIGARETTE .X~ 4.0- ~ !R. 2.0- va~ 00 aiaB .xh1 ~ ~. t5~ @- .. NON-SMOKER OF CICaARETTESt E 130 mm Hg 10 NON-SMOKER OF CIGARETTEs, W CIGARETTE 130+ mmH~ SMOKER, < 130 mm Hg 5MOKER, 130 t mm Hg 1.8 I.0 3.8 N w I

8 YEAR MORTALITY RATE PER 1,000 MEN 100- CORONARY MORTALITY 80-. 60- 40 sVsgssVzoz EMOKIN 0 STATUa RELAnVi. WEIGHT NO OF EVfcNTS NQ OF MEN AT RISK ALL ALL NEVER, PAST. ClCCASIONAI, LIGNT <1.15 1.15> CIGARETTES (10)/OKY). HEAVY CK:AR,PiPE <I.IS LIS> 40 1 a IS 14 1,329 164 327 449 355 ALL ALL 109 1.329 0 0 62 ••. NEVER. PASr CIGARETTES (i0>/DAY), OCCASIONAL,LIGHT HEAVY CIGAR, PIPE 41.15 I IS> l/.l5 1.15> a 21 45 $1 164 327 449 355 o. W

am 634 ~ BHO M L ES = A R A 1 CHOL. > 2 50 SYS. B. P. J~I6O osseeslVzoZ 4W j SMIOKING > LPK6• CIGARETTES PER MY 60 ' NONE " ` ANY ONE ABNOWML 201 ANY TWO. Al.L THREE ApNOiUMAL ABN4RMAL

= ALt 0 /d NO RISK.FACTORS NIGH ONLY ONE HIGH.. SERUMCHOLESTEROL,. eLOOO- PRESSURE, WEIGHT CIGARETTE SMOKWGONLY HIG)f ANY TWO ONLY HIGH ANY THREE OR.ALCFOUR HIGH 33 22 10 = ALL QNO RlBK'. FACTORS HIGH B ONLY ONE HIGN.-SERUM CHOLESTEROL, BLOOO PRESSURE,WEtGHT - - ® [IGARiTTE SMOKING ONLY NIGH 56 120 oe ::. 106 3 11 20 40 35 1,329 96 230 260 490 253 Figure 14 Four risk factors (hypercholesterolemia, hypertension, overweight, cigarette smoking) an& eight year mortality from coronary heart disease (upper figure) and all causes (lower figure), Peoples Gas Light and Coke Company Study, 1958-1966, men age 40-59 in 1958 (15). The risk factor criteria (1958 status) were: hypercholes- terolemia-250 mg./dl. or greater, hypertension-diastolic pressure 90 mm. Hg. or greater; overweight ratio of' observed weight to desirable weight 1.15 or greater, cigarette smoking-10 or more cigarettes per day.

66 . The Scientific Background available from two important postmortem investiga- tions demonstrating that cigarette smokers have more atherosclerosis of the coronary arteries than nonsmokers I• 18 Figure 15 presents data from one of these autopsy studies, with careful quantitative grad- ing of lesions, to assess severity of atherosclerosis in relation to cigarette smoking habit antemortem. Lit- tle or no atherosclerosis was much more common among those who never smoked than among heavy cigarette smokers. Correspondingly, advanced ather- osclerosis was rare in nonsmokers, much more com- mon in smokers. The New Orleans study recorded similar fin.dings.18 Thus, whatever the mechanism, it is now clear that in our population-with its high prevalence rates for such other known risk factors for atherosclerosis as high-saturated-fat high-choles- terol diets,, hyperlipidemia, hypertension, etc.-ciga- rette smoking does indeed accelerate and intensify the basic underlying disease process, atherogenesis. There is "also evidence that cigarette smoking may not only aggravate atherogenesis per se, but also may lead to increased likelihood of thrombogenesis -through effects on platelet stickiness, on circulat- ing non-esterified fatty acids, catecholamine levels, and . on other mechanisms related to clotting mechanism.'8 Thus cigarette smoking may have a double effect-it may beat us with two sticks -an effect on atherogenesis and on thrombogenesis. And a third possibility remains: interference with devel- opment of collateral circulation. The next figures deal with the very important sub- ject of the effects on OHD rates of cessation of ciga- rette smoking. Figure 16 is from the US. Veterans study.lo• 18 It shows that those who had quit smoking by 1954 for reasons other than doctors' orders had lower coronary heart disease mortality rates over the next' eight years than men smoking cigarettes in 1954. The advantage for those who had quit by age 35-54 was marked; it was definite but smaller for men originally age 55-74.

W9 d PERGENT. WITH Fo DESIGN ATED DEGREE Oti ~ OF ATHEROSCLEROSIS i ~' 100%- ~ ~ ~ 9o%- ea~ ° E;S~ 80 %- 71 l $ ~ m 7n •/ -r_-- 42 35 17 I NEVER SMOKED REGULARLY < 20 CIGARETTES/DAY 20-39/DAY 40 t'/ DAY 55 MODERATE 28 25

68 The Scientific Background MOATALITT RATE 500- 400- 300- 200- Ed NoN-sMOKERe IILCGRRENT SAIOKER! = Elf-SMOKER! AGES Ex-fMOKERE STOOpEOfOR REASON6 OTNER Tu,W CpCtaR3' OROERa. R0Ef ARE OER.i00,0o0 PER r[AR- " 100- eg ~$ 22 AGES 95•44 MORTALITY RATE 2,000- © NON`'.AIOKERS QQ CURRENT SMOUERS~ E%-SMOXERS 1,5 00 - 1,000- 501 s7 ' 912 743 390 45-54 1,701 AGE 5S-64 AGE 65-74 EX-SMOKERSSTOVDEOFOR.REALONf 07wER.TwAn DOC?OR9 OROERS.. EAi" ARE PER 100.000 VEA YEAR Figure 16 Coronary heart disease mortality rates of nonsmokers, current smok- ers of 20-39 cigarettes per day and ex-smokers of 20-39 cigarettes per day, age 35-44, 45-54 (upper figure) and 55-84,, 65-74 (lower figure), IlT. S. Veterans Administration Study,1954,1962 (10).

Jeremiah Stamler 69 Figure 17 illustrates the additional fact, from Dr. Hammond's datay ft t,~t e longer a`go one has'"quit,: the greater the benefit.°.38 This`~s pnziIing to many ofus; p-ar_b'cu_1a_rfy the apparent finding that the ben- efit is relatively slight for the first five years for those who had previously smoked 20 or more cigarettes a day. It may be that this ou was inordinatel weighted with o 1e vin,g significant rosclo- r~~.~s.~ wnihe so~.wZio qui .~s y ~ ^of r-'ht at reason-and that it takes about five ears to _ "weed" these persons ou of e group, so that a\ n'"t "'e' cf'-'~`i "effe ~m~t~ng is apparent only for those plac dun&r prospechve epiii-emiologic observa on o m e year-or-later after Aiiev eave up ci a- rette_s: It does nbeem p°" ossib7e to clarify this matter wrt~presently available data, since medicaL exami- nations and risk factor measurements were not made in this population at onset of the study in 1960. This. is an important matter that needs more investigation. Be this as it may, there is, no doubt from this and the other studies that there is benefit to be gained by quitting cigarettes, even for heavy smokers, and even after long years of smoldng-not only in regard to avoiding.lung cancer and chronic bronchopulmonary disease, but also in terms of coronary prevention. = As to what is needed in the way of further re- search, I would summarize my view very briefly as follows: More work is needed on the mechanism whereby cigarette smoking increases risk of' coronary heart disease and severity of atherosclerosis (cf. di's- cussion of mechanism above). Additional epidemio- logical studies are needed, especially international ones, to compare effects of cigarette smoking in pop- ulations with different diets, serum lipid levels, etc. -in order to get more information on risk factor in- terplay. More animal experimental work is also needed on~ the foregoing questions. More studies are also needed on cessation of smoking and the matter of the time required for benefits to become manifest. Of key importance, definitive large-scale controlled

