Passive Smoking and Coronary Heart Disease in Women
To the Editor:
In a recent article by Kawachi et al,1 the differing prevalence of cardiovascular risk factors and other lifestyle variables in women with and those without tobacco smoke exposure is attributed to the lower socioeconomic status of the female passive smokers. In order not to impute the socially determined elevated coronary heart disease (CHD) risk to the passive smokers, the authors adjusted their data by standardizing for these factors. In so doing, however, they may have compounded another error, casting serious doubt on the validity of their results. They overlooked the fact that women from lower social classes in general pay far less attention to their health and visit their doctors less frequently than do women from higher social classes. This is particularly true in the United States, where financial reasons also play a role. The exclusion of about 3500 women with a history of CHD before the study commenced, as well as the exclusion of an unspecified number of women with CHD in the course of the study inevitably led to a higher prevalence of undiagnosed CHD cases in the group of passive smokers, so that it is really not surprising that a higher incidence of myocardial infarction was found in the women exposed to tobacco smoke. Given the selection bias, however, a causal relation is not substantiated in this study.
Selection bias is supported also by another finding reported in the article by Kawachi et al. Tobacco smoke exposure at the workplace must have decreased during the study (1982 to 1992) not only for the reasons given by the authors but also because of the aging of the study population. The oldest volunteers who, by virtue of their age were most at risk of developing the disease, were already 61 years old when the study began and, at 71 years of age, had long since retired from the workplace at the time the study was concluded. Despite this ever-decreasing tobacco smoke exposure, the relative CHD risk of passive smokers increased steadily in the course of the study, rising from 1.6 after 4 years to 2.0 after 6 years and eventually to 2.3 after 10 years. This finding, too, is best explained by the selection bias described above. Again, it appears to be justified to assume that there remained from the very beginning and in the course of the study a larger number of women with undiagnosed CHD in the group exposed to tobacco smoke than in the control group. Therefore, passive smoking need not necessarily play a role in this increase in CHD risk.
To support the plausibility of a CHD risk of 1.91 from regular passive smoking, Kawachi et al refer to a study of their own,2 which showed a fourfold to fivefold increase in CHD risk in active female smokers compared with female nonsmokers. In other, much better-known studies, such as that of the American Cancer Society Study (CPS II),3 the CHD risk of 1.8 for active female smokers was actually slightly below that reported by Kawachi et al for female passive smokers. Relative CHD risks of active smokers of a similar order of magnitude as shown in the CPS II study have been observed also in the Framingham Study4 and in the study on British doctors.5 It is difficult to believe that this discrepancy between the CHD risks as found in these major studies on active smokers and the Kawachi et al study can be explained, as Kawachi et al claim, by an increased CHD risk in the control groups because these studies included exposed and nonexposed nonsmokers, thus narrowing the gap between the CHD risk of smokers and nonsmokers. The credibility of Kawachi and colleagues’ hypothesis would improve substantially if the selection bias discussed above were to be excluded. The authors could do this by reporting the number of women with CHD they removed from each exposure group at each time point.
Asked for a possible conflict of interest, I declare categorically that I am not in any way, financially, economically, or otherwise, linked to the cigarette industry.
- Copyright © 1998 by American Heart Association
Kawachi I, Colditz GA, Speizer FE, Manson JE, Stampfer MJ, Willett WC, Hennekens CH. A prospective study of passive smoking and coronary heart disease. Circulation.. 1997;95:2374–2379.
Surgeon General Report. Reducing the Health Consequences of Smoking: 25 Years of Progress. US Department of Health and Human Services, 1989; Publication No. CDC: 89-8411.
Kannel WB, Higgins M. Smoking and hypertension as predictors of cardiovascular risk in population studies. J Hypertens. 1990;8:S3–S8.
Doll R, Peto R, Wheatley K, Gray R, Sutherland I. Mortality in relation to smoking: 40 years’ observation on male British doctors. BMJ.. 1994;309:901–911.