Early-life exposure to perfluoroalkyl substances in relation to serum adipokines in a longitudinal birth cohort

Background: Per- and polyfluoroalkyl substances (PFAS) exposure has been linked to metabolic health outcomes such as obesity, and changes in adipokine hormones may be one of the underlying biological mechanisms. We prospectively evaluated the associations between prenatal and early childhood exposures to PFASs and adipokines in children. Material and methods: PFAS concentrations were measured in serum samples collected at birth, 18 months, and 5 and 9 years, and adiponectin, leptin, leptin receptor, and resistin were measured in serum samples collected at birth and 9 years. We used multivariable linear regression models to estimate the percent change in serum-adipokine concentrations for a doubling in serum-PFAS concentrations. The potential sex-specific effect of PFAS was assessed by including an interaction term between PFAS and sex in each model. Bayesian kernel machine regression (BKMR) was implemented to evaluate the overall effect of PFAS mixtures. Results: Significant associations with leptin, leptin receptor, and resistin at age 9 years were observed for serum-PFAS concentrations at 18 months and 5 and 9 years, whereas associations for PFAS concentrations at birth were mostly null. However, we observed a positive association between serum-PFHxS at birth and leptin receptor at birth. We found limited evidence regarding modification effect of sex on serum-PFAS concentrations. BKMR findings were consistent and suggested some significant effects of the overall PFAS mixtures at 18 months and 5 and 9 years on adipokine concentrations at 9 years. Conclusions: Given the associations of PFAS exposure with both adipokine hormones and metabolic functions, future studies should include assessment of adipokine hormones when examining PFAS-associated metabolic alterations.


Figure S1 .
Figure S1.Directed acyclic graph of the hypothesized causal pathway between child serum-PFAS concentrations at birth, 18 months, and 5 years and child serum-adipokine concentrations at birth and 9 years.

Figure S2 .
Figure S2.Directed acyclic graph of the hypothesized causal pathway between child serum-PFAS concentrations at 9 years and child serum-adipokine concentrations at 9 years.

Figure
Figure S5.Sex-specific percent change of serum-adipokine hormone concentrations at birth and age 9 years per doubling of the serum-PFAS concentrations at birth, 18 months, and 5 and 9 years.

Figure S6 .
Figure S6.Dose response function between each serum-PFAS concentration at 18 month and resistin at 9 years.

Figure S8 .
Figure S8.Dose response function between each serum-PFAS concentration at 5 years and leptin receptor at 9 years.

Figure S9 .
Figure S9.Comparison between results using generalized estimating equations to primary results using linear regression models in the overall study population.Results are presented as the percent change of serum-adipokine hormone concentrations at age 9 years per doubling of the serum-PFAS concentrations at birth, 18 months, and 5 and 9 years.

Figure S10 .
Figure S10.Comparison between sex-specific results using generalized estimating equations to primary results using linear regression models.Results are presented as the percent change of serum-adipokine hormone concentrations at age 9 years per doubling of the serum-PFAS concentrations at birth, 18 months, and 5 and 9 years.

Figure S11 .
Figure S11.Comparison of effect estimates from the primary models and models additionally including duration of exclusive breastfeeding for the associations of serum-PFAS concentrations at 18 months and 5 years with serum-adipokine hormone concentrations at age 9 years.

Figure S13 .
Figure S13.Comparison of effect estimates from the primary models and models additionally including child whale consumption at 9 years for the associations of serum-PFAS concentrations at 9 years with serum-adipokine hormone concentrations at age 9 years.

Table S1 .
Percent change of the serum-adipokine concentrations at birth per doubling of the serum-PFAS concentrations at birth, overall (n = 463) and by sex (male = 241; female = 222).

Table S2 .
Percent change of the serum resistin concentrations at age 9 years per doubling of the serum-PFAS concentrations at birth, ages 18 months, and 5 and 9 years, overall and by sex.Models were adjusted for maternal age (years), maternal pre-pregnancy body mass index a

Table S5 .
Percent change of the serum leptin receptor concentrations at age 9 years per doubling of the serum-PFAS concentrations at birth, ages 18 months, and 5 and 9 years, overall and by sex.

Table S6 .
Posterior inclusion probabilities (PIPs) of five PFASs measured at birth in relation to the serum-adipokine concentrations at birth in the overall population.

Table S7 .
Posterior inclusion probabilities (PIPs) of five PFASs measured at birth, 18 months, and 5 and 9 years in relation to the serum-adipokine concentrations at 9 years in the overall population.