Jeremiah Stamler 71 field trials are needed' on the effects of quitting ciga- rettes. It is very difficult to understand why the equivalent of the Diet-Heart Study and the Coronary Drug Project are not being vigorously projected.ls.16 Frankly, however, as valuable as such additional research undertakings would be, I do not believe that they are the main need. The evidence is so ov- erwhelming concerning the health hazards of smok- ing that this Conference must record a primary public health and preventive medicine-rather than a research-need. The most urgent needs are in the area of how to enhance cessation of cigarette smok- ing, and how to prevent young people from begin- ning to smoke. In~ addition, to reiterate, there is the related need to face the tough problem of controlled mass field trials on the primary prevention of prema- ture CHD by cessation of cigarette smoking, The a priori defeatist and cynical notion that this crucial last stage of research cannot be done must be re- jected. Nor should' we be content to rely on our col- leagues abroad to do this job. Rather we should find ways to join efforts. It is particularly appropriate to stress this pro- posal at the present time, not only because it has not been made-for reasons I cannot understand-but also because the research group that did the Na- tional Diet Heart Study has just submitted its report on feasibility to the National Advisory Heart Coun- cil and has recommended that there be a mass field trial on the ability to achieve primary and secondary prevention of coronary disease by dietary means.13 There is a good possibility that such a mass field trial will be organized. It is of interest that in the feasibility study on diet and heart disease, even though intervention involvedd only diet, about 20 per cent of the volunteers from all over the country quit cigarette smoking while they were in the study. Clearly this poses the question: Should 'not consider- atioa be given, in planning mass field trials on prev-

72 The Scientific Background enting coronary disease, to studies not only on diet, but also on cigarette smoking? It seems to me that this deserves intense attention at the present time. . s s REFERENCES 1. AUERBACH, 0.; HAMMOND, E.C., and GARFINKEL, L: Smoking in relation to atherosclerosis of the coronary arteries. New Eng. J. Med. 273:775-779, 1965. 2. BORHANl, N.O.; HECHTER, H.H., and BRESLOW, L: Report of a ten-year follow-up study of the San Francisco longshore- men. Mortality from coronary heart disease and from all' causes. J. Chronic Dis. 16:1251-1266, 1963. 3. COMROE, J:H:, JR.: The physiological effects of smoking. Phys- ioL Physicians 2:1-6,1964: 4. Coronary drug project enters enrollment phase. (Medical News). J.Ai.M.A. 200:37-38, June 19, 1967. 5. DAWBER, T.R.; KANNEL, W.B., and McNAMARA, P.M.: The prediction of coronary heart disease. Trans. Ass. Life Insur. Med. Diu., Amer. 47:70-105, 1964. 6. DOYLE, J.T.: Etiology of coronary disease: risk factors influenc- ing coronary disease. Mod. Cone. Cardiov. Dis. 35::81-86, 1966. 7. DOYLE, J.T.; DAWBER, T:R.; KANNEL, W.B ; HESLIN; A.S, and KAHN, HA.: Cigarette smoking and coronary heari dis- ease. Combined experience of the Albany and Ftamingham stud- ies. New Eng. J. Med. 266:796-801, 1962. 8. DOYLE, J.T.; DAWBER, Z'.R.; KANNEI„ W.B.; KINCH, SJ., and KAHN, H.A.: The relationship of cigarette smoking to cor- onary heart disease. The second' report of the combined' experi- ence of the Albany, NY and Framingham, Mass., studies. J.A.MA. 190:886-890, 1964. 9. HAMMOND, E.C.: Smoking in relation to the death rates of one million men and women. Nat. Cancer Inst': Monogr. 19:127-204, 1966. 10. KAHN, H.A.: The Dorn study of smoking and mortality among U.S. veterans: report on eight and one-half years of observation. Nat. Cancer Inst. Monogr. 19:1-125, 1966. 11: KERSHBAUM, A4 BELLET, S:; HIRABAYASHI, M.; FEIN- BERG, LJ., and EILBERG, IL: Effect of cigarette, cigar and pipe smoking on nicotine excretion. The influence of iuhaling, Arch. Inttern. Med. (Chicago) 120:311-314, 1967. 12. KERSHBAUM, A.; BELLET, S.; JIbIINEZ, JF and FEINBERG, LJ.: Differences in effects of cigar and cigarette smoking on free fatty acid mobilization and catecholamine excretion. J.A.M.A.195:1095-1098,1966. 13. NATIONAL DIET HEART STUDY RESEARCH GROUP: The National Diet-Heart' Study. Final Report. Amer. Heart Ass.

Jeremiah Stamler 73 Monogr. No. 18. New York, American Heart Association, 1968. 428 p. Also in: Circulation 37(SuppL 1):1-428, Mar. 1968. 14. PAUL, O.; LEPPER, M.H.;, PHELAN,, W.H.; DUPERTUIS, G.W.; MacMILLAN, A. McKEAN, H., and'PARK, H.: A longi- tudinal study of coronary heart disease. Circulation 28:20-31i 1963. 15, STAMLER, J.: Lectures on Preventive Cardiology. New York, Gruae & Stratton, 1967. 434 p. 16. STRONG, J.P.; McGILL, H.C.; RICHARDS, M.L., and EGGEN, D.A.: Relationship between cigarette smoking habits and coro- nary atherosclerosis in autopsied males. Circulation 34(SuppL 3):31, Oct. 1966. 17. U.S. NATIONAL HEART INSTITUTE: The Framingham Heart Study. Habits and' Coronary Heart Disease. Public Health Ser- vice Pub. No. 1515. Washington, D.C., U1S. Govt. Print. Off., 1966. 13 p. 18. U.S. PUBLIC HEALTH SERVICE: The Health Consequences of Smoking. A Public Health Service Review: 1967. Public Health Service Pub. No. 1696. Washington, D.C., U.S. Govt. Priuti Off., 1967. 199 p. _ 19. U.S. SURGEON GENERAL'S ADVISORY COMMITTEE ON SMOKING AND HEALTH: Smoking and Health. Report of the Committee. Public Health Service Pub. No. 1103. Washing- ton, D.C., U.S. Govt. Print: Off., 1964., 387 p. 20. ZUKEL, WJ.; LEWIS, R.H.; ENTERLINE, P.E.;, PAIIVTER, R.C.; RALSTON, L.S.; FAWCETT, R.M.; MEREDITH, A.P, and PETERSON, B.: A short-term community study of the epi- demiology of coronary heart disease. A preliminary report on . the North Dakota Study. Amer. J. Public Health 49:16304639, 1959. I 1 ACKIVOWLEDGMENTS: It is a pleasure to acknowledge the cooperation and support of Eric Oldberg, M.D., President, Chicago Board of Health and Chairman, Chicago Health Research Foundation, and Samuel L. Andelman, M.D., M.P.H., Commissioner of Health and Secretary, Chicago Health Re- search Foundation. It is also gratifying to pay tribute to my senior col- leagues cooperating in the research briefly presented in this paper, David M. Berkson, M.D., Morton B. Epstein, Ph.D. and Howard A. Lindberg, M.D. We are also grateful to Paul Meier, Ph.D., of'the De- partment of Statistics and the Biological Sciences Computation Center, University of Chicago. It is a further pleasure to express appreciation to the Peoples Gas Light and Coke Company, its Chairman, Remick McDowell and its Presidentj Leslie A. Brandt. The research of our group presented in this paper was made possible by grants from the Chicago Heart Association and the National Heart Institute, National Institutes of Health, United States Public Health Service (HE 04197 and HE 09426). Finally, it is a pleasure to acknowledge permission of colleagues and publishers to reproduce their data-Oscar Auerbach, Nemat O. Borhani, Thomas R. Dawber, Joseph T. Doyle, E. Cuyler Hammond', Harold A. Kahn, William B. Kannel and the American Heart Association, the Journal, of the American Medieali Association, , ~.

