American Journal of Respiratory and Critical Care Medicine

This study aimed to estimate exposure-response relationships for mesothelioma and environmental exposure to crocidolite. All 4,659 former residents of Wittenoom, Western Australia (WA) who lived there between 1943 and 1993 for at least 1 mo and were not directly employed in the crocidolite industry, were followed-up through the WA death, cancer and mesothelioma registries, electoral rolls, and telephone books. In 1992, all subjects who should be traced were sent a questionnaire. Exposure levels were estimated from results of periodic environmental surveys and duration of residence. Incidence rates were standardized to the World Population and Cox Regression was used to estimate the effects of exposure on incidence. To the end of 1993, 27 cases of mesothelioma were diagnosed. Mesothelioma cases stayed longer at Wittenoom, had a higher average intensity of exposure, and a higher cumulative exposure to crocidolite than control subjects. The standardized incidence of mesothelioma was 260 per million person-years, and was similar for males and females. The rate increased significantly with time from first exposure, duration of exposure and cumulative exposure. At these levels of crocidolite exposure, there is a significantly increased risk of mesothelioma, which is dose-dependent.

Epidemiological studies have shown that occupational exposure to the amphibole varieties of asbestos causes mesothelioma (1-4). Exposure has been assessed by a variety of methods, but all these studies reached the conclusion that the rate of mesothelioma increases as the level of exposure increases. The development of this disease has a long latency, with cases usually arising more than 10 yr after first exposure, and continues to rise exponentially with time since first exposure (5, 6). The highest rates of mesothelioma are experienced by those employed in the milling and mining of the crocidolite asbestos minerals. It is difficult to interpret the different risks associated with various processes because they could be due to either the degree of exposure to the fiber type, or both.

That nonoccupational or environmental exposure to asbestos, particularly the amphiboles, is associated with an increased risk of mesothelioma is no longer in doubt (7-18). Exposure is experienced by individuals living with asbestos workers (7, 10, 13-15, 18), and those living near asbestos mines and mills (8, 9), or factories manufacturing asbestos products (11, 12, 17). Similar to the occupational setting, it appears that exposure to crocidolite confers the greatest risk, especially if that exposure was gained by living in the neighborhood of a crocidolite mine or mill. No previous study of environmental exposure to asbestos and the risk of mesothelioma has been able to utilize exposure levels to derive quantitative exposure-response relationships.

Nearly 7,000 people were employed by the Australian Blue Asbestos (ABA) Company at the Wittenoom crocidolite mine and mill in Western Australia (WA) between 1943 and 1966. They experienced an increased risk of mesothelioma, which was dose-dependent (5). A cohort of former residents of Wittenoom, who lived there between 1943 and 1993 for at least 1 mo and were not directly employed in the mining and milling of crocidolite, has been established (19). This study aimed to estimate exposure-response relationships between environmental exposure to crocidolite and mesothelioma in these subjects.

The Wittenoom residents' cohort was established using information from numerous sources (19). These included the local school register, hospital attendances, and birth registrations, information provided by former Wittenoom crocidolite workers in 1979, and participants in a cancer-prevention program using vitamin A supplements. These residents include the wives and children of former ABA workers, government employees (such as teachers, hospital staff, and police), people who worked for other mining companies that used Wittenoom as a base camp for exploration of the surrounding area, people who were self-employed, and families of these people.

Names of subjects in this cohort had already been checked against the Wittenoom workers' cohort and records of a benevolent fund to which all Wittenoom mine workers contributed until 1956, so the residents' cohort did not include names of anyone directly employed in the milling and mining of crocidolite at Wittenoom. To determine the number of subjects who had other occupational exposure to crocidolite at Wittenoom, the following sources were used: Wittenoom residents who could be traced were sent a questionnaire and asked to provide details of their occupation at Wittenoom; many records used to establish the residents' cohort had details of the person's occupation while at Wittenoom, for example, electoral rolls, hospital and local doctor records, and school records which gave information on occupation of parents of students enrolled; and Wittenoom workers provided information about relatives and friends living with them at Wittenoom, including occupation.

