Background. Considerable controversy and uncertainty surround the reproductive health threat from exposure to bisphenol A (BPA). This common chemical to which most humans are exposed has long been known to have estrogenic properties, but much more research is needed on the effects of BPA exposure during pregnancy.
Objectives. To address this uncertainty, we investigated exposure to BPA and associations with birth outcomes with particular focus on the vulnerable subpopulation of low-income, African-American women in the first trimester of pregnancy.
Methods. We conducted cross-sectional analyses of the National Health and Nutrition Examination Survey (NHANES), a nationally representative sample of the U.S. population, examining relations between demographic and occupational factors and urinary BPA concentrations. Logistic regression was used to model the relation between current urinary BPA concentration and reproductive history. This secondary data analysis was complemented by a clinic-based primary data collection effort where we enrolled low-income African-American women in the first trimester of pregnancy (n=16) to compare urinary BPA concentrations to self-reported contact with potential sources of BPA and to assess feasibility of these methods for a future longitudinal study of prenatal BPA exposure and birth outcomes.
Results. Women of reproductive age in NHANES had increased odds of having delivered a low birth weight baby (3% per 10% increase in urinary BPA; p=0.046) and decreased odds of having delivered a high birth weight baby (5% per 10% increase in urinary BPA; p=0.02). History of ever being pregnant, ever giving birth to a live infant, ever experiencing a pregnancy loss, or ever having been diagnosed with gestational diabetes were not associated with BPA exposure (p≥.39). U.S. workers (male and female) employed in the "retail-food stores" industry had higher urinary BPA concentrations than workers in other industries (p=0.05).
With 11 of 16 samples above the limit of quantitation in our feasibility study, the geometric mean (GM) urinary BPA concentration was 0.46 μg/L (95% CI: 0.20, 1.1 μg/L), and the GM of creatinine-adjusted urinary BPA was 0.35 μg/g (95% CI: 0.12, 1.0 μg/g). Contact with common potential sources of BPA exposure identified through the personal interviews was infrequent. We observed breakfast consumption prior to enrollment as the only factor associated with creatinine-adjusted urinary BPA.
Conclusions. Taken together, these findings contribute to elucidating the potential risks that BPA exposure may pose for reproductive health and infant development. Our studies have indicated that birth weight may be adversely impacted by prenatal exposure to BPA. Furthermore, employment in the retail industry may contribute to increased exposure to BPA through dermal contact with receipts. Our feasibility study in first trimester, low-income African-American women helped refine methods for use in future studies of occupational and other BPA exposure sources and adverse effects on birth weight in this particularly vulnerable subpopulation.