Alcohol and cardiac arrhythmias.

likely to have occurred. Longstanding evidence that fetal exposure to medical radiation is a risk factor,' together with the report by Gardner and his colleagues6 linking risk to paternal occupational exposure to radiation immediately before conception, suggest that residence at birth (or before birth) may be more relevant than residence at the time of registration or death. A recent study of the cases arising in children living around Dounreay,7 however, found that the risk was no higher in the children born in the area than in those moving there after birth. Has the large amount of research money that has gone into the epidemiology of this small but unquestionably important problem been justified? Certainly it has stimulated the development ofnew statistical approaches to spatial analyses, which may be applicable to other circumstances. It has also strengthened the validity of what was previously weak evidence on the presence of clusters of cases, even if we have learnt little more about the effects of environmental radiation. Considerably more research is needed, probably on an individual basis, before the causes of these devastating diseases are fully understood. EVA ALBERMAN Professor Emeritus Department of Preventive and Environmental Medicine, Wolfson Institute of Preventive Medicine, Medical College of St Bartholomew's Hospital, London EC1M 6BQ