Absence of evidence for differences in the dose-response for cancer and non-cancer endpoints by acute injury status in the Japanese atomic-bomb survivors.

PURPOSES: To ascertain possible evidence for radiosensitive subgroups in the Japanese atomic-bomb survivors.

MATERIALS AND METHODS: Time- and age-adjusted relative risk models were fitted to the Japanese atomic-bomb survivor mortality data, taking account of the modifying effects of acute injury status and adjusting for random dosimetric errors.

RESULTS: For leukaemia, there were statistically significantly higher relative risks among those persons reporting either epilation (two-sided p=0.03) or burns (two-sided p=0.02), but after adjustment is made for the effects of dosimetric error these results become statistically non-significant (two-sided p>0.05) For all other endpoints (cancers other than leukaemia, benign neoplasms, cardiovascular disease, non-cancer and non-cardiovascular disease) there were no statistically significant modifying effects on relative risk of acute injury status, whether or not adjustment is made for the effects of dosimetric error.

CONCLUSIONS: Apart from leukaemia, there are generally no indications for cancer or other disease endpoints of variations in dose-response by acute injury status. The increased risks observed for leukaemia among those reporting epilation or burns are much diminished after adjustment for dosimetric error, and so there is no strong evidence of a sensitive subgroup within the Life Span Study cohort.