Early-pregnancy plasma per- and polyfluoroalkyl substance (PFAS) concentrations and hypertensive disorders of pregnancy in the Project Viva cohort

Abstract

Background: Hypertensive disorders of pregnancy (HDP), defined here as hypertensive disorders with onset in pregnancy (i.e., gestational hypertension, preeclampsia, and preeclampsia superimposed on chronic hypertension), affect up to 10% of pregnancies in the United States and are associated with substantial maternal and neonatal morbidity and mortality. Per- and polyfluoroalkyl substances (PFAS) are associated with adverse cardiometabolic outcomes during pregnancy, but associations between PFAS and HDP are inconsistent and joint effects of PFAS mixtures have not been evaluated.

Methods: We studied 1,558 pregnant individuals from the Project Viva cohort, recruited during 1999-2002. We quantified concentrations of eight PFAS in plasma samples (median 9.7 weeks of gestation). Using clinical records, we calculated trimester-specific mean systolic (SBP) and diastolic (DBP) blood pressure and categorized HDP status [no HDP (normotensive & chronic hypertension), gestational hypertension, preeclampsia]. We estimated associations of individual PFAS with HDP using multinomial logistic regression and estimated associations with blood pressure using linear regression. We used Bayesian kernel machine regression (BKMR) and quantile g-computation to assess joint effects of the PFAS mixture on HDP and blood pressure measures.

Results: Four percent of participants developed preeclampsia and 7% developed gestational hypertension. We observed higher odds of gestational hypertension, but not preeclampsia, per doubling of perfluorooctanoate (PFOA) [OR = 1.51 (95% confidence interval: 1.12, 2.03)], perfluorooctane sulfonate (PFOS) [OR = 1.38 (1.04, 1.82)], and perfluorohexane sulfonate [OR = 1.28 (1.06, 1.54)] concentrations. We observed higher mean DBP per doubling of PFOA [2nd trimester (T2): 0.39 mmHg (-0.01, 0.78); 3rd trimester (T3): 0.56 mmHg (0.14, 0.98)] and PFOS [T2: 0.46 mmHg (0.11, 0.82); T3: 0.43 mmHg (0.05, 0.80)]. The PFAS mixture was positively associated with odds of gestational hypertension [75th vs. 50th percentile: OR = 1.14 (95% credible interval:1.03, 1.25), BKMR] and mean DBP [T2 = 0.17 mmHg (-0.06, 0.40); T3 = 0.22 mmHg (-0.03, 0.48), BKMR].

Conclusions: These findings suggest that exposure to certain PFAS may increase the odds of gestational hypertension during pregnancy, with potential implications for subsequent maternal and child health outcomes.

Keywords: Blood pressure; Gestational hypertension; Hypertensive disorders of pregnancy; PFAS; Preeclampsia.

Fig. 1.

Overall effect of the per- and polyfluoroalkyl substances (PFAS) mixture on the probability of (A) gestational hypertension and (B) preeclampsia vs. no hypertensive disorders of pregnancyt (HDP) estimated by the probit extension of Bayesian kernel machine regression, adjusting for participant age at enrollment, race/ethnicity, marital status, smoking status, and education. Figures show the probit model estimates and 95% credible intervals for (A) gestational hypertension vs. normotension and (B) preeclampsia vs. normotension when all PFAS concentrations are held at certain percentiles (e.g., 25th, 45th, 75th) compared to when PFAS concentrations are held at the median.

Fig. 2.

Overall effect of the per- and polyfluoroalkyl substances (PFAS) mixture on mean trimester-specific diastolic blood pressure (SBP) (mmHg), adjusting for participant age at enrollment, race/ethnicity, marital status, smoking status, and education. These plots show the estimated differences in mean (A) second trimester and (B) third trimester SBP and 95% credible intervals when concentrations of all PFAS are held at a certain percentile (e.g., 25th, 45th, 75th) compared to when all PFAS concentrations are held at the median.

Fig. 3.

Overall effect of the per- and polyfluoroalkyl substances (PFAS) mixture on mean trimester-specific diastolic blood pressure (DBP) (mmHg), adjusting for participant age at enrollment, race/ethnicity, marital status, smoking status, and education. These plots show the estimated differences in mean (A) second trimester and (B) third trimester DBP and 95% credible intervals when concentrations of all PFAS are held at a certain percentile (e.g., 25th, 45th, 75th) compared to when all PFAS concentrations are held at the median.