Document x1orVEvwow5bewjDdYvw55KG0

<- \fVf Scientific Communications PLAINTIFF'S EXHIBIT The Health of Chrysotile Asbestos Mine and Mill Workers of Quebec J. Corbett McDonald, MD; Margaret R. Becklake, MD; Graham W. Gibbs, PhD; Alison D. McDonald, MD; Charles EL Roositer, MA, Montreal The results of studies of respiratory symptoms and function, roentgenograph*' tc changes, and mortality In relation to dust exposure In the Quebec chrysotile Industry, which has employed some 28,000 workers, are brought together and their Implications tor control examined. Breathlessness on exercise, dimin ished Inspiratory capacity, parenchymal and pleural changes, and respiratory dis ease mortality were related to dust expo sure and to each other. Respiratory can cer was also related to dust exposure. The overall excess of deaths from res piratory cancers. Including five with ma lignant mesothelioma, was at most 50% above expectation, based on age-specific rates for Quebec and the mining region. If safety standards are set, they should be based on epidemiological evidence. From these data lor the chrysotile produc ing Industry ot Quebec, a resonable fig ure, based on a 1% risk ot acquiring clini cally significant disease, would lie between 2 to 4 million particles per cubic foot, calculated for a working life of 50 years. In the absence of epidemiological findings based on fiber counts and lack of a satisfactory means ot conversion, par ticle counts should continue to be used for control In this Industry. In recent years, exposure to as bestos dust has been widely, recog nary function,*** and respiratory symptoms.* The purpose of this report nized as an important etiological fac is to bring together these and certain tor in pulmonary fibrosis, cancer of subsidiary findings and to discuss the lung, and malignant mesothelial their implications for control. tumors. The precise nature of the cau sal association has usually been ob Outline of Study scured by exposure to more than one Records of the mining companies form of asbestos and to other mate were used to identify all persons rials in the manufacture or appli known to have worked in the industry cation of asbestos products. In the since its inception in 1878. Earlier mining and milling industry of Que records tended to be incomplete or bec, where almost half the world's missing and certain sets had been de supply of chrysotQe is produced, expo stroyed, but altogether some 28,000 sure to other types of asbestos or to employees, mostly men, were identi materials used to make asbestos prod fied and information was assembled ucts has been very slight. To assess for this group. Of these, 6,400 were the effects of pure exposure to chryso working on Nov 1,1966, the registra tile, an epidemiological study in this tion date for the study. The informa industry was started in 1966. Our tion collected for each person con aims were to relate dust exposure as sisted of' date of birth and a full sessments to mortality, roentgeno- industrial history. The latest chest graphic change, pulmonary function, roentgenogram of each employee, if and respiratory symptoms. Results of available, was provided by the indus these studies have been presented in trial medical clinics and read by an separate papers on mortality,!-* international group of six readers roentgenographic change,1 pulmo using the UICC/Cincinnati (now Table 1.--Number of Men in Dust Exposure Analyses Submitted tor publication July 2, 1973; ac cepted July 30. From the Department of Epidemiology and Health. McGill University, Montreal Mr. Rossiter b presently with the Medical Research Coun cil Pneumoconiosis Unit, Penarth, South Wales. Reprint requests to Department of Epidemiol, ogy and Health, McGill University, 3775 Univer sity St, Montreal 112, Quebec, Canada (Dr. McDonald). Analysis Mortality study Roentgenographic study Lung function and respiratory symptoms Smokers Nonsmokers Dust Exposure Level, mpef-yr <10 2.810 1,537 10 3.323 1.522 100 1.124 1.133 200 1.007 912 400 837 718 800+ 58S 707 Total 9.692 6.529 91 455 159 134 109 69 381 144 117 9S 22 74 15 17 13 67 1.01S 59 866 8 149 Arch Environ Health/Vol 28. Feb 1974 ^hrysotile Asbestos WorkersAtoDonaTci et al~=61 i 1 ; \ i / > Table 2.