MORTALITY OF A COHORT OF FRENCH URANIUM MINERS EXPOSED TO RELATIVELY LOW RADON CONCENTRATIONS

A cohort mortality study has been performed on French uranium miners having experienced more than 2 years of underground mining, with first radon exposure between 1946 and 1972. Vital status has been ascertained from the date of entry to the 31 December 1985 for 99% of the members of this cohort; causes of death are identified for 95.5% of the decedents. The different causes of death are compared to the age specific national death rates by indirect standardisation and expressed by standardised mortality ratios (SMR). A statistically significant excess has been observed for lung and laryngeal cancer deaths. The Poisson trend test shows a statistically significant trend for the risk of lung cancer death as a function of cumulative radon exposure, assuming a lag time of 5 years; for laryngeal cancer no significant trend has been observed. Poisson regression modelling has been applied to the following exposure groups: < 10 WLM (Working Level Month); 10-49 WLM; 50-149 WLM; 150-299 WLM; greater-than-or-equal-to 300 WLM; it indicates an increase in the SMR for lung cancer of 0.6% per WLM (standard error: 0.4%) with an estimated intercept at 0 WLM of 1.68 (standard error: 0.4). The distinction of two working periods, differing by their annual radon concentration (before/since 1956) does not modify this exposure-response relationship. This coefficient of risk per unit of exposure is lower than in most of the other uranium miners' studies but it lies in the range of the evaluation of the ICRP 50 committee and the 'BEIR IV' report of the U.S. National Academy of Science. It is observed in a cohort having experienced low cumulative exposure to radon (mean: 70 WLM) spread over a mean duration of 14.5 years. Even though occupational exposure in mines differs in several particulars from domestic exposure, this study presents characteristics of low annual exposure comparable to radon gas concentrations in houses of 500-1000 Bq.m -3, and will contribute to the evaluation of cancer risk for the public.