Using time-to-pregnancy data to study occupational exposures: methodology.

Occupational and environmental influences on fertility are for the most part unstudied, partly because sensitive methods for studying them have not been developed. We are developing a measure of fecundability, the monthly probability of pregnancy, by studying time to pregnancy, the number of noncontracepting menstrual cycles each couple requires to conceive. The relationship of this measure of reproductive impairment with others, such as spontaneous abortion, is not known. Preliminary data from two sources suggest that reduced fertility is not highly correlated with increased risk of spontaneous abortion, despite predictions to the contrary from the toxicology literature. A current study of occupational exposures of dental assistants will address questions of data quality by providing comparisons of brief responses from mail questionnaires and telephone interviews with very detailed data from telephone interviews. This study also will allow estimation of the magnitude of two potential selection biases: selection of only planned pregnancies (time-to-pregnancy data for accidental pregnancies are not meaningful), and selection against highly infertile and sterile couples (when studying currently or previously pregnant women, sterile couples are not represented at all and highly infertile couples are underrepresented).