Epidemiologic issues in neurotoxicity research.
Epidemiologic studies of the relationship of exposure to toxic environmental agents and neurologic disorders are essential for understanding mechanisms of pathogenesis and for evaluating dose-response relationships. The stages of execution of an epidemiologic study include problem specification, choice of study type, selection of study population, selection of measures of exposure and effect, specification of relevant confounding factors, data collection, analysis, and interpretation. Although both case-control and cohort studies have been performed, a clear deficiency in the number and variety of case-control studies in this area points to a need for future collaborative research. Typical shortcomings of prior research include failure to adequately quantify individual exposure/dose levels, failure to control for relevant confounding factors, and failure to carefully evaluate selection bias. Test selection in epidemiologic studies of exposed groups should be determined by prior demonstration of a high correlation of "abnormal" test results to documented disease states and by considerations of the biological significance of specific levels of change in test performance between groups. Established criteria exist which should be used in making causal inference from results of epidemiologic studies. If properly designed, executed, and interpreted, epidemiologic investigations occupy a unique position in scientific studies designed to understand and prevent neurotoxic disease in human populations.