Document DGYwoQy0q8MzjnVE77aaVoXko

PLAINTIFF'S EXHIBIT Mortality Effects of Cigarette Smoking Among Amosite Asbestos Factory Workers12 Irving J. Selikoff, M.D., * Herbert Seldman, 4 and E. Cuyler Hammond, Sc.D. 4 5 ABSTRACT--Nine hundred and thirty-three amosite asbestos accumulating data to compare with our original obser factory workers began work from June 1941 through December vations, but we believe that this method has made the 1945. At 20 years from onset of employment (1961-65), 582 were comparisons more secure. known to be alive and to have had asbestos work experience We have been conscious of the fact that our findings solely at the factory. Smoking habits were ascertained at 20 years have been derived from one category of asbestos-exposed from onset of employment, and observation was then maintained individuals--asbestos insulation workers. Whereas it prospectively through 1977, 304 deaths occurred, approximately has been clearly established that these individuals have twice the number expected. The excess mortality was largely due very significantly increased risks of death from asbestos- to malignant neoplasms. 116 neoplasms as compared to 33 associated diseases (<, 5), there remained the possibility expected. Sixty cancer deaths were from lung cancer and 14 that other factors in their work environment could more were due to mesothelioma. Nevertheless, had it not been have interacted with their cigarette smoking expxasure for cigarette smoking, many of the excess deaths would have and that some or all of their increased lung cancel risk been avoided Asbestos alone would have accounted for much of could be so attributed. It has therefore been of interest the increase, including some excess deaths from lung cancer, but to us to investigate the mortality experience of another not for most of the deaths (60) from this neoplasm. With the use group of asbestos-cxpvosed workers who did no insula of smoking-specific mortality retes. It was calculated that the tion work and who were not employed in the construc combination of cigarette smoking and asbestos exposure in tion industry. creased the risk of lung cancer death about BO times. Thus the We now repvort results from this study. We found that experience of the amosite asbestos factory workers agreed with factory workers employed in a plant making amosite that of insulation workers and provided evidence that the asbestos asbestos products for shipyards and other industrial smoking interaction is specific for asbestos rather than other facilities had lung cancer mortality experience consis possible influences in industrial environments.--JNCI 65: SOT- tent with that of the insulation workers. Asbestos SIS. 19B0. exposure alone increased the risk of lung cancer in the absence of cigarette smoking, but foi reasons given, the It was shown in 1967 that asbestos exposure sharply increased the already high risk of death from lung ranrer associated with cigarette smoking among asbestos total number of such cancers was not great. On the other hand, it appears that the remarkable increase in risk among cigarette-smoking asbestos workers is due insulation workers (7). The limited data on which these observations were based were later extended and confirmed by much larger studies, which demonstrated Abbrivistions used ACS = American Cancer Society. BL = be-: evi that between 1 in 4 and 1 in 5 deaths among these dencr available: DC = death cenificaie information onls workers was due to lung cancer if they smoked cigarettes. In addition, as compared to other non- smokers, nonsmoking insulation workers also had an 1 Received August 24. 1979; revised February 29. 19*;'. accepted increased risk of lung cancer, but inasmuch as risk March 17. 1980 among nonsmokers has been relatively low, even multi plying this risk severalfold did not result in many lung cancers (2, 3). Cancer of the esophagus was also considerably increased among smoking insulation workers but not among their nonsmoking co-workers, 1 Supported in pari by Public Health Service grant ESOOfCk Irom the National Institute of Environmental Health Sciences and bv grant R-5S from the American Cancer Society. 