Document qdQErOyByorEmkMVKLKo5xVMj

FILE NAME Talc TALC DATE 1976 Mar DOC TALC140 DOCUMENT DESCRIPTION Journal Article - Mortality Study of Talc Miners and Millers 2003 INIST CNRS Tous droits de propri^'t^i'ntellectuelle r^'serv^'sReproduction repr^'sentationet diffusion interdites Loi du 01/07/92 Articles 5. 6 et 7 des CGV Mortality Study of Talc Miners and Millers Giovanni F. Rubino M.D Giovanni Scansetti M.D Giorgio Piolatto M.D and Canzio A. Romano M.D. Th purpose of this research is to ascertain the causes of death in talc miners and millers Special attention is paid to the incidence of pulmonary mesothelial and gastrointestinal tumors Our study was conducted on a talc which is mineralogically very pure in its natural state This is the type of material used in the pharmaceutical and cosmetic industries All the subjects of our study have worked in the mines and the mills of the talc operation in the Germanasca and Chisone Valley Piedmont This particular talc has been mined for many decades and has continued to be recognized to be of the highest standard of purity see Appendix An epidemiological study which can be taken as reference for our investigation was conducted by Kleinfeld et al and recently reviewed with additional informations 1974 on talc miners and millers in the State of New York.1 2 The incidence of carcinoma of the lung and pleura in those workers was found to be about four times that which is observed in the general population This New York Study dustrial talc which is a was on workers exposed mixture of talc and other to an in- materials Kleinfeld points out that the talc to which his workers were exposed contained not only mineral talc but also other silicates such as serpentine tremolite anthophyllite and other ingredients predominately carbonates Moreover free silica was present in the dust in variable amounts usually averaging not more than three percent Therefore our present epidemiological study provides a basis for comparing subjects exposed to an impure industrial type of talc e.g. Kleinfeld Study to subjects exposed to a very pure type of talc Presented at the 60th Annual Meeting of the American Occupational Medical Association American Industrial Health Conference April 15 1975 San Fran- CISCO Dr Rubino is Professor of Occupational Medicine and Director Institute of Occupational Medicine Turin University Drs Scansetti Piolatto and Romano are assistants Institute of Occupational Medicine Turin University Reprint requests to Dept of Occupational Medicine University of Torino CTO - Via Zuretti 29 10126 Torino Italy 186 Type of Study and Data Collection An historic prospective study has been employed consisting in the follow of a cohort of male subjects who began work at a definite time and including retired as well as active workers We have been able to do so because we had at our disposal a group of subjects fit for this type of investigation Data were collected for each worker who began workin the years between 1921 and 1950 and who has been employed for at least one year in a job involving exposure to talc in the mines or in the mills Necessary informations such as birth date place of birth last known address type and duration of all works performed were abstracted from pay rolls and DUST CONCENTRATION IN AIR ALONG 1000 OBSERVATION PERIOD MPCF FO T CUBIC 100 PER _ MINES --- MILLS PARTICLES o MILON 10 . OF LOGARITHM 1 r 1 1965 4 f po 1800-1845 1800-1845 1960 1965 1960 1965 1970 19755 YEARS OF OBSERVATION' following adoption Decreasing trend of total dust in respirable range cal preventive means from 1950 * the of techni- ~~ interdites 2003 INIST CNRS Tous droits de Reproduction propri^'t^i'ntellectuelle r^'serv^'sReproduction repr^'sentation repr^'sentationet diffusion interdites Lot du 01/07/92 01/07/92 Articles 5. 6 et 7 des CGV records of a mining company of Pinerolo Piedmont Cumulative exposure for each worker was estimated from the results of successive determinations of air dust content from 1948 till today see Figure and quantified by calculating an approximative value of the total amount of inhaled particles during the employement period This one was considered to be finished at the retirement time for retired workers and on June 30 1974 for active workers To compute cumulative dust exposure for each man periods of job along which the dust level was uniform were selected then the number of particles per cubic foot was multiplied by the number of years within each period and the resulting values for each period were summed The cumulative exposure is then expressed as mppcf Thus if a worker has