Ethical approval for this study was obtained from the Committee for Human Rights of the University of Western Australia, and permission to assess Health Department of Western Australia records was obtained from the Confidentiality of Health Information Committee for Western Australia.

All females not known to be at a current address were sought in the marriage register for Western Australia at the Registrar General's Office in Perth. The search for each individual commenced in the year she turned 15 (for those not known to have been married) or the year last known to be alive. This procedure was performed before proceeding with any other searches.

To determine the number and causes of deaths in this cohort, computerized record linkage procedures (20) were used to search the annual death index for Western Australia obtained from the Registrar General for each year from 1969 to 1993. Matches between names of people in the cohort and names produced by the linkage procedure were accepted only after clerical inspection of all identifying data. Deaths occurring between 1943 and 1968 were found by manually searching the death register. For the years 1969 to 1993, the coded cause of death was obtained electronically from the death tapes for Western Australia. For all other deaths, the cause of death was coded by an experienced nosologist.

Records of all cases of malignant mesothelioma in Western Australia, including deaths, are registered by the Western Australia Mesothelioma Registry. The Australian Mesothelioma Surveillance Programme is notified of other cases occurring elsewhere throughout Australia. Records of both these registers were inspected to obtain cases arising in this cohort, and the date of diagnosis of mesothelioma was recorded.

People not known through their participation in a cancer prevention program using vitamin A, and not known to be dead, were linked by computer to the Western Australia Electoral Roll. The residual list of names was searched for in the Commonwealth Electoral Roll for the other States of Australia, and in the Australian telephone directory. Traced relatives of untraced subjects were contacted by letter or phone to ask for their relatives' current address. Subjects were then mailed a questionnaire that gathered information on the time they spent at Wittenoom, smoking and medical histories, and demographic data. Participants of the vitamin A program provided this information when they enrolled in the program.

Methods of estimation of individual exposure levels have been described in detail elsewhere (21). Briefly, subjects not working directly with asbestos were assigned an intensity of exposure of 1.0 f/ml from 1943 to 1957 (when a new mill was commissioned and the town was moved), and then 0.5 f/ml between 1958 and 1966, when the mining operations ceased. Since then, interpolation between periodic hygiene surveys using personal monitors assigned exposures from 0.5 fibers (> 5 microns long) per milliliter of air (f/ml), in 1966 to 0.010 f/ml in 1992. Duration of residence for cases was extracted from detailed information kept by the Western Australia Mesothelioma Registry at the Health Department of Western Australia. For control subjects, this information was provided via a questionnaire either completed at the vitamin A program, or sent to the subject. If a questionnaire had not been completed and that subject was related to a Wittenoom asbestos worker, then dates of residence was assumed to be identical to those of the worker. For other subjects, each family member was assumed to have the same length of residence as each other, providing at least one member of the family had known exposure. For all other residents, dates of Wittenoom residence were taken as those found on various sources when the cohort was established. Duration of residence was then combined with intensity of exposure to give a measure of cumulative exposure for each person in the cohort in fibers per milliliter years (f/mly). Cumulative exposure levels were then adjusted by a factor of 4.2 to account for 24 h a day, 7 d a week exposure.

Details about Wittenoom residence for nonresponders was obtained from information previously collected. The dates of residence for subjects related to a Wittenoom crocidolite worker were taken to be the same as the dates of employment for that worker. For those not related to a Wittenoom crocidolite worker, length of exposure was taken to be the same for each family member, providing at least one family member had completed a questionnaire. Subjects who did not fit these two categories had their dates of residence determined from the various sources used to assemble the cohort (19). Where duration of residence was still unknown a value of 6 mo was assigned. These subjects were thought to have had short exposure, as information on their exposure was limited to only one source. As described previously, those who had longer exposure were found repeatedly in the same source, or in more than one source (19).