5 Environmental Sciences Laboratory. Mount Siru; School ol Medicine ol the City University of New York. 10 Eas: I02d Sc. New York. N.Y. 10029 i However, pleural mesothelioma, peritoneal mesotheli- 4 Department of Epidemiology and Statistics. .American Cancel oma. cancer of the stomach, and cancer of the colon Society, 777 Third Ave.. New York, N.Y. 1001T t and rectum occurred with approximately equal excess ' We thank Mr. Ashley Bodden. Mr. David Newman. M: Edwin i among both smokers and nonsmokers (3). Silverberg. and Ms. Marie Corbo lor their assistance in processing the t These studies w'ere detailed and prolonged because data; Ms. Dorothy Perron. Ms. Shirley Levine. Ms Ravla Margolei. we preferred to utilize only prospective observations made after establishing the smoking history from the and Mr Ovaries V. Nolan for the field investigations of the men in this study; Ms Frances Perez for management of specimens and Xrays. and Ms. Selma Annenberg. Mr. Sidney Sibel. and Ms Judy individual concerned, with subsequent observation, Marmor lor preparation of the manuscript. W'c acknowledge thr help and to compare observed deaths with those expected of the pathologists, clinicians, hospitals, coroners. State health among men of the same age, in the same years, with departments, and other medical facilities providing detailed infuema the same distribution of smoking habits, also observed lion, records, slides, blocks, and roentgenograms We also thank Dt prospectively. This situation led to some delay in J Oturg. Dt. Y. Suzuki, and Di M. Kannrrstetr. foe then help A02872 507 JNCI. VOL. 65. NO S. SIP FEME! tt 1<*V' UCC 015495 508 StUkoff, Seldman, and Hammond to the exposure to the mineral dust rather than other (actors, multiplying the already very high risk of cigarette smoking. MATERIALS AND METHODS In 1941, with the encouragement of the U.S. Navy, a factory was established in Paterson, New Jersey, to manufacture asbestos products for the armed forces and for industrial uses. We have ascertained that amosite asbestos was used almost exclusively (with very small amounts of chrysotile asbestos also being used). Our analyses of fiber retrieved from the factory's storerooms, fillers from workers' dust masks, samples of the prod ucts made, and fibers extracted from workers' lungs at autopsy have confirmed this almost exclusive use of amosite asbestos and are consistent with the fact that the product specification called for such fiber, that the machinery was designed specifically for this purpose, and that the company that supplied the asbestos from its South African mine had provided only this fiber variety (d). This specificity is of some interest because it provides information concerning one of the three major fiber types. Insulation workers had been exposed only to chrysotile asbestos at first and later to both chrysotile and amosite (5). In 1941-45, 933 men began work in this plant. Some worked for a month or less and others until the plant closed in 1954. We have sought to trace the entire cohort of 933 men and to keep them under surveil lance. We have already found that this group has, overall, a significantly increased risk of death from asbestos-associated disease (7). Other studies have shown that the risk of death from neoplasms known to be associated with asbestos expo sure becomes considerable only after 20 years from onset of exposure to asbestos. We therefore believed it useful to limit our analyses to the experience of the men who worked in this plant, starting with their 20th year from onset of employment and excluding all experience subsequent to asbestos employment else where. For some men (those who began work in 1941) the experience was from 1961 onward. For others, observation for current purposes did not begin until 1965 (i.e., men who began work in 1945). Of the original cohort of 933 men, 351 did not qualify for this analysis. Of these 351 men, 39 had prior asbestos work employment elsewhere, 270 were dead, and 42 were lost to follow-up. Almost all of those lost to follow-up were lost immediately after termination of their factory employment. Actually, this cohort exhibited decidedly unfavorable mortality even before the 20-year point (8, 9). Table 1 demonstrates that 582 men came under surveillance at the 20-year point and that the ex perience for each man from that time to December 31. 