Tobacco Documents Online

ETS and Cardiovascular Disease • 147 le Ir The results compared ETS exposed husband's and non-ETS exposed hus- bands, where the husbands had never smoked. None of the endpoints showed' statistical significance between the two groups, before or after adjustment for several variables, including age, baseline blood pressure,,chol'esterol, weight, alcohol consumption and education. However, within the exposed group, increasing levels of cigarettes smoked daily by the wife had a statistically significant dose response re- lationship with husbands' CHD deaths. This is technically significant (p = 0.04) but is based on only one death in the 1-19 cigarettes smoked/day cate- gory. A second analysis lumped never smoking husbands with ex-smoking hus- bands, calling these non-smoking husbands. This group was then evaluated on the basis of the smoking status of the wife. Non-smoking, husbands of smokers did not show a statistically significant result when compared with husbands of non-smoking wives for mortality from CHD (p = 0:15) or from CHD itself as an endpoint (p =.10)L Several methodological problems exist in the Svendsen report. One prob- lem is possible misclassification of husband's smoking status either at entry or subsequently: A second problem is that the wife's smoking status was based on interviews with the husband, andi not on direct questioning of the wife. There is also an alcohol}related bias, as MRFIT ETS-exposed husbands' had two drinks per week, on average, more than non-ETS exposed husbands, and this alcohol! effect could explain the observed statistical significance in, dose response. Finally, by combining ex-smoking husbands with never smokers, Svend- sen confounds any past effects of active smoking by the husband with expo- sure to ETS. The MRFIT study serves as an exemplary prospective trial for its design and conduct. However,, lack of statistical significance, failure to control for several confounding variables (such as alcohol consumption), mi'sclassifica- tion, and misgrouping make it difficult to draw any conclusions from the study. 5. Helsing. The Helsing (1988) paper examines death certificates collected from July 1963 through July 1975 for a population living im July 1963 in Washington County, Maryland; This is based: on underlying cause of death of arteriosclerotic heart disease including coronary disease (International Classification of Disease [1CD] -420) and other myocardial' degeneratiom (1CD 422)i As of July 15, 1963, 98% of the residents were asked questions that included information on sex, age, race, marital status, years of schooling, housing characteristics, information on cigarette, cigar and' pipe smoking habits, as welli as frequency of church, attendance, for each household mem- ber aged 16.5 years or older. Among 22,973' white men and 25,369 white women 25 years of age and

148 • Envi.onmental Tobacco Smoke older in the 1963 census, 4,1162 men and 14,873 women reported'that they had never smoked. The 1971 follow-up population was a subset of these numbers: 3,4541 men and 12,345 women. The results showed that death rates from, arteriosclerotic heart, disease were higher among men (relative risk = 1.31) and women (relative risk = 1.24)who lived with smokers in 1963, after adjustment for age, marital sta- tus, years of schooling and qualiry of housing index. For women, relative risk increased significantly (p <.005): with increasing 1'evel's of exposure,, but for men, there was little evidence of' a dose response relationship. Several methodological' problems exist with the Helsing paper. The first major problem is that the only smoking data that was collected on every person was in~ 1963, Hence, no changes in smoking habits over the 12-year period were ascertained. In additiony no data were collected on other risk factors for heart disease such as diet, exercise, blood pressure and cholesterol'. Finally, no ETS exposure outside the home was measured: B. Case-Control Study 1. Lee.,The Lee (1986) study is a case control (retrospective) study to eval, uate the possible relationship between cigarette smoking and risk of lung cancer, chronic bronchitis, ischemic heart disease and stroke. The original questionnaire was administered in ten hospital regions in England; between 1977 and 1982. Although not recorded initially, ETS exposure data was sub- sequently collected in 1979 for married ~ patients in the last four regions. Two hundred cases and 200 matched controls were collected'for each sex (male, female) and age (35-44, 45-54, 55-64, and 65-74)', grouping to ex- amine the possible relationship between ETS exposure and diagnosis of isch- emic heart disease. Also matched were hospital region ands when possible, }iospital i war&and time of interview. Ischemic heart disease cases and controls did not show a statistically sig- nificant difference in, their exposure to ETS, based either on smoking habits of spouses or on an index accounting for exposure at home, at work, and during travel and leisure.. Although the Lee study is one of the few to attempt to examine non- spousal ETS exposure, it raises the general methodological issues that sur- round retrospective case control studies. In its finding of non-statisticali significance for any trends of association between ETS and cardiovascular illness, the Lee paper confirms the need for execution of better, controlled prospective trials. C. Experimental' Design 1. Aronow. The Aronow (1978), paper, describes an experimentaL design to examine the possible relationship between exposure to ETS and exercise-in- duced angina in both a well ventilated and an unventilated room.