74 The Scientific Baclc$round the Journal of Chronic Diseases, the Life Insurance Medical Directors of America, the New England Journal of' Itiledicine and the United States Public Health Service. The relationship of cigarette smoking to various types of cancer was discussed by Dr. George E. Moore, Direc- tor of the Roswell Park Memorial Institute in Buffalo, N.Y., and Director of Public Heath Research, New York State Health Department. His talk follows: I will refrain from repeating what has already been said this morning by Senator Kennedy and' the other speakers. However, one thing that the Senator said was clearly in error. I can personally testify that the industry has not been totally inattentive to the cigarette-health problem. Indee& during hearings be- fore the subcommittee, Senator Hartke reminded me of remarks I had made to a quasi-private luncheon group; he obviously hoped that they would embar- rass me. Apparently; the industry monitors practi- cally everything we say and do. Since this will prob- ably be my last major address on this topic, I will miss their attentions. One would only hope that they would develop a meaningful scientific program com- mensurate with the seriousness of the problem. Now, as far as cancer is concerned, I think no thoughtful person who has been informed of the facts would deny the fact that cigarette smoking is a causal factor in lung cancer. The problem from a re- search standpoint has been to evaluate supporting evidence, develop biological assays which are rele- vant to the human problem, and search for carcino- genic and cancer promoting agents in tobacco in the hope of preventing this pandemic of diseases. Luns cancer is not as susceptible to research methods as several other kinds of cancer that are caused by smoking or other uses of tobacco. Not enough clinical research work has been done on oral carcinoma. Here we have a site which can be ob- served and biopsied in a serial fashion. We can look

George E. Moore 75 for agents that may stop the progress of oral cancer and perhaps even reverse it. It should be possible to devise clinical studies in which the victim-sorry, the tobacco chewer-is persuaded to refrain from chewing for several weeks or months so as to ob- serve the reversibility of his precancerous lesions. Carcinomas of the pharynx and larynx can give us a very clear idea of the pathogenesis of cancer. Pipe smokers, cigar smokers and chewers who swallow tar or tobacco extracts are most apt' to develop can- cer of the external larynx; while cigarette smokers who mainline the smoke through the internal or in- trinsic larynx are susceptible to lesions of the vocal cord. Little is known about carcinoma of the bladder in cigarette smokers. However, a heavy smoker doubles his risk of contracting bladder carcinoma. With this minute review of a large problem-a .problem which has kept our attention now for 17 years-I wish to indicate the direction which we will now take. What we have done or rather failed to do in the past, because of' a lack of support by physi- cians and scientists alike, is cause for considerable remorse, since so many lives could have been saved! It is gratifying to note that the government will start a task force for lung cancer which we have been requesting for the last fifteen years. May I cau- tion this group about the problems that they may en- counter which could impede or suppress this goal- oriented program. First, no one wishes to tackle the problem directly; the money allocated to the project will be spent on such things as molecular biology, extracts of bee's knees, and anything else that the imaginative scientist can convince himself is related to lung cancer. Many scientists believe that any proj - ect which may provide "practical" information must be bad. I say this with some feeling. Some years ago, a large professional medical organization in this coun- try received funds from the tobacco industry to 1 0±

76 The Scientific Background support' research relating to lung cancer and other smoking problems. We submitted about' 15 research projects for funding consideration. Our Cigarette Cancer Committees were reasonably certain that for public relations purposes, at least one of' our propos- als would be accepted. We, in turn, predicted that it probably would be the proposal least directly asso- ciated' with the problem; and so it was-the least re- lated project was chosen! - ' We would hope that the task force and other study groups seriously consider practical' health measures. It may be a possibility to design a less harmful cigarette which would be salable. _VVe agree that cigarettes should be labeled as "hazardous to health", and that the amount of tar and nicotine should be printed on the packages. However, we would like to see a reverse tar derby started. We would like to have the maximum~ per- missible yield of tar per cigarette set by law at 15 milligrams. This would be a reasonable public health precaution until such time when a safer cigarette could be found. There should be a minimum quality standard for all the components of cigarettes such as additives, residues and~ flavorings. These standards should be similar to those required for all food products cur- rently regulated by. the Food and Drug Administra- tion. We have studied the use of an aluminum band over-wrap on~ a cigarette. This over-wrap automati cally extinguishes the cigarette when it' has burned down to a butt length of 30 mm. We tested it and the outcome was quite reassuring. Cigarettes should be modified so that smokers would find them less desirable for inhaling purposes. ~This could~be accomplished by altering the acidity of ~ the smoke; perhaps this could be done with fiIter .,holders of a selective design. I hope that some of our research will continue to involve both the adverse effects of solid particles as

George E. Moore 77 well as gaseous materials. The gaseous materials may not necessarily be carcinogenic, or even tumor producing, nevertheless, they undoubtedly injure the bronchial epithelium and destroy the protection afforded ciliary movement. This makes the respira- tory apparatus more susceptible to mechanical and chemical damage. As was mentioned by the previous speaker, large cooperative clinical trials should be established for evaluating the changes in cigarette designs and smoking habits. Dr. Bross, a brilliant colleague on the staff at Roswell Park, feels that in, three to four years, with proper techniques, we could actually assay and de- tect whether or not there is a significant clinical re- sponse to-for example-lowering the average tar level of a cigarette from 20 mg. to 15 mg.. If Dr. Bross says that this can be done, I am willing to be- lieve that' it can. These less interesting, difficult,, prognostic studies should be started while we are doing our esoteric, or so-called basic research projects. I would also support Senator Kennedy in his re- quest for cigarette taxes which would be propor- tional to the tar content of a cigarette. I did not reat-e ize that there was a model for this. Apparently, al- coholic products are taxed according to their alcohol content--or, so I was informed by Mr. Ubell, one of my science advisors; and if this is the case, perhaps this can be made acceptable for cigarettes. I doubt whether Congress would pass such legislation; how ever, an attempt should certainly be made. In summary, then, I think that many practical things can be done to protect the consumer and, at the same time, to further the work of the researcher, our work, in the fundamental pathogenesis of the development of lesions. In view of the information presented at' this symposium,' I recommend that the already mentioned suggestions plus the following steps be taken to aid in the finding of a less danger-