--Roentgenographic Change Prevalence Percentages Roentgenographle Chxrg* Irregular small opacities 1/0 2/1 Pleural thickening Grade 1 Grade 2.. Irregular small opacities 1/0 2/1 Pleural thickening Grade 1 Grade 2 Dust Exposure Level, mpcf-yr '<10 10 100 200 400 + Ihetfotd Mines 13 33 S3 8.7 129 173 0 0.2 23 09 23 93 23 43 .13 0.7 Asbestos 63 03 S3 13 83 103 13 29 6.4 3.7 S3 59 0 03 13 03 S3 73 0.6 39 29 23 39 33 0.0 0.6 13 03 4.7' 53 13 0.7 as* 73 29 63 1.1 5.1 03 33 0.7 Rates are standardized for age and years In Industry. Males were 36 to 65 years of age at time of roentgenogram. . Table 3.--Age-Corrected Death Rates Per 1,000 Men Bom 1891-1920: Deaths to December 1969 Cause All causes Respiratory cancers* Abdominal cancers Pneumoconiosis '<10 365 103 18.0 1.6 Dust Exposure Level, mpcf-yr .A. 10 10O 200 400 355 354 313 323 13.1 133 153 213 13.6 18.7 113 263 13 03 49 49 Includes malignant pleural mesothelioma. 800+ 395 32.1 28.7 23.6 Table 4.--Relative Risk of Death From Respiratory Cancer in Men by Dust Group, Estimated From Retrospective Analysis Dust Exposure Level, mpcf-yr < 10 10 too 200 400 800+ Total No. of Cases 32 41 13 14 15 19 134 No. of Controls 186 188 39 61 40 22 536 Relative Risk 19 13 19 13 23 59 Table 5.--Age-Corrected Mortality Per 1,000 Men From Respiratory Caneer and Pneumoconiosis at Thetford Mines and Asbestos Respiratory cancer Thetford Mines Asbestos Pneumoconiosis Thetford Mines Asbestos '<10 9 11 4 0 Dust Exposure Level, mpcf-yr 10 100 2otr 400 12 18 IS 24 13 8 18 13 1 1 76 00 53 800+ 31 43 24 24 ' -'33 UC/ILO) classification.1* A follow-up a of ex-employees was set up and death certificates and autopsy reports were examined for those who had died. One section of this article deals with re sults in women; the rest is concerned : Z with men only. Dust Exposure.--Dust exposure lev els were determined for each recorded job within the industry, and for each year of employment. These levels, were used to calculate a total dust ex- -i posure index for each worker. Gibbs and Lachance" described the tech-' niques used and Gibbs" dealt further with certain qualitative aspects. The main index was based on particle counts from midget impinger sao-lv pies, expressed as millions of particles ~ per cubic foot (mpcf). A total dust ex- > posure index was calculated for each worker by multiplying the measure--!33 ment of dust exposure by the number of years of exposure at that level, and.. l was expressed as mpcf-years. . The results which follow are pre- ' -Sj sented in terms of range of total dust exposure based on particle counts (Tabe 1). Table 1 also lists the number 4 of men included in the main analyses considered in this report. Since disease is probably caused by asbestos fibers rather than dust par- - tides, it has been recommended that standards for occupational exposure. be based on fiber counts obtained by -?sis the membrane filter technique.'* If -32^ there were a reasonably consistent '-*^ relationship between partide counts by midget impinger and fiber counts ^ by membrane filter, the results of our- 'iSE surveys could also be presented and interpreted in terms of fiber counts. :J|| A study was carried out by Gibbs and Lachance" in which 87 side-by-side midget impinger and membrane fil ter measurements were made at nine locations in each of five mines and - -?w mills of the Quebec chrysotfle indus try. The correlation between the two types of measurement was too poor to justify the use of any single conver sion factor. Roentgenographic Changes.--1The two main roentgenogTaphic indices of re sponse to asbestos exposure are ir regular small opacities and, pleural thickening. The prevalences of these, and._of all other roentgenographic 62 Arch Environ Heallh/Vol 28. Feb 1974 Chrysotile Asbestos Workers/McDonald et al=. changes were higher in the Thetford Mines area than in the Asbestos area of Quebec, although the two areas are only 80 km apart and the asbestos mined is geologically similar. The as sociation of the changes with dust ex posure was also stronger at Thetford Mines than at Asbestos. Table 2 shows that at Thetford Mines the prevalence of irregular small opacities of category 1 or more rose steadily with increasing dust ex posure whereas for category 2 or more''the relation to dust exposure was less marked, except for the con siderable rise in the most exposed workers. Even with allowance for dif ferences in dust exposure, the preva lence of category 1 or more was markedly age-related, rising to 14.8% for those 61 to 65 years old. This age effect was much lower for category 2 or more, and amounted to only about one third of the change associated with dust exposure. At Asbestos, the prevalence of ir regular small opacities of category 1. or more