Abstract
Evidence from animal experiments and human studies suggests that electromagnetic fields may be human leukemogens. Epidemiological studies of leukemia in occupational groups and in the general populations are reviewed. The results are inconsistent. In the few studies showing an association between exposure to electromagnetic fields and the risk of leukemia, the temporal relationship between exposure and effect was not established, the observed associations were weak, the dose-response relationships were based on qualitative levels of exposure without regard to the duration of exposure or secondary sources of exposure, and the risk ratios were probably biased due to the population selection procedures used and misclassification of exposure. The proportionate mortality or cancer incidence ratios are unreliable estimates of relative risk. Further epidemiological research is needed to establish an association between exposure to electromagnetic fields and the risk of leukemia.