Mesothelioma incidence rates were standardized to the World Population to enable comparison with other reported mesothelioma incidence rates. Cox regression (22) was used to examine the separate and combined effect of age, sex, time from first exposure, estimated levels of exposure and age at first exposure on the incidence of mesothelioma. Each subject's entry time was the date they first went to Wittenoom, and the “survival time” variable was the time from first exposure. This was not the same as length of stay at Wittenoom since most people only stayed for short periods during the 1950s and 1960s (19). Also, time from first exposure to asbestos is the most important predictor of mesothelioma incidence (5, 6, 23) and should be matched for in any analysis examining quantitative effects of exposure. Age at first exposure, sex, calendar period of first residence, length of stay at Wittenoom, place of residence in Wittenoom, estimated intensity of exposure, estimated cumulative exposure, and whether the subject lived with an asbestos worker, or washed the clothes of an asbestos worker were included as fixed covariates. Length of stay at Wittenoom and cumulative exposure were treated as time-dependent covariates as some subjects had not completed their residence at Wittenoom when cases of mesothelioma were diagnosed. Two censoring schemes were used: date of death, date of loss to follow-up, or end of follow-up (December 31, 1993), whichever occurred first; and date of death, December 31, 1993 or age 85, whichever occurred first. The reasons for using both of these schemes have been discussed previously (4). Briefly, it was considered that the first one leads to an overestimate of disease rates because of the heavy reliance on tracing on death and disease registers, and the second leads to an overestimate of person-years, particularly in the short length of stay (low exposure) groups. Thus using both schemes gives a range of likely values. The best transformation for each variable where appropriate (categorical, log, or untransformed) was assessed by linearity of trends and likelihood ratio criteria, and all variables were included in the final model if they significantly improved the fit (p < 0.05). Detailed analyses were performed to check the proportional hazards assumption of the Cox model. Standardized mortality ratios (SMRs) for all causes were also calculated using these censoring schemes.

The whereabouts of 71.4% of the cohort as at December 31, 1993 was known, and 11.8% of the cohort remained untraced from the time they left Wittenoom (Table 1). The total person-years accumulated to December 31, 1993 by this cohort was 132,986 representing an average length of follow-up of 29.5 yr. A total of 460 deaths from all causes was recorded to the end of 1993. There were significant excess deaths from all causes using data censored at the date of loss to follow-up (SMR = 1.48 95% CI: 1.34–1.64), but not when the data was censored at December 31, 1993 (SMR = 0.96 95% CI: 0.87–1.06).


Follow-up StatusNumber (%)
Returned a questionnaire2,173 (46.6)
On vitamin A program 641 (13.8)
Dead 460 (9.9)
Permanent departure overseas  51 (1.1)
Questionnaire not returned but traced to an address 785 (16.8)
Whereabouts unknown since leaving Wittenoom 549 (11.8)
Cohort total4,659 (100)

Of the full cohort of 4,659 people, 151 indicated that they had occupational exposure to crocidolite at Wittenoom and were excluded from the analysis. To the end of 1993, 27 cases of malignant pleural mesothelioma had been diagnosed in this cohort and histologically confirmed by the Western Australia Mesothelioma Registry or the Australian Mesothelioma Surveillance Programme. A further five cases of mesothelioma arose in this cohort in subjects who had worked with asbestos at Wittenoom, and were not employed by ABA. Sixteen of the 27 cases (59%) had occurred since 1989. Another 17 cases of malignant mesothelioma were reported to the Western Australia Mesothelioma Registry who claimed to have lived at Wittenoom for varying lengths of time. They have not been included in this study because their residence at Wittenoom was not independently established by the methods described previously (19). In fact, examination of records revealed that all 17 cases resided in Wittenoom less than 1 mo, and worked with asbestos while they were there, and were therefore ineligible for inclusion in this cohort.