rose to 9.2% for those aged 61 to 65, and there was virtually no rela tion to dust exposure. With respect to only category 2 or more, the relation to age was mucn lower and preva lence increased among those with the highest dust exposure. For pleural thickening, the effect of age was slightly greater than the ef fect of exposure on the prevalences of grade 1 or more, and of grade 2 or more, for both mining areas. The biggest difference among all the roentgeuographic features was in pleural calcification, for which the prevalence of'grade 1 or more was 0.4% at Asbestos and 5-2% at Thetford Mines. In spite of the relationship to dust exposure, no correlation ex ceeded 0.3. This is largely a reflec tion of the very high proportion of men who showed no roentgenographic change. Some of the differences be tween the two areas were not surpris ing, as overall dust exposure levels at Asbestos were considerably lower than at Thetford Mines, but others cannot yet be explained in terms of dust exposure or geology. If variation in fiber content of dust were an im portant factor, the results in mine workers, for whom the fiber propor tion would be low, should differ from the Quebec chrysotile industry as a these for mill workers, is whom the whole has been at most 50% above ex liber proportion would be higher. Al pectation and probably about 25%. together, there were 506 workers who The primary method of analysis had worked for ten or more years en used in our report on deaths up to tirely in mining or entirely' in mill 1966* had two main weaknesses. The ing. Analysis of roentgenographic first was that length of exposure changes in these men showed that might well have been related to mill workers had a slightly higher . length of survival and thus might prevalence of irregular small opac tend to obscure differences in mor ities, but the differences between tality between exposure groups. We Thetford Mines and Asbestos re looked for evidence of such an inter mained. No other consistent differ action using the parametric approach ences could be detected between the of Berry,*4 and though we failed to mine and mill workers. find a significant effect, the possi Mortality.--This study was limited bility remained. The second weakness to those who had worked for one lay in the fact that, deaths accumu month or more, and who were born lated over many years were used in a 1891 to 1920. This age group was se single calculation of mortality. It is lected because the records of older reassuring that' our subsequent anal persons were frequently incomplete ysis,* which dealt with deaths over a and few of the younger ones had died. three-year period, 1967 through 1969, An initial analysis' was limited 'to gave essentially the same results. To deaths before Nov 1,1966, but the fol some extent these problems are some low-up will continue until at least what academic, since the total excess 1974. By the end of December 1969, ' mortality in the industry from res 87.5% of the 11^72 persons in the piratory cancer, compared with mor cohort and 99% of those who had tality of the general population, was worked ten years or more had been so smalL The data are of practical im traced.* Of these, 3,270 had died, com portance nevertheless in defining the prising 65.4% of those bom 1891 to form of the dose-response relation 1895 but only 9.8% of those bora 1916 ship, essential for the setting of- to 1920. safety standards. The age-standardized mortality per For this reasoson, when the follow 1,000 men for certain causes of death up is eventually completed, an analy are given in Table 3. Cancer of the sis based on man-years of exposure1* lung showed a rising rate with in is planned to minimize these possible creasing dust exposure, particularly errors. Meantime, another method of in the two highest dust exposure analysis (G. Eyssen, MSc, and F. D. groups. Cancers of the gastrointesti K. Liddell, MA, unpublished data) has nal tract also showed a rise in the two been employed that, we believe, elim highest, and pneumoconiosis in the inates certain of these problems, highest categories of dust exposure. though it does not make full use of Of 134 deaths in men from respira the data available and provides only tory cancer, five were from pleural estimates of relative rather than ab mesothelioma. These cases showed no solute risk. For this analysis, the dust clear relationship with dust exposure. exposures of the 134 men included in There were no peritoneal mesothe the 1969 