78 The Scientific Background j ous cigarette. Cigarette advertisements appealing to chiidren should be entirely eIiminated;,and formal' education of children concerning the health conse- quences of smoking should be encouraged and sup- ported. Why should we do all this? The reason is that even though people are fully informed of the dan- gers of smoking, they cannot stop. It is, therefore, up to us to find ways for limiting the hazards of cig- arettes; but more significantly, we must attempt to guide our children until they reach the age whereby they would decide for themselves and choose which personal mismanagement disease they wish to in- dulge in. * . . The topic of cigarettes and respiratory disease was dis- cussed by Dr. C. M. Fletcher, Reader in Clinical Epide- miology, Royal Postgr uate e icaT-School, ondon, and Secretary of the Committee on Smoking and~th of the Royal College of Physicians. His paper follows: I count it an honour to have been invited to speak - at this opening session of the first World Congress on Smoking and Health, but the honour carries a heavy responsibility for it is not easy to summarize in a short space of time the vast amount of evidence which shows that cigare ' oking_plays an im~por- tant part in the causation of disabIing and fatal res- -- --------- -.1._ pirato~ry d~e. Perhaps it is appropriate that some- = one from my country should be called upon to tackle this task since we have for many years had the misfortune of having a higher mortality from chronic respiratory disease than any other country in the world. This confronts us with the fact' that however im- Qortant cigarette smoking may be in the causation of xesgiraio-disease there are other im ,portant a lu- vant' factors. Tfiis is c7ear y seen y constdertng the- L=•---F'='~

C. M. Fletcher 79 ,.,mYort~ality..rates from chronic bronchitis and pneumo- nia in the _three conntnes wTucli` over the'gast thirty rs have. had the highest cigaretfe co`ris~tion among all the countries for which statistics are avail=`„ afiIO T'lie -United States, F°irilan d the Uni'ted Kirigdom. ~ ""As"'shown in Table 1, there has been an increase in bronchitis mortality over this period in all three countries, nearly but not quite counter-balanced by a reduction in pneumonia mortality. But the most striking fact is that the combined mortility`fromC these diseases in England and Wales in both these periods has been more than five times greater than ~ in the U.S.A. or Finland, and bronchitis mortality ~ has been between thirty and fifty times greater. Within each of these . three countries epidemiological studies have shown that the prevalence of chronic respiratory, disease is directly related to cigarette con- sumption.2*' 29. Sq 6; Ibe conclusion, to which I shall . return, must be that care smo is more da -. gerous m's2~ni~ ef[t~ironments an m o e._ ut first' -MMconsi er the evidence of the effects of cigarette smoking in detail. Many studies have shown that mortality from chronic respiratory disease and in particul'ar from bronchitis and emphysema is related to smoking babit's.1'• 1'. b9• Cigarette smokers have been shown to have a greatly increased mortality compared with non-smokers, the mortality increasing steadily and quantitatively with the numbers of cigarettes smoked.590 Cigarette smokers who inhale deeply have a higher mortality than those who do not?' Mortality is reduced im those who have given up amoking,14• 23 and in pipe and cigar smokers it is only slightly greater than in non-smokers.l', 6°` Thus mortality from chronic bronchitis and em-. ~ phy_sema is related'ito smoking habits am manner ~I ver~ ~similar to that o6seiwed`for lun cancer. But ~ eret~ is an imporfant epdemTol-ogical erence. In all countries for which we have records, lung cancer

80 ?ABLB 1. The Scientific Background DATB R6Y8 PER 100,000 ,pRO.r DBQVCHITIS AND Pt7EUW91A IN M ACLD 43-64sIN.Q1S.A.. 7ffiIAUD AND DICGND AND :tALBS Countty Yaar BrcneAttlY Pnaueonla Drouc6ltia 6 Pn.wonia 1952 3 46. 47 U.S.A. 1962 21 17 33 1952 2 33 35 FIN{AND 1962 S 32 37 BNCUND . 1952 125 45 ~ 170 ' 6 WALES 1962 145 40 183 ~; cons''anf -" mortahty ~rom broachitisas~ strikmg; and ~i' ltb fact as een used tii-cfiallenge the concTusion ._..---- from_ othe er`vidence-that cigarette smokiag is a cause__of bronchitis mostality _The explaaation ap- . pears to be tba~`unlike Iung cancer,'`chronic bronfi chitis responds_toIreatment. This is shown by the fe- ,, ~ male mortality trend.-Few Bntisfi women at the age at which death fromironehitis is " common,.._Eave emphysema and bronchitis has been incieasing rap- idly'-in"the United States,ls_ 6°' but it has been sug- gested tbat this may be due in part to physicians be- coming more widely aware of the importance of this disease as a cause of death in recent years. In England' and Wales the contrast between the ~ rapidly rising mortality from` nili` g"~ca_ncer 'and- the world pandemic of cigarette smoking,:7, 48 but in many~countries bronchitis mortality is increasing slly_or-.notat0 alL. It is true that mortality from • 1Wn acrtality ratss for quinqu.nnta . 45-49,30-34, 35-59 snd 60-64 _ mortality has been rising steeply in the wake of the

C. M. Fletcher 81 been life-long cigarette smokers, and their mortality_` imon~~re ~s a: sZight increase in is d€c~g. ~ wr consider the male/female mortality ratio in the past four decades we find a trend very similar to that for lung cancer. It is reasonable to conclude that, in women, modem treatment is reducing mor- tality while in men an increasing incidence and se- verity of disease, attributable to cigarette smoking, is being nearly balanced by a reduction due to treatment! le This balance leads to an interesting economic consideration. In Great Britain many men~ with chronic bronchitis have spent between £2,000 and £4,000 ($642,000) on the life-long cigarette smoking which has been a major cause of their disa- bility. In thousands of these men life is prolonged only by an expenditure of up to £ 100' per annum ($300) for antibiotics, bronchodilators, diuretics and hospital admissions. The disease is expensive to acquire and the economic burden of delaying its final fatal outcome is heavy. This macabre calcula- tion depends on the validity of the evidence that them men have really been disabled by the cigarettes they have smoked and we must return to consider this evidence in greater detail. There are three comp~onents.in-the syndrome_ of bronchitis an empliysema, all of whicli Rave been shown to be relatedjo,cigare ,tteysmok_in.g. F~irs_t.; there s the smokers coug~, dry at first but in many cases productive of increasing quantities of sputum. Surveys in maIIycountries s•, 3, 13, , zl„24, 27-80; 43- s, as, sa, ar; so have shown that the proportion of ciga- rette smokers who have a chronic cough increases with increasing cigarette consumption and falls nearly ~ to non-smoking levels in those who have stopped. In ~ most instances the cough clears within a few weeks or months of stopping smoking and there can be no doubt that it is due to and not' just . associated with smoking. Pathologicai studies °-7,*1•b" have c horytn_. •- --

82 The Scientific Background hypertrophy of_ bronchial mucous glands in smokers tivhich is absent in those who have not smoked, Second~ there are recurrent illnesses in which_ the __ spuum becomes _more pio uF`s `e an3 often purulent. In these ~lnesses the patfent may-have to rest in bed or stay away from work and if he has obstructive bron- chitis or emphysema his difficulty in breathing may increase to dangerous levels. That cigarette smokers are more prone to respiratory illnesses has been shown repeatedly.'s•S4 In adult surveys the liability to recurrent' chest illnesses is related to sputum voI- __-- _._. _ , -. .. ._- ume. -.•- . ~ In,a prospective_study which my colleagues and' I have carried out' over the past five years in men aged 35-60 in London,20 we con ff&ed this buf r----------•-- , ...,.. found no relationshxp between frequency of illnesses d_ci~ette-smoking m men witli smular sputum,~ voTume_(Figure 1). Thus; althougb in the labora- MEAN PERCENTAGE OF MEN REPORTING CHEST EPISODES' AT SIX MONTH INTERVALS FOR $VEARS First Hour 40 k N I ._..~ I  ISputum Yolume 30 PER CENT 20 10 NON-SMOKERS < 15/dar 15or more/daX CIGARETTE SMOKERS Figure 1