Of the 18 female and 9 male subjects, 12 were wives, 11 were children, and one was a brother of men who worked with crocidolite at Wittenoom, so that these 24 cases arose in people who also experienced so-called “domestic” exposure. So far, only one case has been diagnosed in a resident who first went to Wittenoom after the mining operations ceased. Age at first residence at Wittenoom ranged from birth to 51 yr (median 18.2 yr) and time from first residence to diagnosis from 20 to 45 yr (median 35.9 yr) (Table 2). The median period of survival from diagnosis was 7 mo with a range of a few days to 26 mo. The median duration of residence for cases of mesothelioma was 60 mo, with a range of 2 mo to 17 yr. Estimated intensity of crocidolite exposure ranged from 0.15 f/ml to 1.0 f/ ml, with a median value of 0.9 f/ml. Median cumulative exposure was 15.2 f/mly, with a range of 0.53 f/mly to 40.4 f/mly (Table 3). When compared with the rest of the cohort, mesothelioma cases stayed longer at Wittenoom (mean stay 64.8 mo versus 32.8 mo, p < 0.001), had a higher estimated intensity of exposure (mean value 0.8 f/ml compared with 0.5 f/ml; p < 0.001), and a higher cumulative exposure to crocidolite (mean 16.3 f/mly versus 5.4 f/mly; p < 0.001). Nine of the 27 subjects (33%) were younger than 40 yr of age at the time of diagnosis.


CharacteristicSubjects (n = 27) (%)Control Subjects (n = 4,481) (%)
Family Group
 Relatives of ABA workers21 (77.8)2,586 (57.7)
 Other workers2 (7.4)   703 (15.7)
 Relatives of other workers 4 (14.8)1,192 (26.6)
 Male 9 (33.3)1,913 (42.7)
 Female18 (66.7)2,568 (57.3)
 Head of household2 (7.4)   703 (15.7)
 Spouse11 (40.7)1,183 (26.4)
 Children11 (40.7)2,393 (53.4)
 Other relatives and friends 3 (11.1)   202 (4.5)
Year of first residence at Wittenoom
 1943–50 9 (33.3)   300 (6.7)
 1951–5712 (44.5)1,008 (22.5)
 1958–66 5 (18.5)1,905 (42.5)
 1967–931 (3.7)1,268 (28.3)
Lived with an asbestos worker
 Yes24 (88.9)2,758 (66.1)
 No 3 (11.1)1,723 (38.4)
Age at first residence at Wittenoom
 Born there2 (7.4)   394 (8.8)
 Under 1 yr 3 (11.1)   363 (8.1)
 1–10 yr 8 (29.6)1,420 (31.7)
 11–20 yr2 (7.4)   435 (9.7)
 21+ yr12 (44.5)1,869 (41.7)
Time since first residence at Wittenoom*
 Under 20 yr0 (0.0)   560 (12.5)
 20–29 yr 5 (18.5)1,380 (30.8)
 30–39 yr16 (59.3)1,882 (42.0)
 40+ yr 6 (22.2)   659 (14.7)

*Time from first residence to December 31, 1993 or death.


Measure of ExposureSubjects n (%)Control Subjects n (%)
Duration of residence, months
 1–11 3 (11.1)1,685 (37.6)
 12–23 3 (11.1) 878 (19.6)
 24–351 (3.7) 556 (12.4)
 36–59 6 (22.2) 677 (15.1)
 60+14 (51.9) 685 (15.3)
Estimated intensity of exposure, f/ml
 ⩽ 0.5 7 (25.9)3,128 (69.8)
 0.501–0.75 3 (11.1)387 (8.6)
 > 0.7517 (63.0) 966 (21.6)
Estimated cumulative exposure, f/mly
 ⩽ 7.00 7 (25.9)3,423 (76.4)
 7.01–20.0011 (40.7) 820 (18.3)
 > 20 9 (33.3)238 (5.3)

When data were censored at the date of loss to follow-up, the standardized incidence of mesothelioma of people aged 15 and over to 1993 in this cohort was 260 per million person-years (pmpy) and was similar for males and females (236 and 255 pmpy, respectively) (Figure 1). The rate rose from 207 pmpy 20–29 yr since first exposure, to 1,073 pmpy 30–39 yr since first exposure and 1,649 pmpy 40 or more years since first exposure. Within each category of time since first exposure, the rate increased with both duration of exposure (Figure 2), and estimated cumulative exposure (Figure 3), albeit not consistently.