analysis of respiratory can liomas. Mortality from all causes fell cer mortality were compared with a with increasing dust exposure jop to.:, sample of men, four for each case, se the highest dust group, probably be lected at random among persons-liv cause those who died young could not ing at the time of the death of the attain a high dust exposure. On the respiratory cancer case and born in basis of Quebec death rates, the ex the same year. The distribution of pected number of respiratory cancer cases and controls by dust exposure deaths in the cohort was 139, or about category is presented in Table 4. It 93 on the basis of estimated rates in can be seen that the pattern of rela the mining region. This suggests that tive risk obtained by this approach is mortality from respiratory cancer in quite similar to that for respiratory. Arch Environ Health/Vo! 28. Feb 1974 Chrysotile Asbestos Wor-kefs/McBonatd et at 63==-- fi k s* ' Table 6.--Mortality From All Causes and Roentgenographic Changes at Thetford Mines 189MS9S No. of Observed and Expected Deaths by Year of Birth Cohort* 1896-1900 1901-1905 1906-1910 1911-1915 Change* . Cbs. Parenchymal changes only IT Pleural changes only It Both parenchymal and pleural chances 14 Exp ' Obs IQjS 7 10.3 15 11A 17 Exp ' ' Obs 7J& 4 16Z 10 12.7 8 Exp ' Obs 3J9 11 9.2 5 6.6 6 Exp ' Obs 43 T 5.2 2 2.2 3 Exp 1.6 3.4 0.6 * Expected number was calculated from death rates In men without roentgenographic change. 1916-1920 ' Obs Exp 2 1JS 2 2.4 1 0.2 Total Obs Exp 48 30.0 45 47J5 49 33.7' cancer mortality shown in Table 3. It seemed possible that the roent genographic differences between Thetford Mines and Asbestos might "also be reflected in mortality. There was little evidence of this. The agestandardized mortality per 1,000 pop ulation, for all causes, all malignant neoplasms, all' respiratory diseases, and all circulatory diseases were 342, 54, 20, and 120, respectively, at As bestos and 362,61,22, and 121, respec tively, at Thetford Mines. All detailed comparisons showed the same sim ilarity. For example, in Table 5, the age-standardized mortality by dust exposure is presented for respiratory cancer and pneumoconiosis. The only apparent difference is that the res piratory cancer rate rose in relation to dust slightly earlier but to a less extent at Thetford Mines. Mortality and Roentgenographic Changes.--Of 10,120 persons traced in the mortality study, 9,692 were men and of these 5,082 had chest roent genograms and 785 had died. At Thetford Mines, there were 354 deaths in 2,448 men traced. Death rates were calculated by date of birth, dust exposure, and the presence or absence of parenchymal or pleural roentgenographic changes. The rate from all causes combined was in creased in the highest dust exposure group but, with allowance for expo sure and date of birth, those whose roentgenograms showed parenchymal changes had a higher mortality (220/1,000) than those without roent genographic changes (131/1,000) or with pleural changes only (126/1,000). In Table 6, the observed number of deaths by year of birth and roentgen ographic change is compared with the expected number based on those without roentgenographic change. Of 97 deaths in those with parenchymal changes, 33 were in excess of the expected figure. Considering only deaths from respiratory disease, in cluding tuberculosis and cancer, we calculated in a similar way that there were 32 deaths in those with paren chymal change compared with eight expected, an excess of 24. Thus of the 33 excess deaths in this group at Thetford Mines, 24 were attributed to Table 7.--Age-Corrected Prevalence Percentages.by Dust Exposure and Cumulative Smoking Habits - Bronchitis Never smoked Up to 100 cigaretteyears 100-499 500-999 (.1.000 Never smoked Up to 100 cigarette- Breathlessness years 100-499 500-999 (.1.000 <10 10 Dust Exposure Level, mpef-yr' --............. - - --* - - 10 100 200 400 800+ . 19 19 46 21 49 18 0 12 0 46 0 41 28 21 33 42 43 32 38 44 28 47 58 100 47 45 55 54 35 0 14 24 31 13 44 0 12 39 21 41 43 16 14 8 20 42 18 9 19 21 19 14 39 0 23 27 31 34 46 respiratory causes. At Asbestos, the difference be tween the mortality of those with and without roentgenographic changes was less. Calculations similar to those used for Table 6 showed that only seven of the 431 total deaths were as sociated with parenchymal change, and of these 2.4 were attributable to respiratory causes. Dust exposure levels at Thetford Mines were much higher than at As bestos; thus, in the highest exposure group there were 386 and 69 men, re spectively, on whom roentgenograms had been performed and who had been subsequently traced. Though the excess death rate in this highest ex posure group was about 9% at both places, the numbers of excess deaths were therefore 35 and 6 respectively. Calcified pleura! plaques were not related to mortality as there were 60 deaths in persons with pleural plaques compared with an expected 59 deaths." Pulmonary Function.