1977, included 6.311 man-years of observation. Table 1 also gives the age distribution of these men. Most of the men in this factory had a history of cigarette smoking. Table 2 outlines the smoking habits of the 582 men we have studied; 430 (73.9%) had a Table 1.--Number of men at 20-year point and man-years of observation, 1981-77. 20 or more years from onset of employment. 191,1-1,5. for 582 amosite asbestos factory workers by age Age group, yr 35-39 49-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Total No. of men 84 99 90 86 65 56 40 35 20 6 2 582 Man-yr 150 663 1.062 1.192 1.037 787 560 406 266 126 62 6.311 history of regular cigarette smoking (and almost 85% of these were still smoking at onset of observation), whereas 137 men had no such history. Of these 137 men, 59 (10.1%) had smoked pipes and/or cigars but no cigarettes, whereas only 78 (13.4%) had no history of any tobacco smoking. Smoking histories were unob tainable for 15 men (2.6%). The entire cohort has been under surveillance since 1961, with the majority undergoing periodic clinical examination. For men (or whom this periodic exami nation was not possible, intermittent contact was achieved. Whenever a death occurred, all available clinical and pathologic material was sought and re viewed, including reports of personal medical attend ants, hospital records, and histologic material ob tained at operation and/or autopsy. For each death a certificate of death was obtained. Each individual in the cohort remained in the analysis "at risk" from the point at which he reached 20 years from onset of employment during the period 1961-65 until he took up other asbestos employment or otherwise until December 31, 1977, if he was still alive or until date of death if he died before then. The group as a whole then included 6,311 man-years of observa tion (table 1): 4,811 man-years for individuals with a history of cigarette smoking, 868 for those who never smoked regularly, 528 for those who had smoked pipes and/or cigars, and 104 for those whose smoking history was unknown (table 3). Using the attained age of each man in each calendar year period after admission to the cohort following the Table 2.--Number of men at 20-year point and man-years of obsenation. 1961-77, 20 or more years from onset of employment, for 582 amosite asbestos factory ivorkers by smoking habits at 20-year point Smoking history History of cigarette smoking Current smokers Exsmokers No history of cigarette smoking Never smoked regularly Pipe and/or cigar only Smoking history unknown No. of men 430 359 71 137 78 59 15 Man-yr 4.811 3,993 818 1.396 868 528 104 JNCI. VOL., 65, SO S. SEPTEMBER 1980 UCC 015496 A02873 Mortality. Smoking and Aabetlot Factory Work 509 Table 3.--Number of mm at 20-year point and man-years of obsermtion. 1961-77. 20 or more years from onset of exposure. 191,1-16. for 582 amosite asbestos factory workers by smoking habits at 20-year point and by age Age group, yr 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Total History of cigarette smoking No. of men Man-yr 67 121 80 529 70 837 70 942 52 814 38 630 25 433 19 287 1 148 1 49 1 21 430 4.811 Never smoked regularly No. of men Man-yr 10 19 10 80 15 140 8 166 6 132 8 93 9 69 7 67 4 56 1 37 09 78 868 Pipe BTid/or cigar only No. of men Man-yr 36 9 39 3 64 6 67 3 69 8 47 6 51 9 52 8 62 3 39 1 32 59 528 Smoking historyunknown No. of men Man-yr 44 0 15 2 21 2 17 4 22 1 17 07 00 10 11 00 15 104 20-year-from-onset point. 1961-77. we computed the number ol expected deaths tor this cohort tor these periods of years. Calculations were made in several ways. First, ex pected deaths were based on age-specific data of the I'.S. National Center for Health Statistics for New Jersey white males (1963-66, 1967-71, and 1972-77) without regard to smoking habits. Expected deaths were also computed in relation to smoking habits on the basis of age- and smoking-specific experience (1963-66 and 1967-71) of white males in the ACS prospective study (JO): this analysis was limited to the experience of men other than farmers with at most a high school education and a history