Epidemiological study of mortality among veterans of French nuclear tests in the Pacific (1966-1996)

France conducted 193 nuclear tests from 1966 to 1996. The mortality in a cohort of male veterans (n=26,514) who were monitored for external radiation exposure during their participation in the tests has been studied for the period from 1966 to 2015. The mortality was compared to those of the French population using standardized mortality ratios (SMRs), then estimated according to dosimetry status using categorical Poisson models adjusted for attained age, calendar period, and military unit. “Non-null dosimetry” (at least one positive dosimetric result) was reported for 8% of the veterans. In total, 8,094 deaths were recorded. We observed sub-mortality in the cohort relative to the French population for all causes and all malignant tumors. Mesothelioma was the only pathology for which a significantly high SMR was observed. There were non-significant excesses in the 'non-null dosimetry' group relative to the 'null dosimetry' group for deaths related to mesothelioma and thyroid, central nervous system, bone, and hematological malignancies. A sensitivity analysis performed on veterans (after the exclusion of health services) who wore their first dosimeter during aerial tests highlighted significant excess mortality for bone tumors, hematological malignancies and non-Hodgkin lymphomas. Further analyses should consider extended follow-up and include ionizing radiation doses and the dates of exposure.