ETS and Cardiovascular Disease • 149 The design induded ten men (eight non-smokers and two smokers) who exercised upright on a bicycle ergometer with a progressive work load until the onset of angina pectoris. Subjects were randomized to three groups: noo smoking, smoking in a well ventilated rooms or smoking in an unventilated room. Aronow has suggested that the results of his study demonstrate that, un- der the conditions of the experiment, ETS exposure causes angina] pain to develop soon after exercise. In addition, the data from the study indicate that exposure to ETS causes an increase in carboxyhemoglobin, more after ETS exposure im am unventilated room than~ after ETS exposure in a ventilated room.. Severat major criticisms of the Aronow study include: (1) ~ the use of sub- jective pain as an end point without double blinding; (2) a very small sample size that can lead to a large variance based on just one or two subjects chang- ing their responses, (3) problems associated with the Hawthorne effect [sub- jects tend to produce symptoms suggested to them], and (4) failure to control for stress. D. Conclusions The Surgeon General's Report of 1986 (1986) examined the studies of Hirayama, Gillis, Garland and Aronow,, and concltyded that "further studies on the relationship between involuntary smoking and cardiovascular disease are needed in order to determine whether involuntary smoking increases the risk of cardiovascvlar disease." The National Research Council (1986): in 1986 reviewed the prospective studies of Garland, Gillis, Hirayama and Svendsen, as well as severaliexper- imental designs examining the biological plausibility of the association of ETS and cardiovascular disease, and concluded that: 1. No statistically significant effects of ETS exposure on heart rate or blood pressure were found inihealthy men, women, and school-aged children dur- ing resting conditions. During exercise there is no difference in the cardio- vascular changes for men and women between conditions of exposure to ETS and control conditions. 2. With respecrto chronic cardiovascular morbidity and mortality, although biologically plausible„there is no evidence of statistically significant effectss due to ETS exposure, apart from the study by Hirayama in Japan. It is the opinion of' this author that none of the studies critiqued in this paper provides any basis for altering the Surgeon General's and' NAS's con- clusions concerning ETS and cardiovascular disease. This conclusion is reinforced by the findings of Schieveblein and Richter (1984). They report that, under real-life conditions, persons exposed to ETS inhale only approximately .02 to .01 of the amount of particulate matter

150 • Environmartat Tobacco Smoke taken up by active smokers. Also, nicotine concentration in serum~ of ETS- exposed individuals is within a range that is barely distinguishable from the background level, and'the increase in carboxyhemoglobin rarely exceeds 1%. The authors conclude that exposure to ETS "is not likely to have an effect on the development and progression of CHD." Il. Recommendations for Future Research To provide meaningful recommendations for future research, it is necessary to evaluate the existing studies of ETS exposure and cardiovascular disease in light of the five criteria for causality discussed at the beginning of this paper. The NAS concluded that a relationship between ETS exposure and car- diovascular disease is biologically plausible, and each of'the studies reviewed in this paper appears to provide an adequate temporal association between ETS exposure (as measured by spousal smoking) and the onset of cardiovas- cular disease. However, the studies fail to meet one or more of the remaining criteria for causality. Of the six studies concerning ETS exposure and cardiovascular morbidity and mortality, only two (1-lirayama and Helsing) 1 reported statistically signif~ icant relative risks for exposed' compared to non•exposed populations, and neither study reporte& a relative risk greater than 2. Hirayama reported a dose dependent relationship but Helsing did not. None of the studies demonstrate a specificiry, of association between ETS exposure and cardiovascular disease. Each of the studies fails to control for one or more important confounding variables, including lifestyle, blood pres- sure, serum cholesterol; obesity and socioeconomic status. None of the stud- ies provides an accurate measurement of' ETS exposure. Alll of the studies suffer from one or more serious methodological problems, including small sample size and possible misclassification of spousal smoking status. These confounding variables and methodological problems also preclude any dem- onstration of consistency of association among the existing studies. in view of' the inadequacy of existing studies, it is logical to consider whether the Framingham Heart Study might provide an adequate basis for a definitive evaluation of the relationship between ETS exposure and heart disease. The Framingham Heart Study, initiated during 1948-1950,,is comprised of a study cohort from~ a random subsample of the adult residents of Fra- mingham, Massachusetts, of which 69% responded~ No reports on ETS and heart disease have been published under the Study, but spousal smoking hab- its could be determined from the Study's data base. Although an effort could be made to measure ETS effects on~ heart disease in the Framingham Study, this would not be likelw to provide an adequate basis for a definitive evalu- ation of' ETS an&hean disease. I

