Problems with HBRA studies in health risk assessment

Epidemiological studies in High Background Radiation Areas (HBRAs) offer the prospect of large study populations and therefore appear attractive as a way of obtaining information about the effects of low doses of ionising radiation. However, bias and confounding are problems in all epidemiological studies and their residual effects may make it very difficult to detect relative risks lower than about 1.4. The easiest studies to propose will often be of ecological design in which data are aggregated over a population, for example, those living in a particular area, rather than cohort or case/control studies, which use data for individuals. Ecological studies are subject to problems of correlations between aggregated disease rates and aggregated measures of exposure which do not arise with studies which use individual data on disease, exposures and other risk factors. UNSCEAR suggests that only cohort and case/control studies should be used in quantitative assessments of radiation risks. © 2005.