of occupational exposure to dust, fumes, chemicals, gases, or radiation. The 1967-71 ACS study rates were extrapolated to 1972-77 according to changes in age-specific death rates among total I'.S. white males between 1967-71 and 1972-77. Expected deaths for the few men with unknown smoking habits were computed on the basis of the age-specific and calendar-year-specific rates of all the men (regardless of smoking history) in the ACS study. A third set of expected deaths was computed on the basis of the death rates of the amosite asbestos factory workers who never smoked regularly. To ensure more stable figures, we initially used age-specific death rates for all of the men with experience past 20 years from onset of employment (regardless of smoking habits), 1961-66, 1967-71, and 1972-77. The resulting expected deaths were then proportionately adjusted to levels such that totals of expected and observed deaths were equal for the workers who never smoked regularis. Observed deaths were then compared with deaths expected. We classified the causes of death in the observed deaths in two ways, according to DC and according to BE. Some investigators may prefer to use the DC figures and argue that the deaths in the populations used as controls have not been subject to a rrview similar to that of the BE deaths. Their prefer ence assumes that whatever errors exist in the DC figures are much the same in the population being investigated (factory workers) as in the control popula tions (ACS or ITS. National Centet for Health Statis tics data). This procedure is certainly far less laborious. However, some serious difficulties arise. When the distributions of causes of death in the population being investigated and control groups are markedh different, one might anticipate the distributions of error to be different. Mesothelioma and asbestosis have been decidedly underreported on the death certificates of asbestos workers. However, because of the infre quency of these conditions, they have not posed corre sponding difficulties in the general population. Whatever one does, one cannot ensure strict compar ability between the cause-of-death classifications for the population investigated and the control populations. In the interest of showing as accurate results as we can for the population being studied, we have oriented the exposition in this paper to the BE findings. However, the DC results have also been shown in the tables for those who prefer such data. Paralleling the observations for insulation workers (3, 11). some differences were observed among the various categories between cause of death as recorded on the death certificate and cause of death established after review (table 3). However, on the whole, this set of findings speaks well for the diagnostic acumen of the medical attendants of these men. RESULTS Three hundred and fouT deaths occurred among the 582 men observed sometime between 1961 and 1965 (the point at which each man reached 20 yr from onset of employment) to December 31, 1977. Cancer of one or another site accounted for 116 of these deaths (38.2%), and 18 deaths were due to asbestosis (5.9%). Table 5 shows that the following cancers in five areas caused most of the deaths: Jung cancer (60 deaths); pleural mesothelioma (7 deaths); peritoneal mesothelioma (7 deaths); gastrointestinal cancer including esophagus (16 deaths); and cancers of the oral cavity, pharyngeal, and laryngeal tissues (5 deaths). Altogether, there were approximately three and one-half times as many cancer deaths as expected. The increase in cancer deaths, A02874 JNCI. VOL 6S. NO J St.PTLMBI R lew- UCC 015497 510 S*Nkof1, Seldman, and Hammond Table 4.--Categorization of SOi deaths. 1961-77. 