Over all times from first exposure, the incidence of mesothelioma rose from 80 pmpy for those who lived at Wittenoom for less than one year, to 181 pmpy for residents with duration of residence between 1 and 5 yr, to 904 pmpy for those who lived at Wittenoom longer than 5 yr. The incidence of mesothelioma was 106 pmpy for subjects whose estimated cumulative exposure was less than 7 f/mly, rising to 367 and 1,953 pmpy for estimated cumulative exposure of 7–20 f/mly and over 20 f/mly, respectively.

In all regression models, age at first exposure, sex, year of first exposure, place of residence in Wittenoom, and whether a subject lived with an asbestos worker or washed work clothes had no significant effect on incidence of mesothelioma. Both length of stay at Wittenoom and estimated cumulative exposure were significantly associated with an increased incidence of mesothelioma. This result was true whether either measure of exposure was included in the model as a grouped or continuous variable. The association of incidence rate with estimated intensity of exposure was not significant (Table 4). None of the other variables significantly (p > 0.05) affected the rate of mesothelioma.


VariableCensored at Date of Loss to Follow-upCensored at Dec. 31, 1993
RR95% CIRR95% CI
Length of stay at Wittenoom, months
 Continuousper log month1.61.1––2.5
60+6.7 2.0–22.26.2 2.6–14.6
Length of stay at Wittenoom (months) and intensity of exposure (f/ml)
 Continuousper log month1.61.1––2.6
60+6.1 1.7–21.48.1 2.4–27.0
 Intensityf/ml0.60.1–4.60.9 0.1–6.8
Cumulative exposure, f/ml
 Continuousper log (f/ml)1.591.09–2.331.791.22–2.60

There was no significant effect of age at first exposure when treated as a continuous variable on mesothelioma incidence, after adjusting for all other variables (RR = 1.02 [95% CI: 0.99–1.04]), and those first exposed as children under 10 yr had a lower rate of mesothelioma than subjects first exposed after that age (RR = 0.7 [95% CI: 0.3–1.5]).

This cohort study of Wittenoom residents shows that the incidence of mesothelioma increased significantly with increasing time following first residence at Wittenoom and with increased level of exposure to crocidolite. This result holds whether level of exposure is measured by duration of residence or by cumulative exposure. The incidence of mesothelioma increased from about 210 per million person-years (pmpy) at 20–29 yr since first exposure, to over 1,600 pmpy at 40 or more years from first exposure. The corresponding figures for the Wittenoom workers' cohort were approximately 900 and 7,000 pmpy (23). No figures are available from studies of other environmentally exposed subjects, but in other occupational cohorts, rates typically have ranged from about 1,000 pmpy after 20 yr from first exposure to over 5,000 pmpy after 40 yr (23).

This is the first published study to show exposure-response relationships between incidence of mesothelioma and environmental exposure to any form of asbestos. The study of a birth cohort of exposed subjects near South African crocidolite mines has been unable to determine the intensity and duration of exposure for members of the cohort (8).

Evidence of an exposure-response relationship is also apparent in this cohort when relative risks of mesothelioma are obtained from Cox Regression. When both duration and estimated intensity of exposure were included in the model together, either as categorical or continuous variables, the association between rate of mesothelioma and estimated intensity of exposure was not significant, but showed some evidence of a trend. This result was not unexpected as intensity of exposure was measured with much more error than duration of residence (21). The model with exposure assessed by just duration of residence appeared to be a better fit than the one which used estimated cumulative exposure, probably because of the large error introduced into the estimate of cumulative exposure by the estimate of intensity of exposure.

The Wittenoom residents aged 15+ yr in this study experienced a standardized incidence rate of mesothelioma of around 260 pmpy. This rate is substantially higher than the Western Australian rate of 50 pmpy for men and 8 pmpy for women in 1988 (23), and is one of the highest population incidence rates in the world. It is similar to that found among Australian Aboriginal residents of the Pilbara region of Western Australia, who experienced a crude rate of 250 pmpy (25), but with a much higher age-standardized rate. Other studies of environmental asbestos exposure have reported high rates among specific populations, including residents near an asbestos-cement factory in Italy who experienced age-standardized incidence rates for males and females of 114 and 73 pmpy, respectively (15). A pilot study of an asbestos-exposed birth cohort in a crocidolite-mining area of South Africa has also reported high rates for both men and women (8). They reported an equivalent crude mortality rate of mesothelioma of approximately 215 pmpy (based on a life expectancy of 70 yr).

