Cancer risks associated with protracted versus acute exposures to ionizing radiation: results from INWORKS and the Life Span Study

Background: The appropriate use of results from acute exposure settings for elaborating radiation protection guidelines in settings of protracted low dose-rate exposures remains an open question. Aim: We present a comparison of cancer mortality radiation risk estimates derived from the Life Span Study (LSS) of A-bomb survivors and INWORKS, an international cohort of nuclear workers exposed to protracted low doses of external radiations to assess similarities of dose-response associations in settings that differ with respect to exposure conditions. Methods: Restrictions have been made so that the characteristics of the two populations are similar with respect to ages and periods of exposure. Dose-response relationships based on excess relative rate (ERR) and excess absolute rate (EAR) models were fitted using Poisson regression, for both solid cancer and leukemia mortality. Results/Conclusions: Applying restriction criteria led to selecting 45,625 A-bomb survivors and 259,350 workers from INWORKS, who respectively accrued 1.48 and 6.18 million person-years of follow-up. For solid cancer, we estimated ERRs per Gray (Gy) colon dose similar in the two populations: ERR per Gy = 0.28 (90% confidence interval (CI): 0.18, 0.38) in the LSS and ERR per Gy = 0.29 (90%CI: 0.07, 0.53) in INWORKS, based on a linear ERR model. A larger EAR per Gy estimate was noted in the LSS (8.03, 90%CI: 3.74, 13.07) than in INWORKS (1.68, 90%CI: <0, 7.55), with evidence of modification of the EAR per Gy by attained age. For leukemia, ERR per Gy red marrow dose estimated from a linear model were also close in magnitude (2.75 (90%CI: 1.73, 4.21) in the LSS and 3.15 (90%CI: 1.12, 5.72) in INWORKS). These analyses offer useful comparisons of the LSS risk estimates with those derived from occupational exposures occurring in settings more comparable to those encountered in contemporary occupational and environmental settings.