Abstract
Follow-up of the atomic bomb survivors in Japan in the Life Span Study (LSS) has been fundamental for the understanding of the dose-response curve. We compare our risk estimates from a newly published epidemiological study on cancer in Sweden after the Chernobyl Nuclear Power Plant accident with the LSS data. Hazard ratios (HRs) with 95% confidence intervals (95% CIs) were calculated using conditional logistic regression adjusted for rural/nonrural habitat, education level, and pre-Chernobyl cancer incidence from 1980 to 1985, respectively. Adjusted HRs by sex were calculated in deciles for all cancer sites combined for 1986 to 2020. These risk estimates were translated to excess relative risk (ERR) to allow comparison with LSS incidence data. ERRs per decile were compatible with ERR in the low-dose range <100 mGy for both sexes. The CIs in each decile need to be taken into account when interpreting the dose-response curve. Risk estimates in dose categories add important information at very low doses on the dose-response curve when compared to LSS data.