Associations of per- and polyfluoroalkyl substances with maternal metabolic and inflammatory biomarkers in early-to-mid-pregnancy

Abstract

Background: Per- and polyfluoroalkyl substances (PFAS) can disrupt metabolism. Early-to-mid pregnancy is characterized by amplified metabolic processes and inflammation to support maternal adaptations and fetal growth. Thus, we cross-sectionally evaluated whether PFAS are individually and jointly associated with these processes in early-to-mid pregnancy.

Methods: Pregnant Illinois women (n = 452) provided fasted blood samples at median 17 weeks gestation. We quantified serum perfluorononanoic (PFNA), perfluorooctane sulfonic (PFOS), perfluorooctanoic (PFOA), methyl-perfluorooctane sulfonamide acetic acid (Me-PFOSA-AcOH), perfluorohexanesulfonic (PFHxS), perfluorodecanoic (PFDeA), and perfluoroundecanoic (PFUdA) acid. Key outcomes were plasma glucose, insulin, C-peptide, insulin-like growth factor 1 (IGF-1), adiponectin, leptin, triglycerides, free fatty acids, total cholesterol, high-density lipoprotein (HDL) cholesterol, C-reactive protein, tumor necrosis factor alpha (TNF-α), monocyte chemoattractant protein-1 (MCP-1), and interleukin 6. We calculated homeostatic model assessment for insulin resistance (HOMA-IR), low-density lipoprotein (LDL) cholesterol, and very low-density lipoprotein (VLDL). We evaluated associations of PFAS with each metabolic/inflammatory biomarker individually using covariate-adjusted linear regression and jointly using quantile-based g-computation.

Results: In linear regression, all PFAS (except Me-PFOSA-AcOH) were negatively associated with insulin, HOMA-IR, and leptin, whereas all PFAS were positively associated with HDL cholesterol. We also observed negative associations of some PFAS with TNF-α and MCP-1; positive associations with adiponectin and total cholesterol also emerged. Additionally, PFOS was positively, whereas Me-PFOSA-AcOH was negatively, associated with triglycerides and VLDL. Each 25% increase in the PFAS mixture was associated with -31.3% lower insulin (95%CI: -45.8, -12.9), -31.9% lower HOMA-IR (95%CI: -46.4, -13.4), and -9.4% lower leptin (95%CI: -17.3, -0.8), but 7.4% higher HDL cholesterol (95%CI: 4.6, 10.3). For most outcomes, the major contributors to the PFAS mixture often differed compared to single-PFAS analyses.

Implications: Individual and joint PFAS exposures were associated with markers of maternal metabolism and inflammation in pregnancy. Further investigation is needed to elucidate possible mechanisms and consequences of these findings.

Keywords: Inflammation; Metabolism; Mixtures; PFAS; Pregnancy.

Figure 1.. Serum PFAS concentrations in I-KIDS pregnant women (n=452) and reproductive-aged women in NHANES 2013-14 (n=376), 2015-16 (n=328), and 2017-18 (n=315) survey cycles.

Data are presented on a log scale as 1.5 times the interquartile range below and above the 25^th and 75^th percentiles (lower and upper endpoints of whisker), the 25^th and 75^th percentiles (lower and upper edges of box), median (line inside box), and mean (large circle). I-KIDS PFAS levels < MDL were replaced with the MDL/sqrt(2), while NHANES PFAS levels < LOD were replaced with the LOD/sqrt(2). Concentrations of PFDoA (2017-2018), PFHpA (2015-2016, 2017-2018), PFBS (2015-2016, 2017-2018), PFOSA (2013-2014, 2015-2016, 2017-2018), and Et-PFOSA-AcOH (2013-2014, 2015-2016, 2017-2018) were not measured in all NHANES cycles and this is denoted by ‘NA’. Et-PFOSA-AcOH, ethyl-perfluorooctane sulfonamide acetic acid; I-KIDS, Illinois Kids Development Study; LOD, limit of detection; MDL, method detection limit; NA, not available; NHANES, National Health and Nutrition Examination Survey; PFAS, per- and polyfluoroalkyl substances; PFBS, perfluorobutane sulfonate; PFDeA, perfluorodecanoic acid; PFDoA, perfluorododecanoic acid; PFHpA, perfluoroheptanoic acid; PFHxS, perfluorohexanesulphonic acid; PFNA, perfluorononanoic acid; PFOA, perfluorooctanoic acid; PFOSA, perfluorooctane sulfonamide; PFOS, perfluorooctane sulfonic acid; PFUdA, perfluoroundecanoic acid.

Figure 2.. Joint associations of PFAS with metabolic and inflammatory biomarkers (n=447-452).

Data are presented as the percent change (%Δ, filled circle) and 95% CI (horizontal lines) in metabolic/inflammation biomarker concentrations for each 25% change in the PFAS mixture. QGComp models accounted for age, early pregnancy daily meat/poultry/fish intake, perceived stress score, pre-pregnancy BMI, gestational age at blood collection, race/ethnicity, educational attainment, annual household income, parity, smoking in the first trimester, alcohol intake in the first trimester, and employment status. Potentially meaningful associations are denoted with a red asterisk. Sample sizes for each exposure-outcome relationship can be found in Supplemental Figure 1. CI, confidence interval; CRP, C-reactive protein; HDL, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; LDL, low-density lipoprotein cholesterol; MCP-1, monocyte chemoattractant protein-1; QGCOMP, quantile-based g-computation; TNF-α, tumor necrosis factor alpha; VLDL, very low-density lipoprotein.

Figure 3.. Contributions of PFAS to overall joint associations of the mixture with metabolic and inflammatory biomarkers (n=447-452).

Data are presented as the relative chemical weight (in percent). QGComp models accounted for age, early pregnancy daily meat/poultry/fish intake, perceived stress score, pre-pregnancy BMI, gestational age at blood collection, race/ethnicity, educational attainment, annual household income, parity, smoking in the first trimester, alcohol intake in the first trimester, and employment status. PFAS contributing to the negative direction are denoted in dark blue and PFAS contributing to the positive direction are denoted in light blue. The sum of all weights in the negative direction equals 100% and the sum of all weights in the positive direction equals 100%; overall direction and strength of the mixture can be found in Figure 2 and Supplemental Table 4. Sample sizes for each exposure-outcome relationship can be found in Supplemental Figure 1. CRP, C-reactive protein; HDL, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; IGF-1, insulin-like growth factor 1; LDL, low-density lipoprotein cholesterol; MCP-1, monocyte chemoattractant protein-1; Me-PFOSA-AcOH, methyl-perfluorooctane sulfonamide acetic acid; PFAS, per- and polyfluoroalkyl substances; PFDeA, perfluorodecanoic acid; PFHxS, perfluorohexanesulphonic acid; PFNA, perfluorononanoic acid; PFOA, perfluorooctanoic acid; PFOS, perfluorooctane sulfonic acid; PFUdA, perfluoroundecanoic acid; TNF-α, tumor necrosis factor alpha; VLDL, very low density lipoprotein.