been employed for 12 years at the follow- ing exposure to dust 10 years at 566 mppcf 1 year at 440 mppcf 1 year at 360 mppcf his cumulative exposure would be 10 x All 566 + workers 1 x 440 were then + 1 x 360 = 6460 mppcf classified on the basis of three ex- posure levels for miners and millers as follows see also Table 2 Miners Exposure Level 1 mppcf Exposure Level 2 mppcf Exposure Level 3 mppcf Millers 566 - 1699 1700 - 5665 5666 - 12750 Exposure Level 1.mppcf 25 - 141 Exposure Level 2 mppcf 142-424 142-424 142-424 Exposure Level 3 mppcf 425-906 425-906 425-906 However we should remember that miners and millers ex- posures are not comparable because air dust in the mines in- cludes a certain amount of inhalable silica particles see Appendix Therefore we have separately studied miners and millers obviously excluding from the cohort of millers all sub- jects who have worked in the mines even for a very short period of time We gathered a total of 1514 miners and 478 millers qualified by our definition of the sample For purpose of comparison with a exposed group we Table 3 Vital Status of 1346 Miners and 438 Millers With Completed Follow and Respective Controls on June 1974 Dead Alive Miners No. | 704 . % 523 - Controls = - No. 851 vf % 63.2 - 642 47.7 ' 495 ak ee Millers No. | 227 _ 211 % 51.8 48.2 Controls =. No. *] > 275 163 rn) 62.8 ee Oe a E Known Cauosf Deeasth ae .? eee aN ge BS SE ie Eee thought it advisable to use an ad hoc control sample matched by age to the exposed workers sample rather than a comparison with national death rates which could introduce_ some confounding variables such as feeding habits ethnic characteristics differences etc . ) Control subjects were taken from the population of the town of Alba which presents similar ethnic characteristics and com- parable social and economic conditions Alba is a Pied- montese center about 38 miles from the Pinerolo Valleys with a dominant traditional agricultural activity and very little industrial development not involving any obvious hazard of occu- pational diseases or carcinogenesis to its workers Every subject in the exposed groups was individually matched to a control subject by the following procedure From the City Hall records all subjects born in the same year as a given exposed worker and still alive at the time of his first employement were ordered and listed according to their month and date of birth Among these the first available i.e. not already used for matching subject was selected as the matched control subject As it turned out by this procedure almost the full list of subjects potentially suitable for matching was in the end included in the control sample Journal of Occupational Medicine 18 No. March 1976 Follow of Study Population A follow procedure was instituted for every exposed status subject who had left employment and whose vital could not be determined at the plant as well as for the whole control group The City Hall registries of birth place and death place provided information on vital status Follow is 88.9 complete on miners and 91.6 complete on millers - Table 1 shows the follow and vital status of the exposed population and their controls as of June 30 1974 The distribution of all workers on whom follow was complete according to the Exposure Level is shown in Table 2 187 2003 INIST CNRS Tous droits de propri^'t^i'ntellectuelle r^'serv^'sReproduction repr^'sentationet diffusion interdites . Loi du 01/07/92 Articles 5 6 et 7 des CGV Latency periods interval between first exposure and death Table 5 --- Overall Mortality Observed Deaths were 1-10 11- 21 - 30 31 - 40 41 - 50 > 50 by Interval Between First Exposure and Death General outline is shown in Table 4 by Exposed vs. Controls Comparison Number of deaths to be expected on null hypothesis for Latency yrs Miners fo ... Millers Obs./Exp Ratio Obs./Exp Ratio each latency period group at risk under examination ZE was computed as ,,+ ,, 1-10 84.2 0.81+ |. 29.6 =. 0.54+ EXPECTED DEATHS = ~ X ZEI 11-20 | 119.6 21.30 | 208.8 ' 0.92 .. 0.97 F.41.7 |. 67.9 1.10 0.97 ZEI + Za While for the comparison between exposed workers and | 31-40 207.0 =. 