Gestational PFAS concentrations in association with maternal depression and perceived stress during pregnancy and postpartum in the ECHO cohort.

Abstract

Depression and perceived stress are common perinatal morbidities. We assessed associations between prenatal per- and polyfluoroalkyl substances (PFAS) with prenatal and postpartum depression and perceived stress. We analyzed concentrations of 7 PFAS in maternal plasma or serum in the Environmental influences on Child Health Outcomes (ECHO) cohort (n = 4403). We ascertained prenatal depression diagnosis; prenatal and postpartum depressive symptom severity using a Patient-Reported Outcomes Measurement Information System Depression (PROMIS-D) t-score; and prenatal and postpartum perceived stress using a Perceived Stress Scale (PSS) t-score. Linear and modified Poisson models assessed associations of PFAS with outcomes. Quantile g-computation assessed PFAS mixtures. We evaluated effect modification by nativity and Social Vulnerability Index. N-MeFOSAA was positively associated, PFDA was inversely associated, and PFOS and PFOA were nonlinearly associated with multiple outcomes. For example, the highest category of N-MeFOSAA was associated with 0.93 (95 %CI: -0.17, 2.03) units higher postpartum PROMIS-D t-scores, and 1.88 (95 %CI: 1.05, 2.72) units higher postpartum PSS t-scores. In contrast, the highest category of PFDA was associated with 25 % (risk ratio [RR]: 0.75 [95 %CI: 0.56, 1.01]) lower risk of depression diagnosis and 0.83 (95 %CI: 0.14, 1.51) units lower prenatal PSS t-scores. The middle and highest categories of PFOS were associated with 26 % (RR: 1.26 [95 %CI: 1.05,1.52]) higher and 38 % (RR: 0.62 [95 %CI: 0.37, 1.05]) lower risk of depression diagnosis; and 0.97 (95 %CI: 0.09, 1.84) units higher and 0.04 (95 %CI: -1.53, 1.62) units lower prenatal PSS t-score. Some PFAS may be associated with altered depressive symptoms and perceived stress; investigation into mechanisms is needed.