The exposure-response relationships demonstrated in this study take account of neighborhood exposure only. Domestic exposure, from living with an asbestos worker and/or washing the work clothes, was taken into account in the survival analysis via two binary variables, but found to have no significant effect, additional to cumulative exposure, on the risk of being diagnosed with mesothelioma. Measurement of the level of exposure from domestic sources would be difficult, even though exposure levels are known for each ABA asbestos worker (4). In other studies, higher concentrations of chrysotile were found in the homes of mine and mill workers than in neighboring homes of nonminers (26). Few quantitative data are available regarding the actual respirable dose in such settings. Huncharek and coworkers (27) suggested that the total amount of inhaled fibers in domestic washing of work clothes may be comparable with that inhaled by asbestos workers, although this statement was based on one case of mesothelioma arising after domestic exposure. So far, 24 of the 27 cases of mesothelioma have arisen in people who lived in the same house as a Wittenoom crocidolite worker. However, subjects experiencing this domestic exposure also had higher levels of neighborhood exposure, and were first exposed to crocidolite between 27 and 45 yr before diagnosis. It is therefore not surprising that there was no association between mesothelioma incidence and “domestic” exposure after adjusting for “neighborhood” exposure. Since no cases of mesothelioma in this cohort have arisen within 20 yr of first exposure, many potential cases of mesothelioma exist, especially considering that 16 of the 27 cases have been diagnosed in the last 5 yr.

Most population studies of mesothelioma report differences in the incidence rates for men and women with the rates for males being typically five to nine times greater than those for females. This is thought to be due to a much greater proportion of male cases having prior occupational exposure to asbestos than females (28) and this suggestion is confirmed here. Females are more likely to obtain their exposure environmentally. Similar mesothelioma incidence rates for males and females shown in this study reflect the nature of exposure: both males and females were exposed to crocidolite by living in Wittenoom, and men with known occupational exposure to crocidolite at Wittenoom were excluded from the analysis.

The question of whether asbestos exposure in childhood increases susceptibility to carcinogenesis, due to such factors as stem cell expansion while still growing, or different patterns of deposition and clearance of fibers in the immature lungs of infants, has not been addressed before (29). Among Wittenoom residents, exposure to crocidolite in childhood did not increase the risk of mesothelioma in comparison with exposure in adulthood: there was no significant effect of age at first exposure on incidence after adjustment for time since first exposure and duration of residence. Mesothelioma in young people is an uncommon event with most cases occurring in the sixth and seventh decades of life (18). Of the 27 cases to arise so far in the Wittenoom Resident's cohort, nine (33%) were younger than 40 yr of age at the time of diagnosis, resulting from first exposure to crocidolite during childhood. This proportion was higher than in other environmental studies in New Caledonia and Greece, where 25% and 0% of 12 and 6 cases, respectively, were 40 yr old or younger at the time of diagnosis (30, 31) and in other studies which have reported that 5.8% (18) and approximately 17% (32) of cases were aged 40 or less at diagnosis. However, this is simply a result of larger proportions of exposed children in the Wittenoom cohort. Mesothelioma risk is approximately proportional to the fourth power of time since first exposure, and hence calculated risk increases rapidly after 30 or 40 years has elapsed since first exposure (24, 33). People first exposed as children have potentially 60 or 70 yr to experience the increasing risk, and thus are at considerably higher risk compared with people of the same age who have been exposed to the same levels later in life. In this study, those who were first exposed to crocidolite at Wittenoom as children under 10 yr had a slightly lower risk of mesothelioma compared with those whose first exposure occurred after that age.