0.91 72.5 0.79 2 controls the combined group exposed + controls was eS 89+ 258.4 41-50 TS 157.5 - 080+ 38/42 > -0.90 taken as the reference group ZE + Zo the comparison among classes of differently exposed workers the whole ex- 50+ LY 14.1 0.64 4.5 0.89 | posed group was taken as reference : So Total 791.2 7 S > 0.88+ > Deaths to be expected were computed for overall mortality * Statistically significant at % level Statistically significant at % level and for 22 major causes of death Expected and observed values of deaths were added up for The information relating to cause of death was obtained from death certificates for both exposed and controls and was coded according to the International Intermediate Classification of 150 causes of deaths with additional items mesothelioma tuberculosis related to the possible effect of occupational exposure Death certificates are the source of information used by the Italian Central Institute of Statistics ISTAT to compute national mortality rates Although they may not always be completely accurate they are available in a high percentage of cases and are suitable for epidemiological purposes and widely used by many authors allowing to compare two populations with homogeneous diagnostic different ages within each latency period cell in order to examine the total impact of mortality along the observation period see Tables 5 6 7 Pooling for age and latency was performed for comparison according to specific causes of death see Tables 8 9 10 11 12 The ratio of the number of observed deaths to the number to be expected provides an indication of the excess of mor- tality in the group with risk under examination in respect of the standard population reference group One degree of freedom chi square tests were used to assess statistical significance of individual comparisons exposed vs controls exposed at Level 1 vs. exposed at Level 2 criteria A total of 1840 death certificates were collected When the cause of death was not found by death certificates or if there were not enough indications for diagnosis e.g. cardiac failure reported as death cause a further enquiry was started with relatives of deceased attending physicians or examining hospital records for hospital deaths A further number of 112 causes of death 59 for miners 39 for miners controls 5 for millers 9 for millers controls were so obtained to give the total number of known causes of death reported in Table 3 The subsequent statistical analysis was conducted on 1346 miners and 438 millers on whom follow was complete Workers reported dead but for whom neither death certificates nor the above sources of information were available were con- sidered as deaths due to unknown cause No smoking data were available for dead people in exposed and control groups However as the two populations exposed and controls come from areas previously studied from ethnic social and economic point of view we have no grounds for assuming that there are differences in smoking as well as in feeding habits * Statisticalty significant at % level Statistically significant at % level Lot an mo Methods of Analysis Statistical analysis was conducted by a comparison between exposed groups miners and millersand their matched con- trols and within exposed samples by an internal comparison among the three above mentioned groups with different cumulative dose of inhaled particles . - For both exposed versus controls comparison and internal comparison a modified life table technique adjusting for age by indirect standardization method was used to compute ex, : pected number of deaths Age intervals were < 30 30- 50 51 - 70 > 70 188 * Statistically significant at 5 . level Statistically significant at % level 2003 INIST CNRS Tous droits de propri^'t^i'ntellectuelle r^'serv^'Rseproduction repr^'sentationet diffusion interdites Loi du 01/07/92 Article 5 6 et 7 des CGV and 3 etc. etc. ) standardized as previously described for age or for age and latency In each table a single asterisk indicates the ratios which had a probability of 0.05 of occurring by change and a dagger indicates those which had a probability of 0.01 or less Statistical significance of comparison between cells containing less than five cases was not tested . Results By exposed versus controls comparison the observed overall mortality of talc miners and millers is significantly lower than expected There are 704 observed deaths compared with 791.2 expected among miners and 227 observed compared with 258.4 expected among millers Table 5 shows that no relationship exists between the ratio of observed to expected deaths and interval between first exposure and death Among miners a number of individuals who died more than 41 years after their initial exposure has been found significantly below its expected values By the comparison among different exposure classes the ratio does not increase with increasing exposure As shown in Tables 6 and 7 in both miners and millers an excess of ob- served to expected deaths occurs at Exposure Level 2 for the total sample In the high exposure