C. M. Fletcher 83 tory, cigarette smoke has been shown to inhibit"y ciliary action,s'a to impair the activity of macro- phagesz' and to slow down the clearance of bacteria from the lung,'• it ma be h rsecretion of bron- chial mucus tha is the mostim-'orast nvorm eaakmg own the defense of the lung agaan's'mf'ec-_ " f`iori in man. '"This sion is supported by the observation that' even between acute episodes of illness, the bronchial tree of subjects with productive cough is no longer sterile as it' is in normal people.8• °- a5, 86 The excess secretion caused_by- smoking.seems_to __ ioRg;r thP no=al sazaterilizing,pQwer~_the.lung and thus to ope~~~the door to,recurrent_in_fection It has, however, ~shown that even young people who smoke cigarettes-nurses in training and col- lege students-most of whom have presumably not yet developed persistent' expectoration, are more prone than~ non-smokers to recurrent illnesses4°, *° The thir& and most im __..rtant_aspecY of chronic bronclu~tis and e.mphysema is the velopment.o ' creasm& airws narrowin~,which causes obstruction tothe fr'ee flow of air in and especially out of the ' lungs. This gradually results in disabling breathless- ness and finally fatal respiratory failure. In some cases this narrowing xisdue to emp~sema wc d~- AM s the normal eiasticity of_tbe_,,,lptng. The small bron`"7uc ai"~tu~es are no longer held open by t~e re- tra.`lMe~orce'orthe lung and collapse during e~ua-_ tion. _ ` ` "'W'e now know that this rsistent auw_a~''snarrow- iU can a1o develo wi out em sfi ema w"en -t~`e r ~ smaller respi atory passages are naowec in a - ner we do no e y a con wec omc o s ron tis. -veraL aufopsy s ies ave s own at emphysema is more prevalent and severe in the lungs of smokers than in non-smokers 1, 7,50 ' Epidemiological surveys have repeatedly demon- strated that in the general population the average N O N ~ CD Ot Cl ~ ~

84 The Scientific Background level of breathing capacity is significantly reduced in smokers.2• g• z7• Z9,30, 44, s,, as Severe impairment of' res- piratory function from emphysema or obstructive bronchitis in the general population and in the hos - ,._ pital-clinic is almost confined to cigarette smolc- , el$.13,8T,42 ` TYfa y meticulous studies, using a wide variety of pulmonary function tests, have demonstrated that not only is the ventilatory function sz, 61, 62 of the ktngs impaired by airways obstruction in cigarette smokers but gas-transfer capacity, the ability to get oxygen from the air into the arterial blood, is also re- duced even in young and apparently undisabled smokers:81• $1• °a In all these studies those who have stopped smoking cigarettes have been shown to have pulmonary function that is almost' as good as that of non-smokers, which suggests that' the impairment caused by smoking is at first, in most cases, revers- ible. In one study a three-week period of abstinence in young smokers resulted in a significant improve- ment in both ventilatory function, gas transfer and cardiovascular performance on exercise.3z It' is, however, the unhappy experience of clinicians that ia most cases of established chronic obstructive bronchitis or emphysema the damage done to the bronchioles and alveoli is irreversible and the im- pairment of pulmonary function is often unaltered. This is not surprising since ~anatomical changes _ i of' emphysema and bronchitis are found to persist in__ t ex-smokers.1T Stopping smoking may often lessen ~ the severity of cough and the patients may feell better.lE• 23 Indeed there are occasional patients who benefit greatly with "a disappearance of cough and recovery of lung function when they stop. These are mostly younger patients. In the prospective study to which I have referred, wejpgnd,pg,ci ificant difference inAe average rate ; of decline of' ventilatory cagacity in men~ aged 30-59 ' over a riod of ve as-be e"i en a-oup-of"43~ , smakersn who stopped smoking an ' 383 me`n vv_~ I

C. M. Fletcher 85 continued to smoke the same number of cigarettes___-- le RATE OF DECLINE OF FORCED EXPIRATORY VOLUME (FEV L.O) OVER 5 YEARS IN MEN AGED 30-59 Smoking Habits Number of Men Decline of FEV- (ml. Year) Non-smokers 110 8.5 (4.3) Ex-smokers 112 15.5 (4.2) Smokers who 45 ' 29.0 (8.9) Stopped Smokers of 1-14 213 24.6" (3.9) Cigarettes daily Smokers of 1S+ 170 36.8 (3.8) Cigarettes daily S.D. in brackets Table 2 o smoliabil'lty_ tA_resgira w$Fi some common conshtuho a dise may mcrease the eslre smo ana ~roneness to c es ease. e at e preva- lence o aa ' mo ty om chronic respiratory dis- on ls causal or due to an m e`ndent~" assocla , ~, ,---~-,..--- ing, it ls necess fo`es~a~ilis~`whet~er the associa- but, as in the case of other diseases related jQ_Srnqlc chronic respiratory disease with cigarette smoking, The evidence that I have summarized shows a striking association of prevalence and severity of inhale, strongly suggest that the association is due to ease increase with increasing cigarette consumption, are low in ex-smokers and higher in smokers who

86 The Scientific Background cause and.e.ffect. But there is additional convincing evidence that the association is causal. -First, iChas been shown that in identical twins with different smoking habits respiratory, symptoms t are more frequent and respiratory function is im- (~ paired' in the heavier smoking member of each ` pair.l3. ms ) Secondly, when smokers are matched in pairs with non-smokers in relation to a large number of unconnected personal factors which should control most constitutional differences, the smokers retain a fifteen-fold excess mortality from emphysema.24 Thirdly, in animal experiments, cigarette' smoke has been shown to induce severe bronchitis E3 and to cause destruction of the lung by emphysema.'.2e I have already emphasized that' smoking is not' the only environmental cause of respiratory disease. The _ y_country-nay,. _. _. vast excess of chronic bronchitis in in be largeby _due',to. air pollution but the excess is found even in our most rural areas. Miners and other workers exposed to chronic dust inhalation have been shown to have an excess of respiratory disease and impaired lung function compared with non-miners which indicates that dust exposure can " cause respiratory disease. , ~ But it seems that the adverse effects of air pollu- ti'on a`nd dust - ~xposure _ chiefly affect those who _._ smokg-cigarettes, In non-smokers the difference be- tween the prevalence of respiratory disease in the United Kingdom and the U S.A. is small 1° and some studies have shown that differences in prevalence of bronchitis between workers exposed and not ex- posed to inhalation of dust are confined to cigarette smokers.21• 68 But it is with the effects of cigarette smoking that we are concerned here today and there is no doubt ,,_.....__ --.-- at even in the cleanest air and the cleanest ocaupa- tion cigarefte sinoking causes peisistenr cougfi`and exnectora no ~ferferes'with--employmeat by ea-~=°' ses and ultimately disables couraging recurrent illnes'se s

C. M. Fletcher 87 and kills an important minority of those who adopt . the h'abit. --- At present we know of no means of preventing ,f the onset and of delaying the progression of severe ~ pulmonary disability in persistent cigarette smokers. Control of infection by antibiotics in early bronchitis ~ appears to be ineffective,40 and we have found in our prospective study that the rate of decline of ventila- ' xory capacity in cigarette smo~fastei m men ` with simple ~bronchitis ai with iecurrenf cb~~i - nesa'nin t~iose without=eith~i fact~os -~so that' treatmen ec to -ese adments is uMeIy~o J The only_luea_ns of~prevention wluch we _know to - be effective is to sto suscephble sub'Lct~from smolQng before'4ixeg°arab~e damage has been done. Although we have no proven means _by-which we can detect' _sus`"' c g~1es, our own iesults hint that the may be found among th_s_e__snio_~ ewho at an e~y~age ave ea~y some significant evidence_ of ainvays narrowmg.. We have found that th_erate.,_.- of decline of ventilatory capacity in the course of five years, is partiaularly accelerated in such men _ (F.lgure 2), ._... .. _._ ., .. Perhaps we shall one day produce cigarettes. that will be as harmless as pipes and cigars appear to be today but the only safe road for men and women to follow at present is one of complete abstention from cigarette smoking. If at this conference we can advance our under- standing of how to help young people not to start smoking cigarettes and their elders how to stop, we shall have found a way to prevent a lot of most un- pleasant hawking and spitting; we shall have count- ered much misery and economic loss due to recur- rent illnesses, and we shall have saved future genera- tions of doctors from contending in vain with the breathless disability which is the distressing prelude to premature death in too many cigarette smokers throughout the world today.

