Overall design considerations in male and female occupational reproductive studies.

Epidemiologic studies are extremely useful for examining the possible relationships between occupational exposures and reproductive outcomes in occupational populations. In this paper, we examine some of the major epidemiologic design considerations in male and female occupational reproductive studies. An important consideration in all epidemiologic studies is to choose appropriate outcome variables. One of the unique characteristics of reproductive studies is the wide range of potential outcome variables: from problems of infertility, through fetal development, and into childhood diseases and diseases of young adulthood. Usually, it is infeasible to look at all possible outcomes. By omitting certain outcomes, however, important associations between exposure and outcome may remain undetected. Epidemiology relies on observations of populations. Factors in data availability, collection, and analysis, and in study design often limit the inferences that may be drawn. The "exposed" or "at risk" groups are not likely to be randomly selected and the investigator has little control over the exposures of the study population. Unlike experimental investigations, epidemiologic observations are made on individuals exposed to a variety of hazards during their lives. The size of the population available for study is critical. Bias from potential confounding factors can be minimized through the use of appropriate matching and/or statistical techniques. Many of the topics discussed here will be expanded in subsequent chapters. Although this paper is designed to assist in planning occupational epidemiologic studies, we strongly recommend that, when attempting to conduct such studies, professional guidance from epidemiologists and biostatisticians be obtained.