The Assessment of Risk of Radiation‐Induced Childhood Leukaemia in the Vicinity of Nuclear Installations

Increases in the incidence of childhood leukaemia have been reported in the vicinity of several nuclear installations. Conventional models of risk of radiation‐induced leukaemia are clearly inconsistent with the assumption that these increases are due to radiation exposure resulting from the proximity of these nuclear installations. Alternative interpretations are that the reported increases are (in some way) an artefact, that the increases are real but are not primarily due to radiation, that the population, or individuals within it, have received higher doses than expected, that present risk models are seriously flawed or that more indirect (e.g. genetic) pathways involving radiation are implicated. Present evidence makes it unlikely that population radiation doses or radiation risk factors could be in error by such large magnitudes as would be required. Attention has therefore focused on indirect mechanisms. Areas where further investigation may prove rewarding include radiation genetics, the chemistry of transuranic elements, personalized dosimetry and models of the role of hypothetical fluctuations in dose delivery, heterogeneity of dose and sensitivity within the population and the interaction of radiation with other agents.