Tobacco Documents Online

~ e theory that a~herosclerasis e s:. ur _~ r~ ul~ beg~n~c era ~ ~' 0 a 0 .: ~,. by T}r. gto~ie ea o eart mor 'Pr t an 3.•`millic bl~ood vessel diseaes were~at~ i6ut a arfiei~os .x the causesw o f atherosc<~lerosas'~, t~e tnost - com ,~r~t n. fo ontanicec~ einphasls rr~tas~'be pt~cetl on resea s_ ~. 0 co afa.gn's killer~s ~o s ~nc~ ~h~~796E1's ~ise~kdeah rg% -30 er an Q.f. ; Yolaria, tyctr 0 ~~ heart ±di. ~br~n_aiy-_ e; rleclyieiii: cor~i% d AMh 7_n;th~e ~ . .- ~~~"-s~': d: its ciisease op_u_rrt.e.noscMW_ iore 0 0 ~..:- e ' aCciiunts w a1~ a cen •- The leading cause of death: in theviTnited 5tate~s : : ..-.~ two-thirds of all cardiovascular deathsr„-. _ ' ' • Cardiovascular diseases have led the list af and they remain so today despite a steady ~ec ne • Cardiovascular mortahtvvhas decre dechne has accelerated so much~that _oyez: In eontras~, many countries sucli a~ 'Romani have repoited increases in mortality rates from ~ • Although there is general agreem~ a t___~ tlie,U.S.] is real, the probable cause ar_causes_l~o r ~.sT there are,too many potentLa,l.causes:rather than

Cigarette Smoking and Heart Disease The Tobacco Institute 1875 I Street Northwest, Washington, D.C. 20006 1983

Cigarette Smoking and Heart Disease Introduction Eighteen years ago, the first Surgeon General's report reached a narrow and simple conclusion about smoking and heart disease*: "Male cigarette smokers have a higher death rate from coronary artery disease than non-smoking males, hiit it is not clear that the association has causal significance." A nd : "...rTlhe basic cause or causes of coronary heart disease are obscure..."1 Four years ago, in his 1979 report, the Surgeon General revised those conclusions: "...It can be concluded that smoking is causally related to coronary heart disease in the common sense of that idea and for the purposes of preventive medicine." *For the purpose of this paper, heart disease, coronary heart disease (rHD) and ischemic heart disease (IHD) are assumed to he synonymous. With the exception of direct quotation, the text uses "heart disease" only. i

And in another passage from the same report: "There is no reasonable doubt that cigarette smoking as a risk factor for...cardiovascular diseases has been proven."2 Early in 1983, the Surgeon General is expected to issue a further report, devoted entirely to this subject. While we can speculate on what it may say, it seems proper to consider the sig- nificance and meaning of what has already been said and to indicate some of the subsequent contributions to scientific knowledge. The operative words in the previous conclusions are "associa- tion," "causally related," "cause" and "risk factor." A few com- ments on these are appropriate. On the first of those words, the 1964 Surgeon General's report said that "results of investigations must be considered to deter- mine first whether an association actually exists between an at- tribute or agent and a disease....The causal significance of an association is a matter of ,judgment...."1 That first report listed "consistency of the association" as a major criterion for such a,judgment. As will be seen, there are many inconsistencies in results of investigations of smoking and N) cn heart disease. o ~ -p -~ ~ 0 The meaning of the phrase "causally related" is obscure. The N ii

advisory committee that prepared the first Surgeon General's report said the meaning of "cause" was "discussed vigorously" in "debates" among the members, and that no member "used the word 'cause' in an absolute sense." The concept of "risk factor," perhaps the most gentle of these operative words in heart and other diseases, arose mainly from the Framingham study. Tn 1949, the U.S. Public Health Service began a close surveil- lance of more than 5,000 adult men and women in the community of Framingham, Massachusetts. Tts major objective was to attempt to determine why individuals would develop evidence of heart disease. The Framingham researchers were to utilize direct observation and questionnaires to record the variables -- genetic traits, environ- mental characteristics, lifestyle and any other factors -- believed to be related to heart disease. They were to look at which of these variables were most common in those persons who did develop symptoms of heart disease. And they would attempt to determine statistically the relative importance of each in the occurrence of those symptoms. Those deemed important would be called "risk fac- tors," in the sense of common presence, not necessarily cause. The Framingham study originally found "relationships" between heart disease and high serum cholesterol level, high blood pres- sure, obesity, low lung capacity and cigarette smoking.3 These ~ were adjudged to be "risk factors." o F-4 44 ~ V W