groups the excess does not persist and on the contrary statistically lower than expected number of deaths is detectable Neither in this last group nor in the others are there increases in ratio with increasing interval between first exposure and death for miners and millers When the overall mortality of study populations is subdivided according to cause of death some major patterns emerge for miners Table 8 shows that by exposed versus con- trols comparison death from all respiratory diseases is significantly higher than expected as opposed to malignant neoplasms as a whole cardiovascular diseases gastrointestinal diseases and all other causes which show a number of ob- served cases significantly lower than expected Splitting respiratory diseases by kind provides evidence that the excess of mortality in exposed miners is due to the high incidence of pneumoconiosis with or without tuberculosis Among all malignant neoplasms lung cancer shows an incidence significantly below its expected values By comparison among different exposure classes as shown in Table 9 the ratio of observed to expected deaths from respiratory cause constantly increases with increasing exposure because of the excess of pneumoconiosis in people with greatest exposure An increasing trend is also shown by larynx cancer but the number of cases is too small to assess significance of comparison For all malignant neoplasms oesophagic and gastric cancer as well as for tuberculosis and senility the ratio decreases with increasing exposure but no significant values are obtained in any cell to refuse the hypothesis of no dif- ference among the three groups No clear decreasing or increasing trends are detectable for all other causes of death in miners Millers show an observed number of deaths below its ex- pected values for cardiovascular diseases Table 8 As shown in Table 10 for all causes of death there are no consistent relationships between the ratio and the exposure level Considering the trend of mortality by cause and by interval between first exposure and death all malignant neoplasms and lung cancer decrease in miners with increasing latency see Table 11 as opposed to respiratory diseases as a * Statistically significant at % level Statistically significant at % level Journal of Occupational Medicine 18 No. March 1976 2003 INIST CNRS intelectuele Tous droitdrsoits dede propri^'t^i'ntellectuelle r^'serv^'s r^'serv^'Rseproduction repr^'sentationet diffusion interdites Loi du 01/07/92 Articles 5. 6 et 7 des CGV whole which significantly increase with increasing latency We should underline that for silicosis with or without tuberculosis worked at least one year therefore including a substantial number with a very long period of observation were included the ratio shows unchanged values over time because of the absence of pneumoconiosis in control sample see computation of expected deaths but the number of observed cases shows a constant increase with increasing latency which explains the excess in ratio for all respiratory diseases As shown in Table 12 for millers only one pattern emerges in the study Cause b seems very unlikely methods of data collection on mortality were the same for the two groups exposed and controls respectively and in both groups a similar number of cases were lost to follow for reasons unrelated to mortality as accidental destruction of record files at the local City Hall registries the ratio for all tumors increases with increasing latency but carelessness of clerical workers etc .. the number of observed deaths remains less than expected in Cause c is less easy to discard although a definite in- dication that it cannot provide the entire explanation for the any cell in control comes from the uniformity of Two cases of pleural mesothelioma were found in controls higher mortality group death certificates confirmed by hospital records but no the mortality ratio during the observation period Assuming autopsy data were available to the investigators The occupa- that lower mortality of the exposed group is due to selection of individuals the effect would have vanished after tional anamnesis carried out with relatives has not revealed more healthy one or two decades on the contrary it remains practically un- anW y eexsphoosuulrde atlosoasubnedsetrolsindeusthtat in both miners and millers a changed for the whole period of observation see latency tac} penms number of oesophagic cancer cases higher than expected was statistical tables Cause d could be the main contributor to the observed dif- found even if the small number cannot provide ferences As