RESUMEN
INTRODUCTION: Bisphenol A (BPA) is frequently used in the production of plastics. It is an endocrine disruptor, and BPA exposure in mice has been associated with reduced offspring growth due to insufficient milk production. However, human studies of associations between BPA exposure and duration of breastfeeding are sparse. METHODS: Pregnant women from the Odense Child Cohort (n = 725) donated a third trimester morning urine sample, which was analyzed for BPA by LC-MS/MS. Information about duration of exclusive and any breastfeeding was obtained through questionnaires three and 18 months postpartum, and a subgroup of women responded to weekly text messages about breastfeeding. Associations between pregnancy BPA exposure and duration of breastfeeding were analyzed using Cox regression adjusting for potential confounders. RESULTS: The median urine BPA concentration was 1.29 ng/mL. Compared to women within the lowest tertile of BPA exposure, women in the second and third tertile were slightly more likely to terminate breastfeeding at any given time; HRs (95% CI) were 1.05 (0.87; 1.26) and 1.06 (0.89; 1.27), respectively, and to terminate exclusive breastfeeding at any time up to 20 weeks after birth, HRs (95% CI) were 1.07 (0.88; 1.28) and 1.06 (0.88; 1.27), respectively. However, confidence intervals were also compatible with no effect or even a protective effect. DISCUSSION: This study indicated that high BPA exposure in pregnancy was associated with shorter duration of breastfeeding. Although our findings were not statistically significant, all estimates were above one suggesting increased risk of early breastfeeding termination with high exposure. Using a single spot morning urine sample to measure BPA has likely caused imprecision as it might not adequately reflect long term exposure. Future studies should consider measuring BPA more than once, including other timepoints during pregnancy and after birth.