Tobacco Documents Online

I I I I I I I I I I I I I I I I I I Congressional Research Service • The Library of Congress • Washington,'D.C. 20540-7000 March 23, 1994 . TO FROM : Jane G. Gravelle Senior Specialist in Economic Policy -Oftlce of Senior Specialists and C. Stephen Redhead Analyst in Biomedical Science Science Policy Research Division SUBJECT : Discussion of Source of Claims of 50,000 Deaths from Passive Smoking This memorandum is in response to your request for information on the possible source of an estimated premature 50,000 deaths from passive smoking effects. This estimate is much larger than the (disputed) estimate of 3,000 premature deaths from lung cancer due to passive smoking effects that was reported in a recent study by the Environmental Protection Agency (hereafter EPA Report).' The 3,000 estimate is the only one mentioned in a recent article on the causes of death in the United States3 and lung cancer is the only passive smoking illness that is officially recognized by a government agency. The approximately 60,000 number was mentioned in testimony by the American Medical Association' which states that passive smoking 'may kill as many as 53,000' Americans annually. This.statement in turn appears to I Environmental Protection Agency. Bespiratory Health Effects of Passive Smoking: Lung Cancer and Other Disorders. December 1992. = J. Michael McGinnis and William F. Foege. Actual Causes of Death in the United States. Journal of the American Medical Association. November 10, 1993, pp. 2207-2212. s Statement of the American Medical Association, Health and the Environment Subcommittee, House Committee on Energy and Commerce, Re Adverse Health Effects of Exposure to Environmental Tobacco Smoke, July 21, 1993.

CRS-2 be ultimately traceable to an article published in 1988 in Environment International ` This article used existing epidemiological studies (statistical studies of incidence of disease among huutan populations) to estimate these deaths which included, under one set of calculations, 3,700 lung cancer deathi, 12,300 deaths from other cancers, and 37,400 deaths from heart disease. (Wells actually reported estimates ranging from 38,000 to 63,000, with a preferred estimate of 46,000) Each issue of Environm,ent Interreational contains an editorial; the one in the issue containing the Wells article was directed at that article. The editorial indicated that the study received mixed reviews from referees (two recommended publication after revision and the third recommended against publication on the grounds that it was too speculative), but the editors chose to publish the paper despite mixed reviews. In the following three years there were a series of critiques and rejoinders related to this paper j The main criticisms related to two points: that the evidence, particularly with respect to heart disease, was at odds with information on physical levels of exposure, and that there are a variety of serious problemo with epidemiological studies. The original article and correspondence are enclosed.s The following discussion elaborates on these issues. Generally, there are two ways that one might try to estimate the number of premature deaths, if any, from passive smoking. One could rely on estimates of physical absorption of the components of smoke and then use those estimates to extrapolate based on studies of the effects of active • A. Judson Wells, An Estimate of Adult Mortality in the Unitod States from Passive Smoking. Environment International. Vol. 14, No. 3, 1988, 249-265. s Letters from Alan W. Katzenatein, Peter M. Lee, and Larry Holcolm criticizing the Wells results; a clarifying letter from Takeehi Hirayama, a rebuttal to Katzenetein and I,ee from James L. Repace and Alfred H. Lowrey, and a response from Wells were published in 1990 (Vol. 16, no. 2, pp. 176- 193). In 1991, a letter from Stanton A. Glantz criticizing Lee was published along with Lee's reply (Vol. 17, no. 1, pp. 89-91). Later in 1991, a response of Les to the 1990 letters of Repace and Lowrey, and Wells, a letter from Muin J. Khoury clarifying a point raised in Lee's letter, a joint lotter from Glantz and Lee clarifying an issue raised earlier in the year, and a response from Ftepace and Lowrey, and Wells to Lee's letter were published (Vol. 17, no. 4, 379-387). In 1992 (Vol. 18, No. 3, pp. 315-317, 321-325) another letter from Lee and response from Wells was published. ' A more recent paper of the same general type as the Wells paper has been published on heart disease. See Kyle Steenland, Passive Smoking and the Risk of Heart Disease, Journal of the American Medical Association, Vol. 267, No. 1, January 1, 1992. I I I I I I I I I I I I I I i I I I

