Abstract
Since 1990 we have been conducting a cohort study on nuclear industry workers in Japan using the exposure dose record for each worker filed at the Radiation Dose Registration Center (RADREC) and resident record cards transcribed at municipal offices. The causes of death were identified by record linkage with national vital statistics death records. A prospective follow-up during 1991-2007 was successfully made for 203,904 males, and there were 14,224 deaths from all causes and 5,711 deaths from all cancers. The mean cumulative dose of this population was 13.3mSv. The SMR (95% CI) was 1.00 (0.84-1.18) for leukemia excluding CLL, 1.04(1.01-1.07) for all cancers, 1.13 (1.06-1.21) for liver cancer and 1.08 (1.02-1.14) for lung cancer. In the trend analyses, while mortality for leukemia excluding CLL did not show a significant increase with cumulative doses (p=0.841), that for all cancers excluding leukemia did (p=0.024). There was a positive correlation with doses for esophageal cancer (p=0.039), for liver cancer (p=0.025), for lung cancer (p=0.007),for non-Hogkin lymphoma (p=0.028) and for multiple myeloma (p=0.032). The lifestyle survey separately carried out with approximately 80,000 workers from the nested cohort indicated that the habits of heavier smoking and drinking were positively correlated with cumulative doses. With very likely involvement of potentially confounding factors taken into account, this study did not prove that low levels of radiation increased the mortality for cancer.
