Tobacco Documents Online

! Testimony on Passive Smoking Mortality A Review and Preliminary Risk Assessment by A. Judson Wells, Ph.D. Vice Chairman National Council f or Clean Indoor Air before the Natural Resources, Agriculture Research and Environment Subcommittee of the Committee on Science and Technology U. S. House of Representatives September 17, 1986

My name is A. Judson Wells. I am testifying today on behalf of the National Council f or Clean Indoor Air. The area I will cover is adult, nonsmoker mortality from passive smoking including a preliminary estimate of the probable number of deaths per,year in the U.S. from this cause. I have a Ph.D. in physical chemistry from Harvard University. My principal career has been in research in the chemical industry, having retired in 1980. For the past five years, on a part time basis, I have been studying the scientific literature on passive smoking with particular reference to mortality eff ects on adults. The material I am presenting today is taken from a paper I presented at the Annual Meeting of the Air Pollution Control Association in Minneapolis last June. A copy of the paper, updated slightly to clarify some of the tables, is appended to these remarks. As of last January, when the analysis for the Minneapolis paper was prepared, there were eighteen epidemiological studies that I knew of (I have since found one more) that bore on adult, nonsmoker mortality or cancer morbidity from passive smoking. Five of the eighteen studies were eliminated from the analysis because of (1) inadequate controls, (2) data on current exposure only, (3) insufficient data to calculate a statistical variance, or (4) a serious internal inconsistency. Of the remaining thirteen, eleven dealt with the disease categories of broadest interest, namely, lung cancer, cancers other than lung, ischemic -1-

heart disease, and emphysema. If the five papers that were rejected had been included, the results of the analysis would have been essentially unchanged. Also, although I have not yet carried out the mathematics, it is believed that inclusiop of the missing 1984 paper and papers that have appeared in 1986 or are about to appear on passive smoking mortality would not have a large effect on the results I will be presenting. To illustrate the methodology of the epidemiological analysis we will take female lung cancer as an example. Eight of the accepted studies had data on female lung cancer. The relative risks for each study were weighted according to the inverse of the statistical variance for that study. Then the case control data and the cohort study data were each combined separately, and then an overall combination was made. This method was developed at the Harvard School of Public Health. By this procedure the relative risks shown in Table I were obtained f or each sex and each disease category. The f emale "other cancer" risk is based on f our studies, the heart disease on three, and the emphysema on one. For the males the lung cancer risk is based on three studies, the other cancer on two and the heart disease on two. Obviously, the male data are much scarcer than the female data. For those of you interested in statistics, the p value is the probability that the result is unity, and the confidence interval is the range that the true value falls in with 95% probability. -2-

To develop an estimate of annual deaths from these relative risks it is necessary to have inf ormation on the total nonsmoking population, the percent exposed to sidestream smoke, and the death rates for nonsmokers for each disease category. Estimates ., for these parameters from appropriate sources are shown in Table II along with the relative risks from Table I and the resultant calculated annual deaths by sex and disease category. The estimate for lung cancer, 1800 deaths per year, is right in the middle of the range estimated by Repace and Lowrey of 500 to 5000. However, as can be seen, lung cancer may be just the tip of the iceberg. Other cancer and heart disease, with their much higher underlying nonsmoker death rates and relative risks in the same general range as lung cancer, provide the great bulk of the underlying causes of death. It must be emphasized that these results are preliminary, but if the epidemiology continues the trend it has taken since 1981, these numbers represent the likely outcome. Another factor that appears to be coming forth from the epidemiology is that the cancer sites for cancers other than lung are diff erent f or passive smoking and direct smoking. Based on the few studies that we have so far on specific sites from passive smoking we get the patterns shown in Table III. The be entry sites appear tonappropriately different, and direct smoking leads to digestive tract cancers, or close relatives, while passive smoking does not. This difference in sites is thought to arise from a difference in eff ective particle size between -3-