I I I I I I I CRS-3 smoking. These are called dosimotric approachea. Secondly, one might try to directly estimate the efTects of passive smoking by comparing disease rates of individuals who are and are not exposed to passive smoke. These studies are the epidemiological ones? ''i~ypically, they compare the rates of disease in nonsmoking women married to husbands who smoke and husbands who do not smoke. Obviously, one might be interested in the extent to which the estimates derived from these two approaches are similar, or are different. Consider the lung cancer studies first. Even though these premature deaths are only a small part of the total estimate in the Wells article, discussing these estimates will clarify some of the problems with the Wells estimate. The recent EPA study estimated a risk for lung cancer of about 30 percent from passive smoking based on epidemiological studies. That is, according to their analysis of the statistical studies, nonsmoking wives of men who smoker have 30 percent more lung cancer than nonsmoking wives of men who do not smoke. This risk is, in turn, only a tiny fraction of the risk from active smoking (probably around 3 percent),$ and the proportion of passive-smoking premature deaths to active smoking premature deaths estimated by the EPA is between two and three percent 9 A physical extrapolation approach would tend to yield smaller effects. According to data in the EPA report, measures of cotinine in the urine 7 Of course, the original estimates of the effects of active smoking on disease are based on epidemiological studies in large part as well, but there are some problems that occur in passive smoking epidemiological studies that are not as serious in active smoking studies. ' The risk of lung cancer in smokers and ex-smokers depends on intensity, duration, and, in the case of ex-smokers, time elapsed since quitting. Passive smoking would involve three percent of the risk of active smoking if there is a ten fold active-smoking risk (i.e. smokers have an additional estimated risk of lung cancer that is ten times the disease rate of nonsmokers) which is typical of current estimates of the risk for women as reported in the 1989 Surgeon General's Report (Reducing the Health Consequences of Smoking. U.S. Department of Health and Human Services, DDHS Publication No. (CDC) 89-8411). In generating the estimates of deaths from passive smoking, the EPA actually used the additional risk (of wives married to smokers as compared to wives marriedd to nonsmokers) in the U.S. passive smoking studies, which was about 20 percent. If studies from all countries are considered, the estimated risk from these studiee was 30 percent. ° The estimated attributable deaths from lung cancer due to active smoking are 113,000. See C. Stephen Redhead, Mortality and Economic Costa Attributable to Smoking and Alcohol Abuse, Congressional Research Service Report 93-SPR, April 20, 1993.

CKS-4 indicate that, overall, passive smokers have about 1/2 of one percent of the level of active smokers.10 Since the numbsr of current and former smoker. are the same as the number of never-smokers, the estimated premature deaths annually from passive smoking for never smokers would be about 600 using a linear extrapolation." This number is considerably less than the EPA's estimate of 2000 (the remaining 1000 estimate was for effects of environmental tobacco smoke on former smokers who are about half the number of never-smokers). There is also a section in the EPA study that discusses extrapolations based on the physical exposure to passive smokinE; these estimates also tend to be smaller than the epidemiological estimates and some are very low. There are potential problems with both methods. The physical extrapolation method used above assumes a linear relationship between the incidence of a disease and exposure. Based on evidence from the pattern for active smoking, however, a linear method may not be correct. The evidence, however, tends to suggest that such an adjustment would reduce the estimates based on physical exposure. There is some evidence that disease riees with square of the exposure or even with higher powers in the case of lung cancer.12 If the disease were to rise with the square of exposure, then the estimate based on cotia.ine levels would be only 3 people rather than 600 people. Of course, it is possible that the disease rises lest than proportionally with exposure.'s At the same time, it is possible that there is a threshold which is so small that individuals are not expected to experience health damage. It is also possible that cotinine is not the best measure of exposure; some exposure measures show larger and some show smaller effects. Problems also occur with epidemiological studies. It is always possible that relationships found with human population studies are due to chance, even in a perfectly designed study, and while there are statistical methods 10 EPA Report, pp. 3-43. " To extrapolate, multiply the percentage of cotinine by the ratio of ever- smoken to never smoksrs, and by the number of deaths attributable to active smoking. u Surgeon General's Report, p. 44. 13 The argument has been made for a relationship in which passive smoking can have large effects relative to active smoking for some specific events in laboratory settings, which is largely attributed to increased sensitivity of some nonsmokers. See Stanton Glantz and William Parmley, Passive Smoking and Heart Disease, in Circulation, vol. 53, no. 1, January 1991, pp. 1-12. Circutationn is a publication of the American Heart Association. I I I I I I I I I I I I I I I I

