A comparison of the risks of leukaemia in the offspring of the Japanese bomb survivors and those of the Sellafield workforce with those in the offspring of the Ontario and Scottish workforces

Abstract

The cases of childhood leukaemia found among children of the Sellafield (West Cumbria), Ontario and Scottish radiation workers and those observed in the offspring of the Japanese bomb survivors are analysed using exponential and linear forms of a relative risk model and employing dose estimates both for the period 6 months before conception and for total pre-conception doses. Because of the uncertainties that surround the neutron dose estimates in the Hiroshima bomb survivors, analyses of the possible discrepancies between the pre-conception exposure risks of leukaemia in the Japanese and other datasets are performed separately for each city within the Japanese F1 cohort. In general, the discrepancy between the leukaemia relative risk coefficients in the Japanese and West Cumbrian datasets is strongest for Hiroshima, although there are also indications of differences in risks between the West Cumbrian and Nagasaki data. The leukaemia relative risk coefficients for paternal (whole-body) exposure in these pre-conception periods for children in the Canadian and Scottish groups are found to be statistically compatible with the (gonadal dose) coefficients derived from studies of the offspring of the Japanese bomb survivors, whether for each city considered separately or for both cities taken together. There are indications at borderline levels of statistical significance that the leukaemia risks observed in the Canadian and Scottish datasets are not compatible with those derived from the West Cumbrian group, whether for 6-month or total pre-conception dose. Given the lack of a discernibly raised risk of childhood leukaemia in the offspring of the Japanese bomb survivors and the Canadian and Scottish studies, it therefore appears unlikely that the statistical association between recorded paternal pre-conception external radiation and the raised incidence of childhood leukaemia found in the West Cumbrian study represents a causal relationship.