88' The Scientific Background MEAN RATES OF CHANGE OF FEY OVER 5 YEARS IN MEN AGED 30-59 ACCORDING TO SMOKING HABITS AND MEAN FEY LEVELS 3 MEAN STANDAROIS E O FEY Ritresl < 15/daY 115 or more/daY c ----. _ ~ ~ ~ . . YEARS YEARS YEARS NON-SMOKERS ClGARE1 TE SMOKERS Figure 2 Respiratory tract' symptoms (%) reported by school according to smoking habits. children REFERENCES 1. ANDERSON, A.E, JR.; HERNANDEZ, J.A..; HOLMES; W.L., and FORAKER, A.G.: Pulmonary emphysema. Prevalence, se- verity and anatomical patterns in macrosections, with respect to smoking habits. Arch. En4imn. Health (Chicago) 12:569-577, 1966, 2 ANDERSON, D.O., and FERRIS, B.G., JR.: Role of tobacco smoking in the causation of chronic respiratory disease. New Eng. J. 1'Ked. 267t787-794, 1962. 3. ANDERSON, D.O.; FERRIS, B.G., JR., and ZICShTANTEL, R: The Chilliwach Respiratory Survey, 1963. Part 4: The effect of tobacco smoking on the prevalence of respiratory disease. Canad. Med. Ass. J. 92s 1066-1076, 1965.

C. M. Fletcher 89 4. AUERBACH, 0.; HAMMOND, E.C.; KIRMAN, D., and GAR- FINKEL, L: Emphysema produced in dogs by cigarette smok- ing. J.A.M.A. 199:241,251, 1967. 5. AUERBACH, O.; STOUT, A.P.; HAMMOND, E.C., and GAR- FINKEL, L: Bronchial epithplium in former smoken. New Eng. J. Med. 267:119 -125, 1962. 6. AUER'BACH, 0.; STOUT, A.P.; HAMMOND, E.C., and GAR FIIdKEL, L.: Changes in bronchial epithelium in relation to sex, age, resideace, smoking and pneumonia. New Eng. J. Med. 267:111-119, 1962. 7. AUERBACH, O.; STOUT, A.F.; HAMMOND, E.C., and GAR- FINKEL, L.: Smoking habit5 and age in relation to pulmonary changes. Rupture of alveolar septums, fibrosis and thickening of walls of small arteries and artarioles. New Eng. J. Med. 269:1045-1054, 1963. 8. BROWN, C.C.; COLEMAN, M.B.; ALLEN, RD.; STRANAr FIAN, A., and STUART-HARRIS, C.H.:, Chronic bronchitis and emphysema. Significance of bacterial flora in sputum. Amer. J. Med. 17:478-484, 1954. 9. BRUIVIFTIT, W.; WILLOUGHBY, M.L.N., and BROMLEY, L.L.• An evaluation of sputum examination in chronic bronchi- tis. Lancet 2:1306-1308, 1957. 10. BURROWS, B.; FLETCHER, C.M.; HEARD,, B.E:; JONES, N.L., and WOOTLIFF, J.S.: The emphysematous and bronchial types of chronic airways obstruction. A clinicopathological study of patients in London and' Chicago. Lancet 1:830-835, 1966. 11. BURROWS, B.; NIDEN, A.H.; FLETCHER, C.M., and JONES, N.L.: Clinical types of chronic obstructive lung disease in Lon- don and in Chicago. A study of one hundred patients. Amer. Rev. Resp. Dis. 90:14-27, 1964. 12 CAMPBELL, A.H, and ELDER, N.G.: The effect of tobacco smoking on the course of chronic bronchitis. Med. J. Aust. 2:9- 12, 1963. 13. CEDERLOF, R.; FRIBERG, L.; JONSSON, E., and %AIJ, L.: Morbidity among monozygotic twins. Arch. Environ. Health (Chicago) 10:346-350, 1965. 14. DEAN, G.: Lung cancer and bronchitis in Northern iteland, 1960•2. Brit. Med. J., 1:i506-1514, 1966. 15. DEANE, M.: Epidemiology of chronic bronchitis and emphysema in the United States. 2. The interpretation of mortality data. Med. Thorac. 22:24-37, 1965. 16. DEANE, M.: Personal communication, 1967. 17. DOLL, R., and HILL, A.B.: Mortality in relation to smoking: ten yeaxs" observations of British doctors. Brit. Med. J. 1:1399-1410, 1460-1467, 1964. 18. FLETCHER, C.M.: Bronchial infection and macxivity in chronic bronchitis. (In preparation, 1967). 19. FLETCHER, C:M.: Environmental factors in respiratory disease. In: Compston, N., Ed. Symposium on Advanced Medicine. Pro- ceedings of a Conference held at Royal College of Physiciaas. London, 1964. London, Pitman, 1965. pp. 243-254. 20. FLETCHER, C.M.; HILL, LD.• SPIEZER, F:S., and TII+TKER, C.M.: A prospective survey of bronchitis. Report to the Epide- miology Panel of the Medical Research Council's Committee on Research into Bronchitis. 1967. (Unpublished)

