An analysis of leukaemia, lymphoma and other malignancies together with certain categories of non-cancer mortality in the first generation offspring (F₁) of the Japanese bomb survivors

The absence of any significant excess of childhood leukaemia in the offspring of the Japanese bomb survivors has provided strong evidence against a causal interpretation of the association between paternal preconception radiation dose and the incidence of childhood leukaemia in the offspring of employees at the Sellafield nuclear installation, West Cumbria. The use of the Japanese data has, however, been questioned on the basis that leukaemia cases may have been under-recorded in the years immediately following the bombings. In order to investigate possible misdiagnoses of leukaemia cases in the first generation offspring (F₁) of the Japanese bomb survivors, analyses have been undertaken of cases of leukaemia, non-Hodgkin's lymphoma (NHL), and all other malignancies, together with deaths due to blood diseases, deaths due to infectious diseases and deaths from unknown causes. There are some indications of an increasing trend with paternal preconception dose for NHL, but the significance of this effect is almost entirely accounted for by a single case (diagnosed at the age of 6 years) of reticulum cell sarcoma in a female child born in 1966 in Nagasaki. In particular, no positive trend for NHL has been discerned in those born within 14 years of the bombings (May 1946-December 1958), so that this trend most probably does not represent the effect of paternal preconception irradiation. There are also indications of a positive trend with maternal gonadal dose for all malignancies other than leukaemia and NHL among those born before 1951. Again, the effect is entirely concentrated in one of the cities (Hiroshima), in only one of the sexes (males), and is not apparent in the larger number of births from 1951 onwards, so that again this most probably does not represent the effect of maternal preconception exposure. No significant trend with gonadal dose is found for leukaemia and NHL combined (or for leukaemia alone). There are no indications of any stronger trends with dose in any of these categories of malignancy if those offspring of parents with higher gonadal doses are not included in the regressions. In none of the mortality categories 'deaths due to blood diseases', 'deaths due to infectious diseases' and 'deaths due to unknown causes' is there any indication of an increasing trend with paternal, maternal or conjoint parental dose, so that taking these results together with previous analyses (Little 1992) there is little evidence for increased risks of any childhood cancer type being associated with parental preconception dose, or that leukaemia cases in the exposed Japanese F₁ cohort might have been preferentially 'missed'.