Abstract
This paper concludes the third comprehensive report on radiation effects on solid cancer incidence within the Life Span Study (LSS) cohort of Japanese atomic bomb survivors, adding 11 years of follow-up. Over 1958-2009, 22 538 solid cancer cases and 3.1 million person-years were identified among 105 444 individuals. The study utilized improved radiation doses (DS02R1), updated migration probabilities, and adjustments for smoking and lifestyle factors. Poisson regression was used to model excess relative risk (ERR) and excess absolute rate (EAR) per 1 Gy for all-solid cancers combined and specific sites. Among females, a linear dose-response model best described all-solid cancer risk [ERR = 0.64 Gy-1, 95% confidence interval (CI): 0.52-0.77], whereas among males, a linear-quadratic model with upward curvature provided the best fit (ERR = 0.20 at 1 Gy, 95% CI: 0.12-0.28). Radiation-associated ERR declined with attained age, more rapidly in males than in females, while EAR increased with age. Independently, radiation-associated ERR and EAR decreased with age at exposure. A significant dose-response was identified for prostate cancer for the first time; for brain and central nervous system tumors, female pancreatic, and uterine corpus cancers dose-response reached statistical significance. A new age-at-exposure pattern emerged for female breast and uterine corpus cancers, suggesting that radiation-associated risk increased as exposure age approached menarche (∼age 15), decreasing at older exposure ages. Most cancers exhibited linear dose-response, except non-melanoma skin, bone and joint, esophageal, and kidney parenchyma cancers in males. Sixty-four years post-exposure, solid cancer risks remain elevated, reinforcing the LSS's critical role in unraveling lifetime cancer effects of a-bomb radiation after exposure at various ages.