20 or man yean from onset of employment 1911-15. for 58t amotile asbestos factory icorlcen Cause of death ` Based on DC only Based on BE Difference between BE and DC All causes Cancer, all sites Lung Pleural mesothelioma Peritoneal mesothelioma Mesothelioma not spec- ified above Larynx, oral cavity. and pharynx Esophagus Kidney Colon-rectum Stomach Prostate gland Bladder Pancreas Other and unspecified Noninfectious pulmonary diseases, total Asbestosis Cardiovascular diseases Other and unspecified causes 304 103 52 1 0 3 3 1 2 11 4 4 2 5 15 29 8 130 42 304 116 60 7 7 0 5 1 2 11 4 4 2 4 9 24 18 122 42 0 +13 +8 +6 +7 -3 +2 0 0 0 0 0 0 -1 -6 -5 +10 -8 0 coupled with deaths due to asbestosis. serves to explain most of the virtual doubling of the expected deaths for these men over the period under study. In contrast to the insulation workers, the factory workers had an excess, though a limited one, in mortality from cardio vascular diseases (3, II). Data in tables 6 and 7 are organired to shed some light on the excess deaths associated with amosite asbestos exposure, tobacco habits, and amosite asbestos plus tobacco exposures. Overall mortality in nonsmok ing asbestos factory workers was more than doubled as compared with the experience of comparable men who had no asbestos exposure (table 6). This observation shows the strong influence of amosite asbestos expo sure in itself. For the total cohort of asbestos factory workers,, if expected deaths had been calculated on the basis of the experience of asbestos factory workers who never smoked regularly, it could be estimated that 244 deaths would have occurred, whereas 91 deaths would have been anticipated for men who had neither a history of tobacco smoking nor asbestos employment (table 6). The influence of asbestos alone can be seen as the difference, or 153 deaths. Similarly, if expected deaths were based on men with the same smoking habits, but not with a history of asbestos employment, 137 deaths would have occurred instead of the 91 (table 6). The influence of smoking alone would thus account for'46 deaths. The remainder of 14 when 91 + 153+46 is subtracted from the observed total of 304 deaths can be regarded either as the joint effect of asbestos and smoking or as discrepancies in data that do not purport to be absolutely consistent. The 304 deaths JNC1. VOL 65. NO. 3. SEPTEMBER 1SSO then are distributed as 30% neither asbestos related nor smoking related, 50% related to asbestos alone, 15% related to smoking alone, and 5% remaining. Performing the same computations for the asbestos factory workers who were cigarette smokers, we esti mated that of the 214 deaths (table 6), 57 (27%) were neither asbestos nor smoking related. 115 (54%) were related to asbestos alone, 46 (21%) were related to smoking alone, and --4 (--2%) were the remainder. Although these data can hardly be regarded as definitive, they do illustrate the general nature of the composite effects evident for all causes of death. The composite effects include, among others, specific smok ing effects, asbestos effects--pleural mesothelioma, per itoneal mesothelioma, asbestosis, and the excess deaths from cancers of the stomach, colon-rectum, oral cavity, pharynx, and larynx as well as the limited number of excess deaths from lung cancer among the nonsmoking Table 5.--Observed and ezpected deaths. 1961-77. >0 or more years from onset of employment. 19il~i5. among 582 amosite asbestos factory uorkers by cause of death Cause of death Expected deaths deaths Smoking DC BE New Jersey* adjusted: ACS study* All causes 304 304 Cancer, all sites 103 116 Lung 52 60 Pleural mesothelioma 17 Peritoneal mesothe- 07 tioma Mesothelioma not 30 specified above Larynx, oral cavity. 35 and pharynx Esophagus Kidney 11 22 Colon- rectum 11 11 Stomach 44 Prostate gland 44 Bladder 22 Pancreas 54 Other and unspecified 15 9 Noninfectious pulmonary 29 24 diseases, total Asbestosis 8 18 Cardiovascular diseases 130 122 Other and unspecified 42 42 causes 158 6 334 10.1 Fn c Fn c Fn c 1.6 0.8 0.8 52 2.0 2.7 1.4 1.8 7.0 4.7 Fn c 95.3 252 137.1 29.8 10.1 Fn c Fn c Fn c 1.0 0.6 0.9 37 1.4 25 0.9 1.8 6.9 7.4 Fn c 81.2 18.7 " Expected deaths were computed on the basis of age-specific data of the U S. National Center for Health Statistics for New Jersey 'white males (1963-66, 1967-71, and 1972-77) without re gard for smoking habit3 * Expected deaths were computed on the basis of age- and smoking-specific experience (1963-66 and 1967-71) of white males in the ACS prospective study. This analysis involved men with at most a high school education and with a history of occupational exposure to dust, fumes, chemicals, gases, or radiation; farmers were excluded from this group. The 1967-71 ACS study rates were extrapolated to 1972-77 according to changes in age-spe cific death rates in total U S. white males between 1967-71 and 1972-77. ' New Jersey and ACS study death rates are not available, but these diseases have been rare causes of death in the general popu lation. A0287b UCC 015498 Mortality: Smoking and Aibeatoa Factory Work 511 Table 6.