Investigation of stillbirth risk among the offspring of male radiation workers at the Sellafield nuclear installation, west Cumbria

Ionising radiation is a known mutagen, much of the evidence coming from studies on animals, and from studies of occupationally exposed workforces and those exposed after the atomic bombings of Hiroshima and Nagasaki. However, few studies have examined transgenerational effects of paternal exposure in humans. The workforce at the Sellafield nuclear reprocessing plant in west Cumbria, is the most highly exposed such workforce in Western Europe and North America. This epidemiological investigation set out to determine whether there was evidence of an association between stillbirth risk and paternal preconceptional exposure to external ionising radiation or internal radionuclides. Logistic regression was used to analyse the relationship between stillbirth risk and paternal preconceptional irradiation (ppi). Further modelling of the risk of stillbirth in relation to ppi was carried out using overdispersion and variance components models. The form of doseresponse was investigated using both threshold and broken-stick models, in addition to varying the power function of the dose term. To assess the effect of errors in the dose estimates for the immediate preconceptional period resulting from the derivation of doses for short intervals pro rata from annual dose summaries, a nested case-control study was also carried out. This used more precise estimates of external doses estimated directly from monthly film badges and internal doses from special dose assessments, which were very time consuming to collate and would have taken several years to compile for the entire workforce. A significant positive association was found between the risk of a child being stillborn and total external ppi (adjusted odds ratio per 100 mSv, 1.25 95% CI 1.04-1.46, p=0.019). The risk was higher for stillbirths with congenital anomaly and was highest for the nine stillbirths with neural tube defects, eight of which were anencephalic, the other spina bifida. Although, the possibility of an unmeasured risk factor for stillbirth, confounded with the effect of ppi, cannot be excluded, extensive checks confirmed that all statistical models were a reasonable fit to the data, and there was not statistical evidence of unmeasured risk factors. The statistical models predicted that, were the association to be interpreted as causal, up to 32 of the 130 stillbirths to the workforce between 1950 and 1989 might have been attributable to ppi. However, the increased risk to the current workforce is small. Current knowledge of the transgenerational risks of ionising radiation do not allow the observed relationship to be interpreted as casual, nor does it allow the possibility of a causal relationship to be excluded. The findings of an increased risk of stillbirth with increasing ppi are qualitatively consistent with those from animal models, though the risk estimate is higher. They are also consistent with reports of an increased risk of neural tube defects in the offspring of those occupationally exposed at the Hanford nuclear reprocessing plant. The studies of stillbirth and congenital anomalies in the offspring of the survivors of the atomic bombs at Hiroshima and Nagasaki indicate that up to three stillbirths due to ppi would have been expected amongst the offspring of the male Sellafield radiation workers. The higher estimate observed in this investigation is possibly due to the use of more appropriate statistical methods than were available for the analysis of the Japanese data. A comparable logistic regression analysis of the atomic bomb survivor data would allow a more direct comparison. (author)