A hospital-based case-control study of stillbirths and environmental exposure to arsenic using an atmospheric dispersion model linked to a geographical information system.

Although adverse reproductive outcomes have been associated with arsenic exposure, the extent and severity of the effects of chronic inhalation of low levels of arsenic on reproduction are not known. We conducted a hospital-based case-control study of stillbirths in a central Texas community that included a facility with more than a 60-year history of producing primarily arsenic-based agricultural products. We collected data on 119 cases and 267 controls randomly selected from healthy live-births at the same hospital and matched for year of birth. We abstracted medical and demographic data for the period January 1, 1983, to December 31, 1993, from hospital records and estimated socioeconomic status by median income from the 1990 Population and Housing Census data. We estimated arsenic exposure levels from airborne emission estimates and an atmospheric dispersion model and linked the results to a geographical information system (GIS) database. Exposure was linked by GIS to residential address at time of delivery. A conditional logistic regression model was fitted including maternal age, race/ethnicity, parity, income group, exposure as a categorical variable, and exposure-race/ethnicity interaction. The prevalence odds ratio observed for Hispanics in the high-exposure group (>100 ng per m3 arsenic) was 8.4, with a 95% confidence interval of 1.4-50.1.