--Observed and expected total death* 1961-77, SO or more yean from onset of employment, 19hl-i5. among 582 amosite asbestos factory workers by smoking habits at 20-year point Expected deaths based on Smoking histofy History of cigarette smoking Never smoked regularly Pipe and/or cigar only Smoking history unknown Total Observed deaths 214 38 41 11 304 Amosite asbestos factory workers who never smoked regularly* 171.9 38.0 307 3.0 243.6 ACS study Men with matching history of smoking1 Men who never smoked regularly' 102.3 15.5 18.1 12 137.1 56.7 15.5 17.7 0.8 90.7 * Expected deaths were initially computed on the basis of age-specific experience (1961-66. 1967-71. and 1972-77) of all men in the cohort 20 or more yr from onset of exposure. The expected death totals were then adjusted proportionately so that die expected death total for the men who never smoked regularly equaled their observed death total. * Expected deaths were computed on the basis of age- and smoking-specific experience (1963-66 and 1967-71) of white males in the ACS prospective study. This analysis was limited to men with at most a high school education and with a history of occupational exposure to dust, fumes, chemicals, gases, or radiation; farmers were excluded from this group The rates in the ACS study (1967-71) were extrapolated to 1972-77 according to changes in age-specific death rates in total U S. white males between 1967-71 and 1972-77. ' As in footnote 6. except that expected deaths were computed on the basis of the experience of the men who had never smoked regularly. asbestos workers--and the very much larger increase in the lung cancer category as a result of the interaction between asbestos exposure and cigarette smoking. The last phenomenon, the remarkable multiplication of the already high lung cancer risk of cigarette smoking that occurs in asbestos workers, is clearly seen in table 7. The 0.2 deaths were expected to occur among the nonsmoking asbestos workers if their ex perience had been the same as comparable nonsmokers who had no history of asbestos work. However, 3 such deaths (BE classification) occurred among the 78 men in the group who had no history of regular cigarette smoking. This observation suggests that asbestos expo sure alone produces an increased risk of lung cancer and agrees with observations in our cohorts of asbestos insulation workers (3). In terms of total number of lung cancer deaths, the increase is limited although undesirable. However, the increase in number of lung cancer deaths among cigarette smokers is devastating. Among 430 men with a history of cigarette smoking were 55 men who died from lung cancer (BE), approx imately six limes the number of other cigarette smokers without a history of asbestos employment and approx imately 80 times the number expected among men of Table 7.--Observed and expected lung cancer deaths, 1961-77. SO or more years from onset of employment. iSil-iS, among 58? amosite asbestos factory workers by smoking habits at 20-year point omoKing nisvory xation History of cigarette smoking Never smoked regularly Pipe and/or cigar only Smoking history unknown Total DC BE DC BE DC BE DC BE DC BE deaths 45 55 5 3 2 2 0 0 52 60 Expected death; based on Amosite asbestos factory workers who never smoked regularly* ACS study Men with matching history of smoking* Men who never smoked regularly' 28.8 16.5 50 3.0 3.0 1.8 05 0.3 37.3 21.6 96 02 0.2 0.1 10.1 0.7 0.2 0.2 00 1.1 * Expected deaths were initially computed on the basis of the age-specific DC and BE experience (1961-66. 