I I I I I I I I I I I I I I I I I I I f CRS-5 that assign probabilities of error, they are still probabilities. In addition, the precision of the specific estimate is always in question.". There are also some specific problems that have been identified with passive smoking studies. First, the measures of exposure are based on interviews with subjects or their relatives and are subject to considerable potential error. Secondly, the studies might be picking up the effects of active smoking, through misclassification of never-smokers as smokers. That is, some individuals who identify themselves as those who never smoked are actually current smokers or former smokers and they may be more likely to be married to smok~eri. Indeed, corrections were made for this effect in the EPA report, but it is difficult to know whether they are accurate. Finally, these studies do not or cannot fully control for 'confounders" - other factors that might be responsible for the effect that are simply correlated with marriage to a smoker. For example, smokers tend to be less concernedin general about health risks and engage in other behaviors (e.g. diet, lack of preventive health care) that might be shared with their spouses and that may be the cause of the health effects. The EPA chose the epidemiological studies as a basis of their approach, but they nevertheless relied on the cotinine measures for several aspects of their estimates (such as extrapolating from the effects on spouses of smokers to the population in general). While the estimates from at least some of the epidemiology studies are significantly larger than the estimates of physical exposure, these results are not magnitudes apart. The same cannot be said, however, for the Wells estimates of deaths from heart disease. Using the same type of linear physical extrapolation would result in 700 deaths from coronary disease for never smokers, and perhaps another 850 for former smokers, with a total of about a 1000. The portion of the Wells' 63,000 estimate from the epidemiological studies, even for several years ago, is 37,000, a number that is enormously larger. This large estimate occurs because the spidemioloeical studies show a very high risk estimate for passive smoking relative to active smoking for heart diseaae and not for lung cancer. For example, Wells indicates a 30 percent additional risk for heart disease for males and a 20 percent rise for females, as compared to a 70 percent risk for smokers. These relative risks are enormous compared to both the dosimetric ratios and to the epidemiological results for lung cancer. Note that although the risk ratios are not that different from lung cancer, the absolute risk estimates are much larger. The risk of lung cancer for nonsmokers is very low, and any percentage of a small number is still a small number. The risk for heart "' For a discussion of the methods used by the EPA in combining existing studies and the findings of studies published later, see Appendix A in Cigarette Tazes to Fund Health Care Reform, by Jane G. Gravelle and Dennis Zimmerman, Congressional Research Service Report 94-214, March 8, 1994.

CRS-6 disease is much larger initially, and therefore any siPnificant percentage change in the risk is larger. Put another way, even the epidemiological studies of lung cancer produced passive-smoking related deaths of less than 3 percent of active smoking related deaths, while the heart disease studies produced estimates that, were 26 percent of active smoking related deaths. The biological plausibility of passive smoking effects on cardiovascular diseases has been the subject of some discussion. Both Wells and Stoenland refer ts a 1991 review article by Glantz and Parmley that suggested that passive smoking may, in experimental studies, promote the formation of • plaques In blood vessels, increase the tendency of blood platelet cells to aggregate and form clots, and reduce the osygen-carrying capacity of the blood. 1i There arc limited data to support or refute these hypotheses. For example, while some studies of nonsmokers found that passive smoking appeared to promote platelet aggregation, parallel studies of active smokers have not consistently shown any effects on platelet function.is Among the chemical components in passive smoke, carbon monoxide and nicotine are t,he most likely to adversely affect cardiovascular performan,e. Carbon monoxide binds tightly to hemoglobin and diminishes oxygen transport in the blood stream. Nicotine acts in the brain and throughout the body, promoting the release of adrenalin and increasing heart rate and blood pressure. Although these effects might impair performance, exposures from passive smoke are generally thought to be at concentrations below those at which physiological changes would occur in healthy persons.17 The most likely explanation of these large risks from passive smoking epidemiological studies for heart disease is the absence of control for other factors.l• There are many important causes of heart disease (e.g. diet, lack of exercise, lack of preventive health care) that may be engaged in by smokers. That is, there is much evidence that smokers tend to be less ~ Glantz and Parmley (1991). " Samet, J.M., Environmental Tobacco Smoke, In Envfronmen.tal Tosicnnta, ed, M. Lippman, New York: Van Rostrand Reinhold, 1992. 11 U.S. Department of Health and Human Services, The Health Consequences of Involuntary Smoking, 1986, Surgeon General's Report, DHHS Publication Number (CDC) g7-8938. " This position is taken by Gary L Hubert, Robert E. Brockie, and Vuay K. Mahqjan in a paper written for the layman: Passive Smoking and Your Heart, Consumers Research, vol. 7b., April 1992, pp. 13-19, 32. These authors consider the results in the Wells and Steenland studies biologically implausible, and also note that six of the nine epidemiological studies show relative risks that are in excess of risks estimated for active smokers and that most have very few controls for other factors that might affect heart disease. I I I I I I I I I I I I I I I

I I I I I I I I I I I I I I I I I CRS-7 concerned about health risks in general. In general, studies do not, and perhaps cannot, control for many of thess factors. If smokers' vrives share in thQse behaviors, the relationships found in the epidemiological studies are spurious. Tbe Wells estimate of passive smoking deaths from cancers other than lung tancer is even larger relative to active smoking deaths than is the c,ase of heart disesse-about 60 percent. Again, these cancers are influenced by many other factors and the same general criticisms can be made about thess epidemiological estimates as in the case of heart disease. In sum, this analysis suggests that the WelIs estimates are so high relative to measures of physical ezpoeure that they ssem Implausible. It also suggests that the absence of controls or the inability to control for other factors may be a m$jor problem in relying on epidemiological estimates of the health effects of passive smoidng. I