90 The Scientific Background 21. GANDEVIA, B., and MILNE, J.: Ventilatory capacity on expo- sute to jute dust and the revelance of productive cough and smoking to the response. Brit. J. Industr. Med; 22:187-195, 1965. 22. GREEN, G.M., and CAROLIN, D.: The depressant effect of ciga- rette smoke on the invitro antibacterial activity of alveolar mac- rophages. New Eng. .L Med. 276:421-427, 1967. 23. HAMMOND, E.C.: Evidence on the effects of giving up cigarette smoking. Amer. J. Public Health 55:682-691, 1965. 24. HAMMOND, E.C.: Smoking in relation to mortality and morbid- ity. Findings in first thirty-four months of follow-up in a pro- spective study started in 1959. J. Nat. Cancer Inst. 32:1161-1188,. 1964. 25. HAYNES, W.F., JR.; KRSTULOVIC, VJ, and LOOMIS BELL, A.L., JR.: Smoking habit and incidence of respiratory tract in- fections in a group of- adolescent males. Amer. Rev. Resp. Dis. 93:730-735, 1966. 26. HERNANDEZ,, J.A..; ANDERSON, A.E., JR.; HOLMES, WZ.., and FORAKER, A.G.: Pulmonary parenchymal defects in dogs following prolonged cigarette smoke exposure. Amer. Rev. Resp. Dis. 93:78-83, 1966. 27. HIGGINS, LT.T.: Tobacco smoking, respiratory symptoms, and . ventilatory capacity. Brit. Med. J. 1:325-329,, 1959. 28. HOLLAND, W.W.: The study of geographic differences in the prevalence of chronic bronchitis. Statistician (London) 16:5-22, 1966. 29. HOLLAND, W.W'.; REID, D.D.; SELTSER, R., and• STONE, R.W.: Respiratory disease in England and the United States. Studies of comparative prevalence. Arch. Environ. Health (Chi- cago) 10:338-343, 1965. 30. HUHTI, E.: Prevalence of respiratory symptoms, chronic bronchi- tis and pulmonary emphysema in a Finnish rural population. Field survey of age group 4Q-64, in the Harjavalta area. Acta Tuberc. Scand. SuppL 61:1-111, 1965. 31. KRUMHOLZ, RA.; CHEVALIER, R.B., and ROSS, J.C.: Cardio- pulmonary function in young smokers. A comparison of pul- monary function measurements and some cardiopulmonary re- sponses to exercise between a group of young smokers and a comparable group of' nonsmokers. Ann. Intern. Med. 60:603- 610, 1964. 32. SRUMHOLZ, R.A.; CHEVALIER, RB., and ROSS, J.C: Changes in cardio-pubnonary functions related to abstinence from smoking. Studies in young cigarette smokers at rest and exercise at 3 and 6 weeks of abstinence. Ann. Intern. Med. 62:197-207, 1965. 33. LAMB, D.; PASSEY, R.D., and REID, L. McA.: (In preparation, 1967). 34. LAURENZI, G~t1.; GUARNERI, JJ., and' ENDRIGA, R.B.: Im- portant determinants in pulmonary resistance to bacterial infec- tion. Med. Thorac. 22:48-54, 1965. 35. >.AURENZI, GA.; POTTER, R.T., and KASS, ESi.: Bacteriologic flora of the lower respiratory traot. New Eng. 1. Med. 265:1273- 1278,1961. 36. LEES, A.W., and McNAUGHT, W.: Bacteriology of the lower- respiuatory-tract secretions, sputum, and upper-respitatory-tract N O N IPA W

C. M. Fletcher 91 secretions in "normals° and chronic bronchitis. Lancet 2:1112- 1115, 1959.. 37. LEMON,, F:R., and WALDEN, R.T.: Death from respiratory sys- tem disease among Seventh-Day Adventist men. JA.M.A. 198:117-126, 1966. 38. LUNDMAN, T.: Smoking in relation to coronary heart disease • and lung function in twins. A co-twin control study. Acta Med. Scand. 180(Suppl. 455) :175, 1966. 39. MEDICAL RESEARCH COUNCIL (Gt. Brit): Definition and classification of chronic bronchitis for clinical and epidemiologi- cal purposes. Report to the Medical Research Council by their Committee on the Aetiology of Chronic Bronchitis. Lancet 1:775-779, 1965. 40. MEDICAL RESEARCH COUNCIL (Gt Brit): Value of chemo- prophylaios and chemotherapy in early chronic bronchitis. A re- port to the Medical Research Council by their Working Party on Trials of Chemotherapy in Early Chronic Bronchitis. Brit4 Med. J. 1: 1317-1322, 1966. 41. MEGAHED, G.E.; SENNA, G.A.; EISSA, M.H.; SALCH, SZ., and EISSA, H.A.: Smoking versus infection as the aetiology of bronchial mucous gland hypertrophy in chronic bronchitis. Thorax 22:271-278, 1967. 42. MITCHELL, R.S.; VINCENT, T.N., and FILLEY, G.F.: Ciga- rette smoldng, chronic bronchitis, and -emphysema. J.A.M.A. 188:12-161 1964. - 43. MORK, T.: A comparative study of respiratory disease in England & Wales and Noraray., Acta Med. Scand. 172(Suppl. 384):1-100, 1962. 44. OLSEN, H.C., and GIL,SON, J.C.: Respiratory symptoms, bron- chitis, and ventilatory capacity in men. Aa Anglo-Danish com- parison, with special reference to differences in smoking habits. Brit. Med. J. 1:450-456, 1960.. 45. OSHIMA, Y„ ISHIZAKI, T4 KABE, J., and MAKINO, S.: A study of Tokyo-Yokohama asthma among Japanese. (Notes) Amer. Rev. Resp. Dis. 90:632-634, 1964. 46. PARNELL, J.L.; ANDERSON, D.O:, and RINNIS, C.: Cigarette smoking and respiratory infections in a class of student nuxses. New Eng. J; Med. 274:979-984, 1966. 47. PASCUA, M.: Evolution of mortality in Europe during twentieth century. Epidemiol: Vit. Stat. Rep. 5:1-136, 1952. 48. PASCUA, M.: Increased mortality from cancer of respiratory sys- tem. Bull. WHO 12:687-703, 1955. 49. PETERS, J.M., and FERRIS, B.G:, JR: Smoking and morbidity in a college-age group. Amer. Rev. Resp. Dis. 95:783-789, 1967. 50. PETTY, T.L.; RYAN, S.F, and MITCHELL, R.S.: Cigarette smoking and the lungs. Relation to postmorbem evidence of em- physema, chronic bronchitis, and black lung pigmentation. Arch. Environ. Health (Chicago) 14:172-177, 1967. 51. RANKIN, J.; GEE, J.B, and CHOSY, LW.: The influence of age and smoking on pulmonary diffusing capacity in healthy sub- jects. Med. Thorac. 22:366-374, 1965. 52. READ,, J., and SELBY, T.: Tobaoco smoking and ventilatory function of the lungs. Brit. Med. J. 2:1104-1108, 1961. 53. REID, - D.D.; ANDERSON, D.O.; FERRIS, B.G., and I

92 The Scientific Background FLETCHER, C.M.: An Anglo-American comparison of bron- chitis. Brit. Med. J. 2:148T 1491, 1964. 54. ROYAL COLLEGE OF PHYSICIANS OF LONDON: Smoking and Health. Summary and Report on Smoking in Relation to Cancer of the Lung and Other Diseases. London, Pitman, 1962. par. 47, pp. 28-30. 55. SLUIS-CREMER, G.K.; WALTERS, LG., and SICHEL, H.S.: - Chronic bronchitis in miners and non-miners: an epidemiological survey of a eommunity in the gold mining area in the Transvaal. Brit J. Industr. Med. 24:1-12;,1967.. 56. SLUIS-CREMER, G.K.; WALTERS, LG., and SICHEL, H.S.: Ventilatory function in relation to mining experience and smok- ing in a random sample of miners and non-miners in a Witwa: tersrand Town. Brit. J. Industr. Med. 24:13-25, 1967. 57. STANEK, V:; FODOR, J.; HEJL, Z.; WIDIMSKY, J.; CHAR- VAT, P.; SANTRUCEK, M.M.; ZAJIC, F:, and VAVRIK„ M.• A contribution to the empidemiology of chronic bronchitis. Acta Med. Soand, 179:737-746, 1966. 58. THURLBECK,,W.M.; ANGUS, G.E., and PARE, J.P.: Mucous gland hypertrophy in chronic bronchitis, and its occurrence in smokets. Brit. J. Dis. Chest 57:73-78, 1963. 1 59. U.S. SURGEON GENERAL'S ADVISORY COMIvIITTEE ON SMOKING AND HEALTH: Smoking and Health. Report of the Committee. Public Health Service Pub. No. 1103. Washing- ton, D.C., U.S. Govt. Print Off., 1964. (a) p: 90;, (b) p. 92; (c) .p. 94; (d) p. 107. 60. WYNDER, E.L.; LEMON, F.R., and MANTEL, N.: Epidemiol- ogy of persistent cough, Amer. Rev. Resp. Dis. 91:679-700, 1965. 61. ZP.MEL, N.; YOUSSEF; , H.FI., and PRIME, FJ.: Airway resis- tance and peak expiratory flow-rate in smokers and non-smok- ers. Lancet 1:12374238, 1963. 62. ZWI; S.; GOLDMAN, H.L, and LEVIN, A.: Cigarette smoking and pulmonary function in healthy young adults. Amer., Rev. Resp. Dis. 89:73-81, 1964. I The final speaker of the opening session of the World Conference on Smoking and Health was ir Austin Brad- f'ord Hill, Professor Emeritus of Medical Statisttcs vers~ty odon, England. His o ows: In further researches into the effect of smoking upon the public health, we are faced with two domi- natimg.problems-is the researc ra ca e ethical~ nd prac"t`~ica ie'ly to prov~ e iYe ' greater stumblinl ock "To"'~peeific example we shall continually be faced with claims for the "safe" cigarette. How can we satisfy ourselves that the claim has substance?

