Blue asbestos (crocidolite) was mined and milled by the Australian Blue Asbestos Company, a subsidiary of CSR Limited, at Wittenoom Gorge in the remote Pilbara region of Western Australia, 1600 km north of Perth, from 1943 until 1966.

A high percentage of the workforce were post war migrants, 35%–50% of migrants being non-British in origin,1 especially from southern Europe. The workforce was also young (57% were less than 30 years of age), male and transient (the median time spent working at Wittenoom was less than 4 months2). The least transient workers were the miners who worked 7.5 h shifts in cramped and confined conditions. The young age of the Wittenoom workforce has had an impact upon the subsequent incidence of asbestos related diseases, as many workers have survived to experience diseases with long latency periods.3 The first case of asbestosis from Wittenoom was diagnosed in 1946, while the first case of mesothelioma was reported in 1960.4

Follow-up of the Wittenoom workers cohort to 1986 showed excess mortality from asbestosis (SMR = 19.95), lung cancer (SMR = 2.43) and malignant mesothelioma of the pleura and peritoneum, which have all been related to duration and intensity of exposure and time since first exposure to crocidolite.5 6 The workers have also previously shown excess mortality from stomach cancer (SMR = 2.18), tuberculosis (SMR = 2.42), peptic ulcer (SMR = 2.87), hepatic cirrhosis (SMR = 2.38), injuries and poisonings (SMR = 1.51) and mental disorders (alcoholism) (SMR = 3.66). Deaths from non-transport accidents (SMR = 2.48) and ill-defined conditions (SMR = 2.54) have been increased. There has been no significant increase in deaths from laryngeal cancer or colorectal cancer as suggested from some other studies of asbestos exposed subjects.7^–11

A recent examination of mortality from malignant mesothelioma12 suggested that the exponent of incidence with time since first exposure to crocidolite has declined slightly, an observation which is consistent with the slow clearance of fibres from lung tissue; this phenomenon has also been documented in lung fibre content studies of the Wittenoom workers.13

The aim of this study was to extend the follow-up of this workforce to the end of 2000 with more complete tracing of ex-workers who have returned to Italy since 1966 in order to document the pattern of mortality with increasing time since first exposure (now as long as 60 years). Since the last reported follow-up to the end of 1986, an additional 1386 male workers from the cohort have died. This further follow-up permits better understanding of the effects of increasing time since exposure on various asbestos related disease.

METHODS

Subjects

The cohort consisted of 6943 subjects who had worked at the Wittenoom crocidolite mine and mill at any time between 1943 and 1966. These subjects were identified from employment records provided by the company in 1975, supplemented by information from records of subjects attending for compulsory chest x rays taken by the Perth Chest Clinic and subjects who had paid into a contributory fund (The Mine Workers’ Relief Fund) while they were employed in the mining industry.

Follow-up

Establishment of the cohort has been described elsewhere.2 For this follow-up, certified causes of death to the end of 2000 were obtained from the National Death Index via the Australian Institute of Health and Welfare and from the Western Australian Registrar General. The cohort was matched to deaths using surname, first name, middle name, previous name (if any), date of birth and current address. A total of 810 subjects known to have departed from Australia for Italy have been followed up there for cancer and death registrations and are included in the analysis.14

During the period of its operation from 1943 to 1966, fewer than 500 female workers were employed by the Australian Blue Asbestos Company and follow-up has therefore been restricted to the 6498 male workers known from employment records provided by the company in 1975 to have worked in the mining and milling of crocidolite.

Asbestos exposure assessment has been fully described elsewhere.2 13 15 This paper does not examine exposure–response relationships.

Analysis

Standardised mortality ratios (SMRs) were calculated from the ratio of observed deaths to expected deaths by the person-years method using 5-year age, period, cause and sex-specific death rates for Western Australia. As in our previous follow-up studies, the numbers of person-years for estimating expected deaths were counted in two ways: first by assuming that all subjects who were lost to follow-up were still alive on 31 December 2000 or had died at 85 years of age (SMR1), and second after censoring all subjects at the date last known to be alive from searches of electoral rolls, death registers, cancer registries, mail contact, hospital morbidity records and attendance at a cancer prevention program (SMR2).16 17 For both SMR1 and SMR2 the follow-up was censored at 85 years of age.

