External radiation exposure and mortality in a cohort of French nuclear workers.

OBJECTIVE: To analyse the effect of external radiation exposure on the mortality of French nuclear workers. METHODS: A cohort of 29 204 workers employed between 1950 and 1994 at the French Atomic Energy Commission (Commissariat à l'Energie Atomique (CEA)) or at the General Company of Nuclear Fuel (COmpagnie GEnérale des MAtières nucléaires (Cogema, now Areva NC)) was followed up for an average of 17.8 years. Standardised mortality ratios (SMRs) were computed with reference to French mortality rates. Dose-effect relationship were analysed through trend tests and Poisson regression, with linear and log-linear models. RESULTS: The mean exposure to X and gamma radiation was 8.3 mSv (16.9 mSv for exposed worker population). A total of 1842 deaths occurred between 1968 and 1994. A healthy worker effect was observed, the number of deaths in the cohort being 59% of the number expected from national mortality statistics. Among the 21 main cancer sites studied, a statistically significant excess was observed only for skin melanoma, and an excess of borderline statistical significance was observed for multiple myeloma. A dose-effect relationship was observed for leukaemia after exclusion of chronic lymphoid leukaemia (CLL). The relative risk observed for non-CLL leukaemia, n = 20, was 4.1 per 100 mSv (90% CI 1.4 to 12.2), linear model and 2.2 per 100 mSv (90% CI 1.2 to 3.3), log-linear model. Significant dose-effect relationship were also observed for causes of deaths associated with alcohol consumption: mouth and pharynx cancer, cirrhosis and alcoholic psychosis and external causes of death. CONCLUSION: The risk of leukaemia increases with increasing exposure to external radiation; this is consistent with published results on other nuclear workers cohorts.

Cancer mortality among French Atomic Energy Commission workers.

BACKGROUND: All causes and cancer mortality of 58,320 workers employed at the Commissariat à l'Energie Atomique (CEA) between 1946 and 1994 were compared with that of the general population in a retrospective cohort study. METHODS: Standardized Mortality Ratios (SMR) were computed with reference to the French national population. RESULTS: Between 1968 and 1994, 4,809 deaths occurred. A healthy worker effect is observed for men (SMR = 0.57, CI(90%) = [0.56;0.59]) and for women (SMR = 0.72, CI(90%) = [0.67;0.77]). Nine sites of cancer death were found to be in statistically significant deficit among men, none among women. An excess of pleural cancers is observed among men (SMR = 1.79, CI(90%) = [1.27;2.45]) and of malignant melanoma (SMR = 1.50, CI(90%) = [1.04;2.11]). An excess of breast cancer is observed among women on the borderline of significance (SMR = 1.14, CI(90%) = [0.94;1.37]). CONCLUSIONS: Excesses observed will have to be related to occupational exposures in the on-going cohort study on French nuclear workers which includes a retrospective exposures assessment.