--A total of 1,015 current employees underwent pulmonary function studies during the summers of 1967 and 1968.*-1 The sample chosen for study was strati fied to include a higher proportion of older than younger workers, as the former were more likely to show changes in function. Those tested comprised 83% of the sample selected. The variation in some of the lung function indices by dust level in smokers and nonsmokers is shown in Fig 1, 2, and 3. The results of each test were standardized to age 50 years, height 170 cm, and weight 70 kg. Total lung volume fell ..slightly with increasing dust exposure, but exposure had little effect on either 64 Arch-Environ Health/Vol 28. Feb 1974 Asbestos Wbrkers/MCDonakJ e=at i  #* functional residual capacity (FRC) or residual volume (RV). Thus exposure to chrysotile affected only the inspira tory capacity portion of the total lung' volume (TLV). Forced vital capacity (FVC) and forced expiratory volume in one second (FEVJ both declined with increasing exposure, the FVC vi tal capacity falling by .about 18% in the highest dust group and the FEV rather less. Neither steady state nor single-breath diffusing capacities showed any effect of exposure at rest; on exercise, the steady state diffusion declined slightly. Analysis of the relation between lung function and roentgenographic changes* showed that in those with small opacities of category 2 or more there was an average reduction of 20% in three indices; FRC, RV, and single-breath diffusing capacity at rest. Most indices showed a progres-. sive reduction with increasing cate gory of small opacities, but only VC and FVC showed significant reduc tions in association with the earliest roentgenographic changes. Pleural changes were also associated with re ductions in pulmonary function, which ranged from about 3% in those without parenchymal change to about 6% in those with advanced parenchy mal change. An additional small sur vey showed that changes in pulmo nary mechanics possibly precede roentgenographic and other function changes.* Detailed analysis of those aged 61 .to 65 showed that those with an obstructive lung function profile, ie, RV and TLC greater than ex pected and flow rates less than ex pected, had had heavier dust expo sure, more symptoms of bronchitis, and more irregular small opacities than those with normal or restrictive patterns of pulmonary function.** Respiratory Symptoms.--Each of the 1,015 workers in the function survey answered a slightly modified version of the Medical Research Council (MRC) respiratory questionnaire given in French or English by a bi lingual interviewer. Both persistent cough and phlegm' (bronchitis) and breathlessness on exercise were re lated to exposure.* Table 7 shows, however, that after standardizing for age, the prevalence of bronchitis rose n o > *.5. CD oouc. cc--n CD Dust Group, mpcf-yr Fig 1.--Variation in Jung volumes with dust group, standardized for age, height, and weight to about 50% in both the highest dust were both decreased to a similar ex and highest smoking categories. tent with increasing dust exposure, Thus, the effect of heavy exposure but only FEV was lower in smokers. and heavy smoking together were the As mentioned above, smoking and same as that of either separately. The dust exposure were both related to same relationships have been ob the prevalence of bronchitis, but served in other dusty trades.'" In con smoking was not associated with trast, the prevalence of breath breathlessness. lessness on exercise was unaffected Attempts are now being made to by smoking, but increased steadily, '.collect smoking histories in the cohort with dust exposure. study of mortality, both for those who Effect of Smoking.--Smoking habits . are still alive and for those who arc were originally determined onl/"for dead. This may eventually provide- employees in the pulmonary function prospective and retrospective evi survey.