ETS and Cardiovascular Disease • 15l The critical, problem is that the Framingham, Study does not provide a basis for an accurate measurement of ETS exposure, especially outside of the home. Use of'data on spousal smoking habits as a surrogate for, ETS exposure has been shown to present serious methodological and other problems in existing studies. In any event, the recent paper by Seltzer (1989) suggests that the Fra- mingham Study is not likely to; show a significant association between, ETS exposure and heart disease. Seltzer's paper, compares the Surgeon General's statements regarding the association between active smoking and heart dis- ease with the data in the Framingham Study. Seltzer points out that the Fra- mingham data differ from the Surgeon General's conclusions in several im- portant respects: 1. The Surgeon General asserts a four-fold greater CHD incidence in men who are heavy smokers as compared to non.smokers; Framingham re- ports relative risk ratios less than two. 2. The Surgeon General asserts that cigarette smoking among women has a predictive association ~ with CHD; Framingham finds no such~association. 3. The Surgeon General states there is an increase of CHD with increase of duration of smoking; in the Fratningham Study, this increase is absent. 4. The SurgeomGeneral claims that rates of CHD eventually are reduced im ex-smokers to:those somewhere between smokers and non-smokers, and somedmes, after many years, falling to the level of non-smokers. The Framingham data are surprising in that reductions in CHD among ex- smokers is below levels for never smokers! This suggests that a selection bias may exist. Given the relatively small effect of active smoking on heart disease re- ported in the Framingham Study, it appears unlikely that any effect of ETS exposure on heart disease could be measured under that Study. In view of the lack of adequate existing data, future studies need to be performed that carefully examine the relationship between exposure to ETS and cardiovascular disease. It is the hope of this author that the critiques presented in this paper, examining many of the methodological problems as- sociated with existing ETS epidemiological studies, will be of use to well- traine&scientists. Familiarity with the five key points of causality in epide- miology is critical in designing studies that can clearly show whether any association exists between exposure to ETS and cardiovascular disease. Based on the analysis in this paper, a meaningful future study should contain at least the following elements: 1. A representative sample large enough to yield adequate statistical power. 2. A design that provides control'r for important confounding variables, in-

duding blood pressure, diet, alcohol consumption, plasma cholesterol, body weight, sex, socioeconomic status and exposure to environmental substances other than ETS. 3. A mechanism for accurate measurement of ETS exposure, including ex- posure outside the home, and' adequate follow-up of exposure status. 4. A prospective desigm specifically developed to satisfy the criteria for causality. Aronow, W.S. (1978). Effects of passive smoking on~angina pectoris. New England Journal of Medicine 299(1)21-24. Environmental Tobacco Smoke - Measuring Exposures and Assessing Health Effects, National Research, Council, National Academy Press, Washington, D.C.,1986. Garland, C. et. al. (1985): Effects of passive smoking on, ischemic Heart disease mor- tality of non-smokers. American Journal ofEputemiology 121(S). Helsing, K.J,, D.P. Sandlers G.W. Comstock, and E. Chee, (1988). Heart disease mor- tality in non-smokers Gvingwith smokers. American Journal of Epidemiology 127(5). Hirayama„T. (1984). Lung cancer in Japan: Effects of nutrition and passive smoking, Lung Cancer: Causes and Prevention. Verlag Chemie Ihternational; Inc. Hirayama, T. (1981). Non-Smoking wives of heavy smokers have a higher risk of lung eancer: a study from Japan. British Medual Journal 282:183-185. Holt, D.J., C.R. Gillis et. al. (1989). Passive smoking and cardiorespiratoryhealth in a generallpopulhtion in the west of:Scotland. Brirish' Medical Journal 299:423- 427. Lee, P.N:, J: Chamberdain, and~IvS.R. Alderson, (1986): Relationship of passive smok- ing to risk of lung cancer and other smoking-associated diseases. British Journal' of Cancer 54:97-105. Schievelbein, H. and F. Richter (1984). The influence of passive smoking on the car- diovascular system. Prevention Medicine, 13:626-b44. Seltzer, C. (1989). Framingham study data andestablished wisdom about cigarette smoking and coronary disease. Journal of Clinical Epidemiology 42(8):743-750: Surgeon General's Report (1986)1 The Health Consequences of Involuntary Smoking. U.S. Department of Health and Human Services. Svendsen, K.H., Lewis H. Kuller et. al. (1987). Effects of passive smoking in the muW tiple risk factor intervention trial: American Journal'of Epidemiology 126(5): (= ~.•'i r.~ ~7