mainstream and sidestream smoke. Mainstream smoke is moist and concentrated and appears to agglomerate into large particles that deposit mainly in the mouth or in the main airways of the`lung. From these positions it is cleared into the mouth and swallowed, .~ apparently resulting in the cancers shown.' In passive smoking, however, the particles are small and dilute and tend to stay that way. They are breathed through the nose and tend to be deposited either in the nose or deep in the alveolar region of the lung. Clearance from deep in the lung is very slow which may result in partial solution of the particles and clearance into the blood or lymph systems, hence the absence of digestive system cancers and the presence of cancers that may be more blood and lymph related. Basically, environmental tobacco smoke is a carcinogenic air pollutant. It has been known f or years that cigarette-smoke is carcinogenic and that it causes heart attacks. Millions of ( ovQr age yr) people are exposed (I have noted 32 million,\in Table II). Such a material, as Bernard Goldstein has said, is guilty until proven innocent. Individual victims of passive smoking, like victims of the af termath of Chernobyl, will not be identifiable as such. Their numbers will be buried in the approximately 1.5 million total heart and cancer deaths in the U.S. each year. Yet 47,000 preventable deaths is a large number. So is 1800, compared to the 20 to 100 deaths per year that would attract regulatory attention to any other carcinogenic air pollutant. The results, although preliminary, indicate a strong possibility for substantial risk and call f or an appropriate government response. -4-

TABLES for testimony of Dr. A. Judson Wells September 17; 1986

TABLE I COMBINED RISK RATIOS FOR PASSIVE SMOKING Disease Female Exposed * 95% Cases R/R 2-tail p conf. ant. Lung cancer 463 1.44 <0.001 1.2 - 1.7 Other cancer 2091 1.56 43.001 1.3 - 1.9 Ischemic heart dis. 276 1.27 0.04 1.0 - 1.6 Emphysema & chr. br. 102 1.4 0.18 0.9 - 2.1 Male Lung cancer 13 2.5 0.009 1.3 - 4.7 Other cancer 7 1.0 Ischemic heart dis. 14 1.3 0.46 0.7 - 2.6 *Risk ratio or relative risk

TABLE II PRELIMINARY RISK ASSESSMENT FOR PASSIVE SMOKING PROBABLE NUMBER OF U.S. DEATHS PER YEAR Non-smoking Population, 45+ % Exposed Passive smoking Population, 45+ Disease Lung Cancer Other Cancer Ischemic Heart Emphysema Total - four Diseases aRelative risk. Females Males Both 27.6 MM 20.3 MM 47.9 76% 55% 67% 21.0 MM Non-smoker R/Ra Death Rateb 1.44 9 1.56 245 1.27 210 1.4 2 bPer 100,000 person years. Deaths from Pass. Smkg. 600 20,200 9,900 100 30,800 11.2 MM 32.2 Non-smoker Deaths from Total R/Ra Death Rateb Pass. Smkg. Deaths 2.5 13 1,200 1,800 1.0 188 0 20,200 1.3 531 15,300 25,200 1.0 6 0 100 16,500 47,300 ZOOLSLZ6

TABLE III CANCER SITE PATTERNS IN DIRECT AND PASSIVE SMOKING Direct Smoking Passive Smoking Buccal Cavity - - Nasal Sinus Pharynx - Larynx - Lung - largely Lung - largely bronchial peripheral Esophagus - Stomach - Urinary Bladder - Kidney - Pancreas - - Breast - Brain - Endocrine Glands Cervix Cervix

RELATIVE RISKS: LUNG CANCER FROM PASSIVE SMOKING Highest Exposure All Exposures Mantel_Trend Locale Cases R/R 2-Tail p R/R 95% C.L. 1-Tail p Females Cohort Studies: Hirayama Japan 163 1.9 0.002 1.6 1.1 - 2.2 0.002 Garfinkel U. S. 88 1.1 - 1.2 0.8-1.6 - Gillis, et al. Scotland 6 - - 1.1 0.2- 5.6 - Combined Cohort 257 1.34 1.1 - 1.7 Case Control: Trichopoulos, et al. Greece 53 2.6 0.19 2.1 1.2 - 3.6 0.005 Correa, et al. Louisiana 14 3.5 0.017 2.1 0.8 - 5.2 - Koo, et al. Hong Kong 64 - - 1.3 0.7-2.3 - Sandler, et al. N. Carolina 2 - - inf. - - Garfinkel, et al. U. S. 73 2.0 0.05 1.3 0.8 -1.9 0.025 Combined C/C 206 1.54 1.2 - 2.0 Combined Cohort & C/C 463 1.44 1.2 - 1.7 Males Cohort Studies: Hirayama Japan 7 2.3 0.16 2.2 1.1 - 4.8 0.023 et al. Scotland Gillis 4 - - 3.3 0.7-16.4 - , Combined Cohort 11 2.5 1.2 - 5.2 Case Control: et al. Louisiana Correa 2 - - 2.0 0.4-10.0 , Combined Cohort & C/C 13 2.5 1.3 - 4.7 VOqLSLZS