-Figures 1,2, and 3 show that, dence on the interrelationship of with many of the physiological in smoking and asbestos exposure. A dices, the effect of smoking was controlled retrospective study" has greater than the effect of exposure to been made based on deaths from lung asbestos dust. This was most obvious cancer in the cohort study. Each pa in FRC, RV, and perhaps in the tient was matched with a death from steady-state diffusing capacity at lung cancer from the same hospital rest. Forced vital capacity..and FEV, records in a man-of. approximately Arch Environ Health/Vol 28. Feb 1974 JSt Chrysotile.Asbestos Workers/McDonald et al 65 ity was virtually confined to men with . exposure equivalent to at least 400 mpcf-yr. Detailed examination of mortality showed no particular cause, except pneumoconiosis, with a rate above that of the general population of Quebec. An investigation was made of. the 236 known cases of malignant meso thelioma in Canada, I960 through 1970. The manner of death certifica tions of these patients were exam ined,1' and the pathological findings reviewed in detail by the Mesothe lioma Panel of the Canadian Tumour Reference Centre."-*1 Epidemiologi cal inquiries showed that 23% of men and 1% of women had definite or probable occupational exposure to as bestos, and a further 1% of men and 6% of women had lived in the home of an asbestos worker. In the remaining cases, none had ever lived within 32 km of an asbestos mine or milL Of all known cases of mesothelioma in Can ada, only nine have been associated in any way with the Quebec chrysotile mining and milling industry. Seven (five in the cohort) had been employed and two were women whose fathers Dust Group, mpcf-yr Fig 2.--Variation in forced expiratory volumes with dust group, standardized for age, height, and weight " Had worked in the industry. These cancer rates are very low in comparison with the studies of New York insulation workers** and the London croddolite factory workers," the same age and year of death but posed heavily to asbestos dust. One in both of which there were much who had never been employed in the death was ascribed to lung cancer and higher rates of malignant neoplasms, industry. There was a higher propor none to pneumoconiosis. particularly of mesothelioma. Other tion of nonsmokers in patients than Chest roentgenograms were avail studies in the Soviet Union" and controls, which suggests that pulmo able for 294 women, most of whom Italy" support the view that only nary cancer may be caused by as were 30 years old or less. The only high levels of exposure to chrysotile bestos exposure in the absence of roentgenographic changes recorded during mining and milling have an smoking. The controls were on aver were one subcategory 0/1 rounded appreciable effect on mortality. age heavier smokers, which reinforces small opacities, one subcategory 0/1 Although roentgenographic changes the point but does not rule out the irregular small opacities, and one had some relationships to dust expo possibility of synergism between as grade 1 ill-defined cardiac outline. sure, the clinical significance of the bestos exposure and cigarette smok ing. Clear evidence was obtained Comment minor changes which were also re' fated strongly to age is uncertain. Ir from a survey of all known cases of Although death is the most definite regular small opacities of category 2 malignant mesothelioma in Canada and serious manifestation of expo and pleural thickening of grade 2 are that this disease was not related to sure to asbestos, it is not the most usually considered of clinical impor smoking.** sensitive. Excess deaths related to ex tance, and these were much less influ Women.--The number of women posure altogether were probably in enced by age. Changes of this order ever employed in the Quebec chryso- total no more than 2% of the 3,270 occurred in 1.0% and 0.7% of the en tile industry is small. In the mortality deaths in the cohort study. Most of tire working population. In men aged study, 428 of 465 women in the cohort these were attributed to respiratory 61 to 65, employed an average of 20 were traced and of these 54 had died. cancer or pneumoconiosis and almost years in the industry, the rates were Only 79 women had worked for more all were in the two highest dust expo 5.0% and 2.5% at an average exposure than ten years, and few had been ex sure categories. Thus, excess mortal level of 1-3 mpcf. These rates are very 's Arch Environ Health.'Vol 28, Feb 1974 Dhrysotile Asbestos Workers/"MG0ona1d et al-. J- _ -----.