Sixteen years later, the Framingham director, William Kannel, wrote that elevated blood pressure had been confirmed as the "dom- inant contributor" to heart disease in the study, but that re- searchers were continuing to study the possible role of other "risk factors" in the development of heart disease.4 It must he remembered that by 1978 "risk factors" were no longer thought of merely as statistical "relationships" by many in the medical community. In the years between the publication of these two Framingham reports, the role of high blood pressure in heart disease had been elevated to "dominant contributor." The possible role of smoking as one "risk factor" among many was con- tinuing to be studied. In 1979, the author of a chapter in a medical textbook wrote that all known "risk factors" taken together could account for approximately 50 per cent of an individual's "risk" of developing heart disease in the United States. He added that important risk determinants remained to be discovered.5 In 1979, the Surgeon General's report said that relatively little was known about the mechanisms by which smoking was alleged to enhance atherosclerosis* or to increase the "risk" of heart attack. *Atherosclerosis, a form of arteriosclerosis, or thickening of the arteries is thought to be a principal factor in the development of heart disease. rv cn 0 ~ .p .P. w a ~ .P iv

The 1979 report also said that smoking was not a necessary condition for atherosclerosis and heart attack, as these occur in nonsmokers, and that correlation is not synonymous with causation. It is generally recognized that statistical correlates derived from studies such as Framingham could identify certain characteris- tics that might possibly be related to an individual's risk of de- veloping heart disease. Identification of so-called "risk factors" does not necessarily mean that the cause or causes of heart disease have been discovered. Although the 1979 report alleged smoking is "causally related" to heart disease in the common sense of the idea and for the pur- pose of preventive medicine, it suggested that additional research on mechanisms and on a more precise quantification of certain "risk factors" through epidemiological studies* was an important topic for medical science.2 It would be difficult to find fault with these suggestions for additional research. Without understanding of and knowledge about disease mechanisms, there can he no certainty about what causes disease. In fact, the entire concept of "risk factors" might well deserve reevaluation by the medical and other scientific communities. *F;pidemiology is a statistical science in which a group of people is studied to determine how often a disease occurs and what factors might be related to or associated with it. nj rn 0 ~ -p ~ w 0 v cn v

The material that follows examines whether smoking has, in fact, been established scientifically as "causally related" to heart disease. It looks also at whether support for the claim that smoking is an important "risk factor" is as strong as some have suggested -- all in light of some of the scientific evidence presented mainly since 1979 and reviewed here. N CJ1 O ~ ~ -p W 0 V ON vi

Intervention Studies Multiple Risk Factor Intervention Trial (MRFIT) If elevated cholesterol level, hypertension* and cigarette smoking were "risk factors" for heart disease mortality,** then re- ducing them in people presumably should lower the mortality rate. However, results reported from recent intervention studies, such as KRFIT, raise doubts as to whether the concept of "risk factors" carries the significance the medical community has ascribed to it up to this time. The results of this "massive, expensive, and lengthy clinical investigation termed 'Multiple Risk Factor Intervention Trial (hRRFIT) "'6 were reported in the Journal of the American Medical Association in September 1982. MRFIT was designed to test the effect of "risk factor" reduc- tion. From among more than 300,000 volunteer American men, 12,866 healthy but "high risk" subjects were assigned at random to one of two groups. "High risk" was determined by smoking history, serum cholesterol levels and diastolic blood pressure readings. *Hypertension is high blood pressure. N vn 0 ~ **Mortality rate and death rate are used synonymously in this 4:k paper. w a ~ ~ 1

Figure 1: Results from Multiple Risk Factor Intervention Trial (MRFIT) Percentage Reported Smoking at Beginning and End of Study 63.8% 63.5% 45.6% Beginning End ® SI is Special Intervention group ~ UC is Usual Care group Death Rates per 1000 at 72 Months All Causes 22.5 CVD* including CHD** CHD** *Cardiovascular Disease **Coronary Heart Disease Source: From Multiple Risk Factor Intervention Trial Research Group, "Multiple Risk Factor Intervention Trial: Risk Factor Changes and Mortality Results," JAMA 248(12): 1465-1477, Sept. 24, 1982 The special intervention group (SI) received special treatment for high blood pressure and physician counseling on smoking and diet. Members of the usual care group (UC) were left to their regular sources of health care. It was anticipated that at the end of six years the death rate from heart disease in the special intervention group would be reduced by 26.6 percent below that of the usual care group. Although there was, after almost six years, a 46 percent re- duction in smoking among the SI group and a 29 percent reduction among the UC group, there was no significant difference in mortali- IV cn 0 ~ ~ 4N w 0 ~ 03 2