Epidemiology of occupational reproductive hazards: methodological aspects.

Both male and female workers may be exposed to occupational agents which affect their reproductive processes. Reproductive outcomes are important in their own right, for example teratogenic effects and impaired neurological development. In addition, reproductive impairment may indicate germ cell mutation, which has far-reaching implications. Research into reproductive effects is beset with pitfalls. Applying the results of toxicological research may be suggestive, but such extrapolation is very unreliable owing to profound differences in reproductive and/or toxicological mechanisms. Epidemiological data are therefore indispensable. Optimal methods of data collection and study design, comparable with similar studies of other populations, are essential. Otherwise reliable conclusions cannot be drawn, and this is a waste of resources and of the work and cooperation of everyone who participates in the research. In considering study design, the key question is the measurement (or minimally, the accurate description) of exposures, which is more problematic than the ascertainment of reproductive outcome variables. From the viewpoint of exposures, case control studies have the same disadvantage as studies of the general population, that exposures are heterogeneous, and most specific exposures are therefore rare, even in a large study; in addition, measurement is particularly difficult in this situation. Occupationally defined populations have the corresponding advantages of relative homogeneity and availability of data, although even here numerous problems exist, which are well known. Outcomes can be assessed by means of biological or questionnaire-based methods. They are best regarded as complementary, as each has its strengths and weaknesses.(ABSTRACT TRUNCATED AT 250 WORDS)