-4 similar to those found in the chrysotile mining and milling industry in Cyprus,** but are much lower than those in the New York insulation workers,** or in the British Royal Na val Dockyards.**-3* For pulmonary function and res piratory symptoms, the effect of smoking was generally greater than the effect of exposure to asbestos. However, for inspiratory capacity, forced flow rates, and breathlessness on exercise, relationships to dust ex posure were found. With respect to these indices, a 10% reduction in nonsmokers occurred after a total expo sure in excess of 100 mpcf-yr. This is close to the lower limit of the dust -- category at which 1% of subjects had grade 2 roentgenographic changes. In the Quebec studies, information on dust exposure was obtained, where as in most other studies quantitative information was not available. Mur phy et al*1 found 11 cases of asbestosis in 101 New England shipyard workers with an average exposure of 120 mpcf-years, compared with one in a nonexposed control series. These workers appear to show a larger ef fect of asbestos exposure than those engaged in chrysotfle production. However, the main constituent of the insulating materials used was amosite, with some chrysotile and no crocidolite. Dust exposure assessments were also made for the Cyprus mining and milling industry," and the rates of roentgenographic change for the same exposure levels were quite sim ilar to those in Quebec. Exposure in dices have also been assessed in a study of two asbestos-cement plants in the United States by Enteriine and Weill" where the main exposure has been to chrysotile and silica, but the results are not yet available. The limit standards for occupa tional exposure to asbestos are based solely on the airborne concentration of dust or fiber, averaged over a life time's work, and the results in this re port have been presented from this point of view. There was some evi dence in our studies that longer expo sures for the same total dust levels are associated with slightty higher prevalence rates of roentgenographic Dust Group, mpcf-yr Fig 3.--Variation in diffusing capacity with dust group, standardized for age. height, and weight change. However, the roentgen ograms were all taken during work ing life, so the correlation between total dust and years of exposure or years since first exposure is high, and the separation of the contributions from these two factors, duration and concentration, would be very diffi cult. The cohort study also showed that respiratory cancer mortality-was greater in those who took more years to reach a given dust level. Therefore, in assessing the overall effect of as bestos exposure, duration of exposure should not really be ignored.'For this reason, it would be unwise to apply the results of this or other descriptive epidemiological studies to widely dif ferent exposure patterns. The British dust standard for chrysotile of 2 fibers per milliliter. averaged over three months, is based on the concept that a 1% risk of ac quiring clinically significant.disease in a 50-year working lifetime of expo sure is acceptable. Considering all facets of disease--death, roentgenographic changes, pulmonary func tion changes, and respiratory symp toms--the 1% risk is reached by men in our third dust exposure category (100 to 200 mpcf-yr). The relationships that exist be tween measurements of total dust and fiber exposure are thus of critical importance if standards are to be set for asbestos production and other as bestos industries. Such standards should be based on dose-response re lationships established by %ound epi demiological inquiries. To date, very few of these have included any qtran- -Arch Environ "Health/Vol 23, Feb 1974 ^~CI?rysotila-Asbes(oirWorkers/McDonald et al 67 1 . titative assessments of dust, let alone of fiber exposure. Though safety stan dards expressed in fiber concentra tions have theoretical and conceptual attractions, there is as yet little or no direct epidemiological evidence on which to base them. Our studies ap pear to provide a reasonable basis for establishing safety standards for chrysotile mining and milling in terms of dust concentration. Without more evidence on the conversion fac tor that should be applied in different parts of the industry, we cannot ex press our results confidently in fiber counts. Further efforts to find a satis- factory means of converting midget impinger to fiber counts are merited but, until new epidemiological evi dence based on fiber counts has been assembled, it appears unwise to dis continue use of particle counts for control in this industry. References L McDonald JC, et ah Mortality in the chrysodle asbestos mines and mills of Quebec. Arch Environ Health 22677-686, 1971. 