Cancer Mortality in Chinese Chrysotile Asbestos Miners: Exposure-Response Relationships

Abstract

Objective: This study was conducted to assess the relationship of mortality from lung cancer and other selected causes to asbestos exposure levels. Methods: A cohort of 1539 male workers from a chrysotile mine in China was followed for 26 years. Data on vital status, occupation and smoking were collected from the mine records and individual contacts. Causes and dates of death were further verified from the local death registry. Individual cumulative fibre exposures (f-yr/ml) were estimated based on converted dust measurements and working years at specific workshops. Standardized mortality ratios (SMRs) for lung cancer, gastrointestinal (GI) cancer, all cancers and nonmalignant respiratory diseases (NMRD) stratified by employment years, estimated cumulative fibre exposures, and smoking, were calculated. Poisson models were fitted to determine exposure-response relationships between estimated fibre exposures and cause-specific mortality, adjusting for age and smoking. Results: SMRs for lung cancer increased with employment years at entry to the study, by 3.5-fold in ≥10 years and 5.3-fold in ≥20 years compared with <10 years. A similar trend was seen for NMRD. Smokers had greater mortality from all causes than nonsmokers, but the latter also had slightly increased SMR for lung cancer. No excess lung cancer mortality was observed in cumulative exposures of <20 f-yrs/ml. However, significantly increased mortality was observed in smokers at the levels of ≥20 f-yrs/ml and above, and in nonsmokers at ≥100 f-yrs/ml and above. A similarly clear gradient was also displayed for NMRD. The exposure-response relationships with lung cancer and NMRD persisted in multivariate analysis. Moreover, a clear gradient was shown in GI cancer mortality when age and smoking were adjusted for. Conclusion: There were clear exposure-response relationships in this cohort, which imply a causal link between chrysotile asbestos exposure and lung cancer and nonmalignant respiratory diseases, and possibly to gastrointestinal cancer, at least for smokers.