already mentioned section Data Collection male eH significance to the comparison on controls were matched by age to exposed individuals and . Uitlaned Discussion On general grounds differences between exposed and con- cPinehalm trols mortality as we observed could be due to several factors a a biased selection of the exposed groups b a difference In recording mortality in the exposed and in the control groups c the well known effect whereby occupational fit groups experience a lower mortality than controls taken from the general population because of a selection upon starting work and d a lack of comparability of the two groups Cause a can be excluded because all the workers who have been enrolled between 1921 and 1950 and who have drawn from a population similar in economic and social structure and general habits However it could be possible that the matched population exhibits for not easily understandable reasons a constantly higher death rate in respect the whole population of the Piedmont Region But this is not easy to verify because of the lack so far back in the past of regional statistical data for this comparison It may be concluded that exposed vs. controls comparison has to be taken with caution and not generalized as confident- ly as if no doubt would exist on control group adequacy However further insight in accordance with exposed versus 190 * Statistically significant at % level Statistically significant at % level Mortality Studoyf Talc Miners and Millers 2003 INIST CNRS Tous droits de propri^'t^i'ntellectuelle r^'serv^'Rseproduction , repr^'sentationet diffusion interdites Loi du 01/07/92 Articles 5 6 et 7 des CGV controls comparison may be gained by the internal comparisons of groups of workers exposed to different cumulative dose of inhaled particles and splitting total mortality by cause For millers no clear effect relationship shows up either for total mortality considered over time of latency or for cause of death For miners the only one clear pattern emerges for respiratory causes of death here in spite of the lower overall mortality in the exposed group an excess of mortality is present in the exposed versus controls comparison and as an increasing mortality with increasing dose and latency period The trend of mortality from respiratory diseases in relation with dose and latency and the different incidence of silicosis respectively in miners and in millers allow us to assume that inducing factor is silica rather than talc The deficit of tumors in the exposed in respect of controls is supported by the absence of any effect pattern and is more easily interpreted as just one aspect of the lower mortality of the exposed group The lack of lung cancer in miners is similar to that observed by several authors Kennaway and Kennaway and more recently Goldman Ashley En- terline15 and Liddell have found a strikingly low rate of lung cancer in coal miners Gooding has found an incidence of lung cancer of 0.5 in the autopsied cases among 592 coal by ft nos, a - - ne ogea wesTable 10- Millers ~~ ExpectedDeaths andRatios Cause InternalComparison * Death Causeof aa . wag FN OY -Tuberculosisall sites , except except associated with silicosis 7 Respiratorydiseases Call except pulmonarytuberculosis ere J Respiratorydiseases -* Silicosis aes tuberculosis = Respiratory chronic diseases infectious = cs : Other ~ <* vt Malignant neoplasms<4... Deeg Malignant bronchus neoplasms sy bronchus = mesothelioma ars Wlee oan ae ce as err Re YH a oy neoplasms primary : Malignant |neoplasms othersit Bt te haD And SZ ss 16.5182 rr . wmes yA sfyf Ae NyoF bein 2 SS PE Be ow SY ngs 18 3Z12R73127 * 1.34 we 4 Sg, "I * Statistically significant at % level Statistically significant at % level Ea 2 Journal of Occupational Medicine 18 No. March 1976 . 191 2003 INIST CNRS Tous droits de propri^'t^i'ntellectuelle r^'serv^'Rseproduction repr^'sentationet diffusion interdites Loi du 01/07/92 Articles 5 6 et 7 des CGV miners in South Wales while Doll in reviewing all the autopsies of male subjects in general hospitals in England has found an autopsy lung cancer incidence of 6.2 No risk of excess of lung cancer was detected by Waxweiler examining death certificates of potash workers Therefore our findings are in accordance with the hypothesis that lung cancer is unrelated with underground mining in absence of specific cancerogenic agents radon daughters asbestos haematite Oesophagic cancer has shown an incidence in exposed groups higher than expected although not statistically significant This needs further studies at a more advanced status of the follow