We propose to conduct a prospective epidemiologic study of early pregnancy loss in a population of 8,000 women office workers. The goals are (1) to determine rates of early pregnancy loss using a newly developed, highly sensitive, immunoradiometric assay for human chorionic gondotropin (hcG) in urine; and (2) to determine, using this assay, whether early pregnancy loss is associated with cigarette smoking, alcohol consumption or exposure to video display terminals (VDTs). The epidemiology of early, subclinical pregnancy loss is not well defined. Both lifestyle and toxic environmental exposures may contribute to its etiology. It may be unrecognized cause of infertility. Previous studies in highly selected populations have demonstrated the utility of immunoradiometric hCG assay as an objective marker of early pregnancy loss. This molecular probe has not, however, been used to assess the epidemiology of subclinical early pregnancy loss in a general population. In an initial cross-sectional survey, we will characterize the population of women office workers, identify women at risk of becoming pregnant, and collect information on potential risk factors for early pregnancy loss. Women identified as at risk of becoming pregnant (sexually active and not using birth control) (estimated N=720) will be followed prospectively for one year. Each women will collect two urine samples each month at the beginning of menses; pregnancy and pregnancy loss will be established by hCG assay of these specimens. Each women will complete monthly reports detailing smoking, alcohol consumption and VDT use. Both job-related stressors (ergonomic and psychologic) and electromagnetic radiation exposures possibly associated with VDT use will be assessed. Maximum likelihood logistic regression will be used to model risk of early pregnancy loss.