1967-71. and 1972-77) of all men in the cohort 20 or more yr from onset of exposure. The expected death totals (DC and BE) were then adjusted proportionately so^that the expected death totals for the men who never smoked regularly equaled their observed death totals (DC and BE) Expected deaths were computed on the basis of age- and smoking-specific experience (1963-66 and 1967-71) of white males in the ACS prospective study. This analysis was limited to men with at most a high school education and with a history of occupational ex posure to dust, fumes, chemicals, gases, or radiation; farmers were excluded from this group. The rates in the ACS study (1967-71) were extrapolated to 1972-77 according to changes in age-specific death rates in total U S white males between 1967-71 and 1972-77. As in footnote 6, except that expected deaths were computed on the basis of the experience of the men who had never smoked regularly. A 0 2 8 71j UCC 015499 JNCI VOL & NO J. SEPTEMBER I9W 512 S*ilkoff, Seldman, and Hammond the same age and same personal background but without a history of either tobacco smoking or asbestos employment (0.7 expected and 55 observed). For other asbestos-related diseases, considerable in terest exists in whether the risks of death associated with asbestos exposure are also influenced by cigarette smoking. However, there are only a small number of deaths at hand for use in investigation of these diseases among the amosite asbestos factory workers, especially among the nonsmoking workers. Table 8 shows the data available for mesothelioma, gastrointestinal cancer, and asbestosis deaths among the factory workers (cigarette smokers and nonsmokers). To supplement these insubstantial findings, table 8 also gives corresponding data for the asbestos insula tion workers (Hammond EC, Selikoff IJ, Seidman H: Unpublished data.) The BE classification of causes of death is used because all comparisons in table 8 are within the asbestos worker groups. Moreover, the DC classification does not provide results of any value with respect to the incidence of deaths from mesothelioma and asbestosis. We found no increased risk of death from mesothe lioma in amosite asbestos factory workers with a history of cigarette smoking compared with nonsmokers. The same was true for cancer of the stomach, colon, and rectum. These observations agree with those for the asbestos insulation workers. However, the factory workers do not show the strong influence of cigarette smoking on risk of death from asbestosis that was so striking in insulation workers (table 8). CONCLUSIONS The mortality experiences of amosite asbestos factory workers are much the same as those of asbestos-exposed insulation workers, i.e., remarkably increased risk of deaths from asbestosis and cancers of several sites (lung cancer; pleural mesothelioma; peritoneal mesothelioma; gastrointestinal cancer; and oral cavity, pharyngeal, and laryngeal cancers) (11, 12). Among men 20 or more years from onset of asbestos exposure, observed pro spectively from a 20-year point reached 1961-65 to December 31, 1977, 304 deaths were seen, approximately twice the number expected. Eighteen deaths from asbestosis occurred. Excess deaths, however, were largely due to neoplasms. Neoplasms caused 116 deaths, three and one-half times the number expected. Sixty cancer deaths were from bronchogenic carcinoma and 14 more were due to either pleural or peritoneal mesothelioma. Nevertheless, had it not been for cigarette smoking, many of the excess deaths would have been avoided. Asbestos alone would have accounted for an important increase, including some excess deaths from lung cancer, but not for most of the lung cancer deaths. Here asbestos exposure greatly multiplied the already high risk that would have been present with cigarette smoking alone. The combination of cigarette smoking and asbestos exposure among these amosite asbestos factory workers was calculated to have increased their risk of lung cancer death about 80 times as com pared with like men of the same age who neither smoked cigarettes nor worked with asbestos. This increased risk is very much the same as that seen among asbestos insulation workers. This observation indicates that the increased risk of death from lung cancer among cigarette-smoking asbestos workers is a specific interaction rather than coincidental, and not, for example, the result of other agents in the environ ment of the construction trades. The factory in which these men worked closed in 1954, and no further occupational exposure to asbestos occurred for most of the men. For most workers, asbestos employment ceased in 1945 at or before the end of World War II. In other studies (8, 9) we found that cessation of occupational exposure to asbestos in this factory was not accompanied by cessation of risk of asbestos-associated disease, including mesothelioma. Table 8.