In conclusion exposure to silica in talc mines increases the risk of death from respiratory causes No increase of risk is detectable for other causes of death in miners nor for any cause of death in talc exposed millers While Kleinfeld's Study provides evidence of cancerogeni- city risk when talc mining or milling is a source of asbestos exposure caused by fibrous contamination our conclusions support the thesis of no cancerogenic effect attributable to pure talc From these considerations emerges the definite indication that adequate environmental analyses must be performed in case of exposure to talc to assess the possible presence of Ra jo Whe Sal ServedExpected rs al Expected ee Par Be Servedty SF W Nesy Expected ve lh EB o&e CO Ba gE Ts fibrous materials contamination and to provide in this last case to an adequate prevention Appendix Results of mineral samples examinations and of environmental determinations are reported in this Appendix Mineralogical Examinations A greatly detailed report on minerals coming from the mines and the mills of Val Chisone was carried out by Dr. Pooley in 1972. The examined samples included footwall contact rocks rock type inclusions carbonate calcite and magnesite inclusions and talc specimens from mines as well as talc powders from mills Specimens were examined by optical and electron microscopy and by rays diffraction ; For specific results of each examination we should refer to complete text However patterns emerging from the conclusions of the authors are the following Footwall contact rocks are mainly composed of mineral quartz muscovite chlorite garnet some carbonate material both calcite and magnesite and other materials in minor amount Rock type inclusions in talc have similar composition to footwall rocks In few specimens of footwall rocks as well as in inclusions be e hee Average of mppcf in the years within each period Not measurable Data not available 8 192 aN 2003 INIST CNRS Tous droits de propri^'t^i'ntellectuelle r^'serv^'Rseproduction repr^'sentationet diffusion interdites Loi du 01/07/92 Articles 5 6 et 7 des CGV SE of rock and a carbonate little amount of tremolite was detect- ed Neither other types of anphybolic asbestos nor chrysotile were detected in any amount in rocks and in inclusions Talc specimens were found very commonly contaminated by chlorite No anphybole or chrysotile minerals were detected in any of the examined talc specimens Free silica content in this talc is discussed by ray diffraction neither Parmeggiani in 1948,20 nor Rubino in 1963,21 nor Pooley in 1972 have found detectable free silica amount quartz peaks in mineral talc powdered specimens By con- trast an ray diffractogram performed by Jerry Krause Colorado School of Mines Research - Institute Personal com- munication in 1974 on samples of commercial Val.Chisone Val.Chisone talc powder has shown a minor amount of chlorite and quartz plus very minor to trace amount of magnesite and dolomite in addition to talc This could be due to some microinclusions in talc observed also by Pooley which as well known are not even detected and then eliminated by means as milling photoelectron screening used in Val Chisone mills Analytical examinations conducted by Grill22 on mineralogical composition of rock strata in Germanasca Valley have shown that is not possible to state an average value for quartz amount in footwall rocks This value can vary in a range which is assumed by Parmeggiani in 194820 to be from 10 to 25 Rubino21 has indicated an average amount of 45 in the strata mined in 1963 Environmental Determinations Environmental data were available for mines and mills since 1948 In sampling and measuring airborne particulate contaminant termic precipitator and optical microscopy for dust count were used until 1964 these measurements were performed by the mining company Since that year determinations were carried out by the Department of Occupational Medicine Turin University using membrane collection technique and optical microscopy Conditions in microscopy examinations were the following enlarging 500 x transmitted light in phase contrast for free silica polarized and dark field for fiber count The trend of workroom air total dust in respirable range 0.5 - 5 ...asdefined by British Medical Research Council criteria and expressed in mppcf is visualized in Fig 1 for the whole period of observation of the exposed workers The trend was obtained by linking the points corresponding to the dust level in the years in which air dust determinations were carried out As no changes modifying the environment occurred before 1950 