I Austin Bradford Iiill 93 In spite of obvious limitations, the simplest and quickest approach would undoubtedly be through animals. If, in comparison with some present stan- dards, we were to find that the "safe" cigarette was no less carcinogenic to the skin of mice or, let' us say, had no less effect' upon the ciliary action of guinea pigs, then I think we should automatically dismiss the claim. Of' course we may be wrong in doing so. What is true of mice and guinea pigs is not necessarily true of man. But we often do have to take action on just that basis-for example, in the introduction of new drugs. No one would dare to market a new drug which had pronounced teratogenetic effects upon rabbits or invariably produced neoplasms in the liver of the rat. And the onus of proof of' innocence lies with the manufacturer. But suppose the cigarette passes our very limited animaI tests? What then about man? Can we.ever make a strictly controlled trial of a cig, arette, such as we have used so snccessfully~to;es, 'tablish'~the--value` of inoculations against infectious, diseases: -a _ _ . -T doubt it. The scale would have to be large and ~~ the time probabiy, very prolonged. And, still more 'f difficult, randomization of the participants in the trial might prove impossible. And then the ethical problem might step in. In such a trial, perhaps in- volving cheap or free cigarettes, we might be encour- aging persons to continue smoking who would other- wise have given up. As in nearly all epidemiological work we shall - -- - - - ---- - -- therefore be forced to rely_up9n_eareful observation of what man _ad _woman _ themselves clwose_~o do and with what result. What they will choose to do is , not mX idea, or your idea, of a nice clear-cut experi- ment. It wdl be riddled-witli seleetiv-c"_ ' c rs. ___... But' we cannof help -that:' ~`e s have'to inter- pret the answers critically in the light of' what' we al- ready know..

94, The Scientific Background In relation to cancer of the lung, there is one fea- ture of what we already know that I have found impressive-the histological findings of Auerbach and Hammond. In their study of sections from the bronchial tree they found certain atypical cells in , cigarette smokers and disintegrating cells in persons. I! who had given up smokin,, Would it be possible to i use that line of evidence? Could we plan to examine e, such sections of persons coming to autopsy and re- late the findings to changed or unchanged smoking l , habits? If the "safe" cigarette is safe we should presum- ably not find,, or less frequently find, the atypical l cells in persons smoking them; rather we should see the disintegrating cells as in the givers-up. There is sometimes one way of avoiding the selec- five factors which are so prominent in Nature's ill- designed experiments. That is to consider the whole _._- -- population. -Foi' instance, with the giving up of cigarettes by an appreciable proportiom of our observed group of British doctors, Doll and I found that in a few years "there had been a fall in their death rate from cancer of the lung. No selection n of genetic factor could con- _ fuse the issue_ because we were observing the whole population-just as 100 years ago we saw a declin= °-- ing death rate in~ the whole population from typhoid fever when we ceased to sell sewage and water as a thirst-quencher. Perhaps the time has come when we should see some such effects if cigarette filters in general are all that the public appears to believe. Is~not' their use sufficiently widespread and sufficiently prolonged to reveal results? Another problem likely to be given attention is the profile of the smokers, their physical and psy- chological features. All smokers do not get chronic bronchitis, coronary thrombosis or cancer of the ~ lung and it would be enormously helpfull if~ we could ~ distinguish who would and who would not. Dr. •' Stamler has lucidly discussed this problem in rela-

Austin Bradford Hill 95 tion to heart disease. But with cancer of the lung we' have no corresponding data. And it is not what the public really wants. If one in X men smoking two packs a day eventually gets cancer of the lung, they want to be assured that they are not the one. This I suspect is a mere will-o'-the wisp-that there's no hope at all of defining persons in this way. The underlying factors are unlikely to be simple fac- tors. Demogra hicall the association of_eiguette smoking ann cancer of e ung has been~shown over an, _- ~orl~de range of genetics,and environmentsy w`"Yf" e onsider animals inbred for generations,-they do not all get skin tumors when painted with a known carcinogen or aII die when injected with a -drug. In the clinicaI trial of a drug all patients do n~o~_t _r~e nd and„gven in spite of'caref_u~'observation a~ n~easurement we cannot dctect wh~ .~~ Tf we are unable to de'tect'tiie eatures in these relativily well-controlle-d situahons, s it~cef yat we can do_sq,,,,,in~he mueh wider and more general field of smokers~ " - ``Perhaps, too, there may be nothing to look for. It may be a1L a question of chance hits. Neither with animals nor humans can we repeat such an experi ment We cannot know whether the same creatures would always respond in a particular way to a par- ticular stimulus. So personally I would not want to put much effort into this problem. The main issue todag,_however, is_how do_ we teach`cIiildren not-to starZ..smorkiag2 and how do we heip"'W quit tliMose af the adult population who seek to do so? . From the point of -view of research, the would be quitters present the easier problem. The conduct of a controlled trial of a drug with alleged helpful prop- erties will be no different from the conduct of any controlled trial in patients. If there is no real evi- dence that the drug is beneficial then there can be no ethical problem in not giving it' to a randomly consti: tuted section of the clientele.

96 The Scientific Background Indeed, if there is no evidence to justify your giv ing it, are you behaving ethically in giving it to ev- eryone? Would it not be more ethical if you were first to seek some justification for your action? And if you have no tablets of proven value, I see no ethi- cal problem in using placebo tablets alongside. However, if this sticks in the gullet, I believe I would settle for a trial' in which clinics were random- ized rather than people. Their success rate has been so very low that any pronounced changes in only those clinics using a particular drug would be diffi- cult to interpret except' as cause and effect. It is sometimes said that we may induce anxiety in those who find they cannot quit Has anyone tried to find out? Are those unable tQ quit any more anxious after than before the event?' And, if so, are we to sacrifice to them those whom we can help quit? There is also a good deal said about the teaching and propaganda disturbing equanimity or making neurotics as well as interfering with personal freer dom. But advances in public_ health always have in- terfered with perso'~Ina fteedom. We make the public drink pastenrised milk and fluoridated water. We in- sist sist on inoculations-for fear of disease-or on driv-- ~j ing cars at prescribed speeds and curb drills for chil- a dren-for fear of sudden death. There is really no s dearth of precedent, whether by dictation or educa- tion. Is this problem so very exceptional? But I stray from the problems of research and would merely add that personally I would always re- main sympathetic to the views of Her Britannic Ma- jesty's Parliamentary Under Secretary of State for Commonwealth Affairs. In a debate in The House of Lords on smoking in commercial aircraft, he was implored to see that smokers were not' entirely elimi nated from the passenger service. "My Lords," he replied, "I quite agree that those who smoke have every right, if they so wish, to ex- pedite their journey to another plaee."

---