--Obaerted and expected deatha from selected causes (BE) 20 or more years from onset of employment for 582 amosite asbestos factory workers. 1961-T7. and 12.051 asbestos insulation workers. 1957-76. for men with a history of smoking cigarettes compared with those who never smoked regularly Causes of death Cigarettes or NSR' Amosite asbestos factory workers Observed deaths, BE Expected deaths. NSR* Asbestos insulation workers Observed deaths, BE Expected deaths. NSR' Mesothelioma Cancer of stomach, colon-rectum Asbestosis Cigarettes NSR Cigarettes NSR Cigarettes NSR 9 3 9 3 11 2 14.8 3.0 10.1 3.0 10.0 2.0 105 108.8 15 16.0 43 41.5 6 6.0 97 328 5 5.0 * NSR = never smoked regularly. Expected deaths were initially computed on the basis of the age-specific BE experience (1961-66, 1967-71, and 1972-77) of all amosite asbestos factory workers 20 or more yr from onset of exposure. The expected death totals were then adjusted proportionately so that the expected death totals for the men who never smoked regularly equaled their observed death totals. c Expected deaths were initially computed on the basis of the age-specific BE experience (1967-71 and 1972-76) of all asbestos insu lation workers 20 or more yr from onset of exposure. The expected death totals were then adjusted proportionately so that the expected death totals for the men who never smoked regularly equaled their observed death totals. JNCI. VOL. 65. NO. J. SEPTEMBER 1980 A02677 UCC 015500 Mortality: Smoking and Aabestos Factory Work 513 gastrointestinal cancer, and asbesiosis. This finding makes it all the more necessary to repeal an admoni tion previously made (/): Asbestos workers who smoke cigarettes should slop smoking immediatelv and those who do not smoke should never begin. Furthermore, all avoidable exposure to asbestos must be prevented. REFERENCES Ut Selixoff IJ. Hammond EC. Chirg J Asbestos exposure, smok ing and neoplasia. JAMA 1968.204 106-112. ,?) Hammond EC. Seuxoff IJ Relation o! cigarette smoking to risk of death of asbestos-associated disease among insulation workers in the United States In: Bogovski P, Gilson JC. Tim brel! V. Wagner JC. eds. Biological effects of asbestos IARC Sci Publ 1973:8 512-317. {5. Hammond EC. SfcLtvOFF IJ. Seidman H. Asbestos exposure, cig arette smoking and death rates .Ann NY Acad Set 1979,330. 473-490. {4) Seuxoff IJ, Chvrc J, Hammond EC. Asbestos exposure and neoplasia. JAMA 1964.188:22-26 (S) Seuxoff IJ. Hammond EC. Qti'RG J Mortalit> experiences of asbestos insulation workers. In: Shapiro HA. ed. Pneumo coniosis. Proceedings of the'international conference. Johan nesburg. Capetown. Oxford Unix Press. 1970180-186 (6) Lancer AM, Rubin 1, Seuxoff IJ. Electron microprobe analysis of asbestos bodies. In: Shapiro HA. ed. Pneumoconiosis. Pro ceedings of the international conference. Johannesburg Cape town. Oxford Unix Press, 197057-69. (7) Seuxoff IJ. Hammond EC, On rg J Carcinogenicity of amosiie asbestos Arch Environ Health 1972:25.183-186. (8) Seidman H, Lius R, Seuxoff IJ. Short-term asbestos exposure and delayed cancer risk In: Nieburgs HE. ed. Prevention and detection of cancer Vol 1. New York: Marcel Dekker. Ine., 1976.943-960. (?) Seidman H. Selixofi IJ, Hammond EC. Short-term asbestos work exposure and long-term observation Ann NY Acad Sci 1970. 33061-B9. (iO) Hammond EC. Garfivkel L. Seidman H. Lew EA. Tar and nicotine content of cigarette smoke in relation to death rates. Environ Res 1976,12.263-274. (//) Selixoff IJ. Hammond EC, Seidman H. Mortality experience of insulation workers in the United States and Canada. 19431976. Ann NY .Acad Sci 1979;330:91-116. 02) SELIXOFF IJ. Cancer risk of asbestos exposure. In: Hiatt HH. Watson JD. Winsten JA. eds. Origins of human cancer. New York Cold Spring Harbor. 1977:1765-1784. A02873 UCC 015501 JNCl. VOL. 65. NO J. SEPTEMBER 19H0