Abstract
Effects of small doses of ionising radiation are currently estimated by extrapolation from effects observed at higher doses modified by knowledge of the results of experiments in animals or in vitro. Three are possible: reduced intelligence if the fetus is irradiated at 8-25 weeks in utero, hereditary disease, and cancer. Direct observation of the effect of small doses has been possible only for cancer. Observations on 131,000 radiation workers in the UK and the USA lead to estimates of the lifetime risks of leukaemia and all cancers close to those obtained by extrapolation, halved to adjust for the lesser effect of doses given at low dose rates, as do observations on children in Utah and the Nordic countries exposed to radioactive fallout from nuclear weapons testing, respectively, in Nevada and worldwide. Clusters of leukaemia and NHL in young people near some nuclear installations in the UK cannot be attributed to chance, local pollution with radioactive waste, or occupational exposure of their fathers before the children's conception. There remains the possibility that they result from the spread of infection in susceptible individuals from the mixing of urban and rural populations.