The risk of mesothelioma in this study has been calculated using two censoring dates, namely, the date of death or loss to follow-up, and the date of death, December 31, 1993, or age 85, whichever occurred first. It is difficult to reach any firm conclusion as where the true value of mesothelioma incidence lies, and the two values could be considered the upper and lower estimates of the true risk. Even though it is unlikely that all untraced subjects are still alive, they are not likely to have been diagnosed with mesothelioma because ascertainment of cases of mesothelioma throughout Australia is thought to be very nearly complete, due to the activities of the members of our research team, the Western Australia Mesothelioma Registry and the Australian Mesothelioma Surveillance Programme. Details on death and disease in this cohort have yet to be systematically sought from states in Australia other than Western Australia, and therefore possible bias may be introduced using the data censored at December 31, 1993 as some untraced subjects may have died from other causes, or returned to their country of origin.

Cases of mesothelioma in this cohort of Wittenoom residents have arisen in subjects with durations of crocidolite exposure as short as 2 mo and estimated cumulative exposure as low as 0.53 f/ml. Thus exposure-response relationships have been confirmed at far lower levels than in any previous study.

This study would not have been possible without the considerable efforts of Dr. Jim Baker and Mr. Steven Woodward in collecting many of the initial paper records in 1979, and the encouragement of Dr. Jim McNulty from what was the Public Health Department of Western Australia. We are also grateful to Mrs. J. Eccles, Ms. L. Watts, Ms. C. Westbury, and Mrs. J. Sleith for assistance in data collection, and members of the Asbestos Diseases Society of W.A. (Inc.) for their cooperation.

This study was funded by a grant from the State Government Insurance Commission of Western Australia to the Sir Charles Gairdner Hospital Research Foundation, and the Public Health Research and Development Committee of the Australian National Health and Medical Research Council.

