The use of epidemiological data in risk assessment

Next to toxicological data, epidemiological studies form the scientific basis for risk assessment. Both approaches have their particular advantages and limitations. The most important limitation of epidemiological studies is the lack of exact data on past exposures. Their most important advantage is that they study the actual occurrence of diseases in humans. Epidemiological and experimental data should be regarded as complementary and should be used in close conjunction in the process of risk assessment. Apart from these advantages and limitations several specific issues are addressed in this paper, such as study populations of particular interest, the problem of low-dose extrapolation, the interpretation of negative results, derivation of no-effect levels from epidemiologic studies, and the application of safety factors for epidemiological studies.