Cataract in atomic bomb survivors based on a threshold model and the occurrence of severe epilation

This report re-examines the relationship of radiation dose to the occurrence of cataracts among 1742 atomic bomb survivors seen in the years 1963-1964 for whom the degree of epilation and Dosimetry System 1986 (DS86) doses are known. Of these individuals, 67 had cataracts. A relative risk model with two thresholds, one for the epilation group and the other for the no-epilation group, has been fitted to the data using a binomial odds regression approach and a constant relative biological effectiveness (RBE) for neutrons of 10. Among these models, a linear-linear (L-L) dose-response relationship with two thresholds presents the best fit. Under the L-L threshold model based on DS86 eye organ dose estimates for the epilation and no-epilation groups, the slope estimate for the epilation group was 1.6-2.0 times greater than that for the no-epilation group, but no statistical difference between the two slope estimates was noted. The estimated threshold for the epilation group was 0.86 Sv and 1.54 Sv for the no-epilation group, but again the difference between the two threshold estimates is not statistically significant. When an L-L relative risk model with two thresholds was fitted to the data assuming the dose estimates to be in error by 35%, or when the data were restricted to the 1105 individuals exposed in Japanese houses at distances of less than 2500 m, where the DS86 doses are thought to be most reliable, the results were almost the same as those for the individuals for whom unadjusted DS86 eye organ dose estimates were used.