1. Dement J. M., Brown D. P., Okun A.Follow-up study of chrysotile textile workers: cohort mortality and case-control analyses. Am. J. Ind. Med261994431447
2. McDonald J. C., Liddell F. D. K., Dufresne A., McDonald A. D.The 1891–1920 birth cohort of Quebec chrysotile mines and millers: mortality 1976–88. Br. J. Ind. Med50199310731081
3. Newhouse M. L., Sullivan K. R.A mortality study of workers manufacturing friction material: 1941–86. Br. J. Ind. Med461989176179
4. Armstrong B. K., de Klerk N. H., Musk A. W., Hobbs M. S. T.Mortality in miners and millers of crocidolite in Western Australia. Br. J. Ind. Med451988513
5. de Klerk N. H., Armstrong B. K., Musk A. W., Hobbs M. S. T.Cancer mortality in relation to crocidolite at Wittenoom Gorge in Western Australia. Br. J. Ind. Med461989529536
6. Peto, J. 1985. Some problems in dose-response estimation in cancer epidemiology. In V. B. Vouk, G. C. Butler, D. G. Hoel, and D. B. Peakell, editors. Methods for Estimating Risk of Chemical Injury: Human and Non-human Biota and Ecosystems. Wiley, New York. 361–380.
7. Joubert L., Seidman H., Selikoff I. J.Mortality experience of family contacts of asbestos factory workers. Ann. N.Y. Acad. Sci6431991416418
8. Reid G., Keilkowski D., Steyn S. D., Botha K.Mortality of an asbestos-exposed birth cohort: a pilot study. South Afr. Med. J781990584586
9. Wagner J. C., Sleggs C. A., Marchand P.Diffuse pleural mesothelioma and asbestos exposure in the North West Cape Province. Br. J. Ind. Med171960260271
10. Vianna J. C., Polan A. K.Non-occupational exposure to asbestos in malignant mesothelioma in females. Lanceti197810611067
11. Hammond E. C., Garfinkel L., Selikoff I. J., Nicholson W. J.Mortality experience of residents in the neighborhood of an asbestos factory. Ann. N.Y. Acad. Sci3301979417422
12. Newhouse M. L., Thompson H.Mesothelioma of pleura and peritoneum following exposure to asbestos in the London area. Br. J. Ind. Med221965261269
13. Dodoli D., Del Nevo M., Fiumalbi C., Iaia T. E., Cristaudo A., Comba P., Viti C., Battista G.Environmental household exposure to asbestos and occurrence of pleural mesothelioma. Am. J. Ind. Med211992681687
14. Siemiatycki, J. 1982. Health effects on the general population: mortality in the general population in asbestos mining areas. In Proceedings of the World Symposium on Asbestos. Canadian Asbestos Information Centre, Montreal. 337–348.
15. Magnani C., Bellis D., Borgo G., Botta M., Ivaldi C., Mollo F., Terracini B.Incidence of mesotheliomas among people environmentally exposed to asbestos. Eur. Respir. Rev31993105107
16. Anderson, H. A. 1983. Family contact exposure. In Proceedings of the World Symposium on Asbestos. Canadian Asbestos Information Centre, Montreal. 349–362.
17. Inase J., Takayama S., Nakayama M., Miura H., Kimula Y.Pleural mesothelioma after neighbourhood exposure to asbestos during childhood. Jpn. J. Med301991343345
18. Kane M. J., Chahinian A. P., Holland J. F.Malignant mesothelioma in young adults. Cancer65199014491455
19. Hansen J., de Klerk N. H., Eccles J., Musk A. W., Hobbs M. S. T.Malignant mesothelioma after environmental exposure to blue asbestos. Int. J. Cancer541993578581
20. Wajda A., Roos L. L.Simplifying record linkage: software and strategy. Comput. Biol. Med171987239248
21. Hansen J., de Klerk N. H., Musk A. W., Hobbs M. S. T.Individual exposure levels in people environmentally exposed to crocidolite. Appl. Occup. Environ. Health121997483490
22. Cox, D. R. 1972. Regression models and life-tables. J. R. Stat. Soc. B; 34: 187–220.
23. de Klerk, N. H., and B. K. Armstrong. 1992. The epidemiology of asbestos and mesothelioma. In D. W. Henderson, K. B. Shilkin, D. Whitaker, and S. L. P. Langlois, editors. Malignant Mesothelioma. Hemisphere, New York. 223–250.
24. de Klerk, N. H., A. W. Musk, B. K. Armstrong, and M. S. T. Hobbs. 1994. Disease in miners and millers of crocidolite from Wittenoom, Western Australia: a further follow-up to December, 1986. Ann. Occup. Hyg. 38(Suppl. 1):647–655.
25. Musk A. W., de Klerk N. H., Eccles J. L., Hansen J., Hobbs M. S. T.Mesothelioma in Pilbara Aborigines. Aust. J. Publ. Health191995520522
26. Nicholson, W. J., A. N. Rohl, I. Weisman, and I. J. Selikoff. 1980. Environmental asbestos concentrations in the United States. In J. C. Wagner, editor. Biological Effects of Mineral Fibres. IARC Scientific Publications, No. 30. IARC, Lyon. 823–827.
27. Huncharek M., Capotorto J. V., Muscat Y.Domestic exposure, lung fibre burden and pleural mesothelioma in a housewife. Br. J. Ind. Med461989354355
28. Leigh J., Corvalan C. F., Grimwood A., Berry G., Ferguson D. A., Thompson R.The incidence of malignant mesothelioma in Australia 1982–1988. Am. J. Ind. Med201991643655
29. Peto J.Fibre carcinogenesis and environmental hazards. IARC Sci. Publ901989457470
30. Goldberg P., Goldberg M., Marne M. J.Incidence of pleural mesothelioma in New Caledonia: a 10-year survey (1978–1987). Arch. Environ. Health461991306309
31. Langer A. M., Nolan R. P., Constantopoulos S. H., Moutsopoulos H. M.Association of Metsovo lung and pleural mesothelioma with exposure to tremolite-containing whitewash. Lanceti1987965967
32. Hillerdal G.Malignant mesothelioma 1982: review of 4710 published cases. Br. J. Dis. Chest771983321343
33. Doll, R., and J. Peto. 1985. Effects on health of exposure to asbestos. Health and Safety Commission. HMSO, London.
Correspondence and requests for reprints should be addressed to Janice Hansen, Department of Public Health, University of Western Australia, Western Australia, 6907. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record