Per- and polyfluoroalkyl substances, perceived stress, and miscarriage risk in a cohort of US Black women
Background: Per- and polyfluoroalkyl substances (PFAS) can disrupt hormonal and metabolic processes that support healthy pregnancy. PFAS exposure has been associated with miscarriage risk, but less is known about exposure to multiple PFAS and psychosocial stress measured at preconception, especially among Black women. Methods: We used prospective data from 538 participants from the Study of Environment, Lifestyle, and Fibroids who self-identified as Black or African American. At baseline (preconception), we measured six PFAS concentrations in plasma samples, and participants completed the Perceived Stress Scale-4 (PSS-4) (higher scores reflect greater stress). Participants reported pregnancy outcomes at follow-up study visits (~20, 40, and 60 months). We used Cox proportional hazards models, with gestational week as the time scale, to estimate adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) characterizing miscarriage risk associated with PFAS concentrations and perceived stress scores. We also used the survival extension of quantile-based g-computation to estimate associations of the mixture (PFAS and perceived stress scores) with miscarriage risk. Results: Ninety-five miscarriages were reported during follow-up. In Cox models, perfluorohexanesulfonic acid was associated with 99% (HR = 1.99, 95% CI = 1.10, 3.58) and 52% (HR = 1.52, 95% CI = 0.73, 3.17) higher miscarriage risk at the 50th–74th and ≥75th percentiles (vs. <50th percentile), respectively. Other PFAS showed little association with miscarriage risk. Perceived stress scores (per 1-standard deviation increase) were associated with 23% (HR = 1.23, 95% CI = 0.98, 1.53) higher miscarriage risk. The mixture was nonlinearly associated with miscarriage risk, but associations were imprecise. Conclusions: Higher perfluorohexanesulfonic acid concentrations and perceived stress scores were associated with miscarriage risk among Black women.