RESULTS

Follow-up rates in this cohort have increased since previous analyses as a result of searches performed in Italy and because increased numbers of previous workers have been identified through death or disease records or family contacts elicited through a cancer prevention program which commenced in 1990.16 Follow-up of the workforce now exceeds 75%.

There is an overall increased rate of death in this cohort, which is probably between 13% and 90% higher than in the general Australian population (table 1). Much of this increase appears to be due to diseases specifically known to be associated with asbestos exposure including mesothelioma, lung cancer and asbestosis. In addition, there are also increases in deaths known to be associated with mining occupations (non-transport accidents), cigarette smoking (lung cancer, bronchitis and emphysema, laryngeal cancer) and social and geographical isolation (alcohol related diseases including liver cirrhosis and suicide). In addition there are increased rates of tuberculosis and the symptoms and signs of ill-defined conditions.

The pattern of mortality is similar to that previously reported (table 1): there is an increase in the proportion of deaths resulting from malignant mesothelioma of the pleura or peritoneum. There has also been an increase in the proportion of peritoneal mesothelioma cases compared to pleural mesothelioma.18

A number of cancers other than those traditionally associated with asbestos showed an increase. As in our previous follow-up studies, there was a borderline increase in mortality from upper aerodigestive and oesophageal cancers. There was also a slightly increased risk of mortality from stomach cancer.

DISCUSSION

This study has utilised data from more than 14 extra years of complete follow-up, with a further 1386 deaths since the previous mortality analysis of the Wittenoom male workers cohort.6

In this study we have estimated a lower limit of SMRs by calculating person-years assuming that those lost to follow-up were alive unless they had reached 85 years of age (SMR1), and a higher limit by censoring all subjects at the date last known to be alive (SMR2). The “true” SMR lies somewhere between these two estimates.

The number of mesothelioma cases has continued to rise in this cohort. This can be attributed to the exponential increase in risk with increasing time since first exposure to asbestos documented for this workforce and by others. This increase has continued despite the institution of a cancer prevention program in the workforce, the results of which have previously been consistent with a protective effect of dietary supplements with retinol (but not beta carotene).17 The increase in the proportion of peritoneal mesothelioma cases compared to pleural mesothelioma is consistent with known differences in their dose and time relationships to asbestos exposure, with peritoneal mesotheliomas seeming to have a longer latency period.18

The maturity of this cohort with extension of the follow-up period has permitted much more precise estimates of exposure–response relationships for crocidolite.12 The cohort is unique in that it comprises an exposed population who had short-term (median duration 3 months) intense exposure exclusively to crocidolite in measured amounts commencing over 60 years ago, allowing for the effect of time since exposure to be minimally affected by duration of exposure. Few of the subjects had previous exposure to any asbestos before or following their exposure at Wittenoom. This has enabled us to disentangle some of the competing effects of other factors in the workforce, particularly smoking, alcohol, country of birth and social factors.

The borderline increases in mortality from upper aerodigestive (lip, tongue, mouth, pharynx and larynx), oesophageal and stomach cancers in this cohort do not appear to be related to indices of crocidolite exposure. With improved follow-up rates and increasing time since exposure, the strength of the dose–response relationships for crocidolite has not increased. Our previous study in a subset of the total cohort on whom smoking information was obtained in and after 1979, demonstrated a significant relationship of upper aerodigestive cancers but not stomach cancer incidence with tobacco smoking,19 stomach cancer mortality having previously been associated with country of birth in this cohort.20

Because this cohort had a relatively intense exposure of short duration, it provides an important resource for dissecting the relative effects of duration and intensity of crocidolite exposure and because of the existence of a parallel cohort of Wittenoom township residents also with exposure estimates,21 it will continue to provide information on dose effects over a wide range of exposure levels in the future. This is especially important for developing countries in South America, sub Saharan Africa and South East Asia where crocidolite is still used and present in the working and general environment. The largest producers today of any asbestos type are Russia, China, Canada, Kazakhstan, Brazil and Zimbabwe.22

Conclusion

Asbestos related diseases (malignant mesothelioma, lung cancer and pneumoconiosis) continue to be the main causes of mortality among this cohort of male former Wittenoom crocidolite miners and millers, followed up to the end of 2000. As this cohort continues to mature, it remains important to keep recording their mortality and cancer incidence, to document the legacy of Australia’s blue asbestos venture at Wittenoom.