2. McDonald JC, et at Mortality in the chrysotile producing industry of Quebec K progress re port. Read before the Fourth International Pneumoconiosis Conference, Bucharest, Hun gary, 19TL 3. McDonald AD, et ah Epidemiology of pri mary malignant mesothelial tumours in Canada. Cancer 26314-919,1970. A McDonald AD, McDonald JC: Epidemiologic surveillance of malignant mesothelioma in Can ada. Can. Med Assoc J 109359-362,1973. 3. Rossiter CE, et ah Radiographic changes in chrysotile asbestos mine and mill workers of Quebec. Are* Environ Health 24388-400,1972. S. Beddake MR. et ah Lung function in rela tion to chert radiographic changes in Quebec as bestos workers. Ball Thytiopathel Bap 6:637- 639.1970. 7. Beddake. MR, et ah Lung function in chrysotile asbestos mine and mill workers of Quebec. Arch Environ Health 24:401-409,1972 8. Jodoin G, et ah Early effects of asbestos ex posure on lung function. Am Bee Bap Die 104325-535,197L 9. McDonald JC, et ah Respiratory symptoms . in chrysotile asbestos mine and mill workers of Quebec. Arch Environ Health 24:358-363.1972 10. UICC Cincinnati classification of the ra diographicappearances of pneumoconiosis. Chat 5337-67.1970. 1L Gibbs GW, Lachance M: Dost exposure in the chrysotile asbestos mines and mills of Que bec. Arch Environ Health 24:189-197,1972 12 Gibbs GW: Qualitative aspects of dust ex- posure in the Quebec asbestos mining and mill ing industry, in Walton WH (cdk Inhaled Partides HISurrey, England, Unwin Bros Ltd, 1971, toI 2, ffp 783-799. 12 Standardfir Asbestos Dust Concentration fir Use With Asbestos Begulations 1969, techni cal note 12 Department of Emjflpyment and Productivity, Her Majesty's Factory Inspector ate, 1970. 12 Gibbs GW, Lachance M: Dust fibre rela tionships in the Quebec chrysotile industry. Arch Environ Health, to be published. 12 Berry G: Parametric analysis of disease in cidences is multiway tables. Biometrics 26372579, 1970. 12 Hill ID: Computing man years at risk. Br J Rev Soc Med 26:132-134,1972 17. Gibb* GW: Epidemiology of Pleural Calci fication, thesis.' McGill University, Montreal, 1972 12 Fournier-Maasey G: Pulmonary Function in Quebec Asbestos Workers, thesis.- McGill Uni versity, Montreal, 1972 19. Gilson JC: Occupational bronchitis. Proc B Soc Mod 63357-864,1972 20. Manfreds if Eyssen G: Long cancer and smoking in chrysotile asbestos workers. Arch Environ Health, to be published. 2L Dude S: LTSxactitude des causes de diets. Can J Public Health 62392402 197L 22 Report by Mesothelioma Panel of the Ca nadian Tumour Reference Centre: Primary ma lignant mesothelial tumours in Canada, 19601968: A pathological review. Cancer 31369-876, 1972 22 MagnerD. McDonald AD: Malignant meso thelial tumours: Histological type and asbestos exposure. N Engl J Mti 287370-571,1972 24. Selikoff U, Hammond EC, Churg J: Mor tality experiences of asbestos insulation work- era, in Proceedings of International Confirmees on Pneumoconiosis, Johannesburg. Capetown. South Africa, Oxford University Prose, 1972 pp 180-186.. 22 Newhouse ML: A study of the mortality of J workers in an asbestos factory. Br J Ind Med 26294-301,1969. 22 Kogan FM, Guseinikova NA, Gulevslcaya MR: The cancer mortality rate among workers in : the asbestos industry of the Urals: Gig Sardt.'3729-32 1972 27. Vlgliani EC: Asbestos exposure andits re- ; suits in Italy, in Proceedings of International ' Conference on Pneumoconiosis, Johannesburg. Capetown, South Africa, Oxford University Press, 1970, pp 192-192 22 McDonald JC: Asbestos in chrysotile mines and mills. Read before the IARC Conference oa the Biological Effects of Asbestos, Lyon, France,. 1972 29. Harries PG: The Effects and Control ofDie-' cases Associated With Exposure to Asbestos in a . Naval Dockyard, thesis. University of London. London 1970, pp 141*272 30. Harries PG, et at Radiological survey of men exposed to asbestos in Naval dockyards: Br J Ind Med 29274-279, 1972 3L Murphy RLH, et at Effects of low concen trations of asbestos. N Engl J Med 235:1271- ' 1278,197L 32 Enteriine P, Weill H: Asbestosia in as bestos cement workers. Read before the L\RC Conference oa the Biological Effects of Asbestos, Lyon, France, 1972 68 Arch Environ Health/Vol 28. Feb 1974 -<S / i Chrysotile Asbestos Workers/McDonald-et 51==- ^