Epidemiologic studies of cancer in populations near nuclear facilities.

We reviewed over 40 epidemiologic studies around nuclear power stations, fuel reprocessing plants, and weapons production facilities and testing sites in the United Kingdom, the United States, France, and Canada. We examined these studies for their potential to support a cause and effect relationship between cancer risk and radiation exposure. The extent to which an epidemiologic study supports a causal relation between radiation exposure and increased cancer risk can be evaluated using a set of criteria that have become known as Hill's postulates. In our review, epidemiologic studies yielded results that were biologically plausible and were supported by experimental data, but in almost all of the studies the methodologies were not adequate for evaluating causality. In the majority of cases, the methodologies did not permit examination of dose-response associations, making it impossible to support or refute causal relations. We suggest that investigators consider these issues when designing studies and employ dose reconstruction methodology to estimate radiation doses for specific individuals and population groups.