dust level in the years 1920 - 1950 was assumed to be uniform and equal to that of 1948. The decrease of dust content from 1950 to the actual values in mines is due to the successive applications of means of technical prevention which can be summarized as follows Period before 1950 was characterized by dry drilling and the absence of any forced ventilation system Since 1950 wet drilling was introduced and more widely applied in the following years Forced ventilatory system was applied in the years 1958 1959 by air introduction and consequent dilution of dust until 1963 and by complete exhaust system after that year It should be noted parenthetically that adoption of prevention means was encouraged by the law n 198 of March 4 1958 on Rules of Policy of Mines which establish in the mines Journal of Occupational Medicine 18 No. March 1976 a limit of 650 particles in the range 0.5-5... 0.5-5... 0.5-5... per cubic centimeter that is 18.41 mppcf Actual values of dust level in the mills were obtained by successive improvement of dust suction systems Table 13 shows a remarkable difference of free silica amount in air dust respectively in the mines and in the mills and within the mines jobs between drilling and other operations This is due to the high content of quartz in footwall rocks and inclusions as opposed to the absence of free silica in talc minerals The small amount of free silica in mills operations is due as above mentioned to the actual incomplete screening of talc inclusions The same explanation could be given for the very small number of fibers in air caused by possible microinclusions of rock containing little amount of tremolite An examination of Italian Mine Samples and Relevant Powders J. Lighfoot G. A Kingston F D Pooley Department of Mineral Exploitation University College Cardiff 1972 Acknowledgment The authors express their indebtedness to Dr. Rodolfo Saracci for his assistance in data elaboration to Dr Alessandro Pauna Dr. Paolo Vergnano Mrs Michela Ghignone and Ms Piera Chignone for their assistance in data collection and to Mr Mario Vallotti and Arnaldo Forneron for their program assistance References 1. Kleinfeld M Messite ) Kooyman O et al Mortality among talc miners and millers in New York State Arch Environ Hith 14-663-667 1967 2 Kleinfeld M MessiteJZaki MH Mortality experiences among talc workers A follow study IOM 345-349 1974 3 MacMahon B Pugh TE Epidemiology Principles and Methods Boston Little Brown and Company 1970 pp 207-239 4. Bradford Hill A Principles of Medical Statistics The Lancet Limit- ed London 1971 pp 14-40 5. Istituto Centrale di Statistica Classificazioni delle malattie trauma- tismi e delle cause di morte VIII Revisione 1965 - Publ 1972 pp 283-292 6. Enterline PE Kendrick MA Asbestos dust exposures at various levels and mortality Arch Environ Hith 181-186 1967 7. Enterline PE De Coufle P Henderson V Mortality in relation to occupational exposure in the asbestos industry JOM 897-903 1972 8. Tabershaw IR Gaffey WR Mortality study of workers in the manu- facture of vinyl chloride and its polymers JOM 509-518 1974 9. Waxweiler RJ Wagoner JK Archer VE Mortality of potash workers JOM 486-489 1973 10. Meurman LO Kiviluoto R Hakama M. Mortality and morbidity among the working population of anthophyllite asbestos miners in Finland Brit | Indust Med 105-112 1974 11. Armitage P Statistica Medica Feltrinelli 1975 pp 375-379 12 Kennaway MM Kennaway EL The incidence of cancer of the lung in coal miners in England and Wales Can| c Cae ncerr 10-18 1953 13. Goldman KP Mortality of coal miners from carcinoma of the lung Brit ) Indust Med 72-77 1975 14. Ashley DJB The distribution of lung cancer and bronchitis in England and Wales Brit J Cancer 243 1967 15 Enterline PE Mortality among coal miners in the U.S.A. Ann NY Acad Sci New York 1971 16. Liddell FDK Mortality of British coal miners in 1961. Brit J Indust Med 15-24 1973 miners 17. Gooding CG Pneumoconiosis in South Wales anthracite Lancet 891-896 1946 18. Doll R Bronchial carcinoma Incidence and aetiology Brit Med } 521-527 1953 possible 19. Blejer HP Arlon R Talc A occupational and environ- mental carcinogen JOM 92-97 1973 mugnai det 20. Parmeggiani L Le pneumoconiosi dei minatori e dei talco nel Pinerolese Rassegna di Medicina Industriale 1-33 1948 21. Rubino GF Maranzana P Pettinati L Scansetti G Aspetti etio- Medicina del patologici e clinici delle Pneumoconiosi da talco La Lavoro 496-506 1963 22. Grill E Pagliani G Sacchi L La genesi del talco della Valle Ger- manasca Alpi Cozie Rendiconti Istituto Lombardo Scienze e Lettere 442-490 1955 i -" ~- . .