Bayesian benchmark dose assessment of per- and polyfluorinated substances exposure-associated thyroid function disruption during pregnancy

Abstract

Epidemiological evidence on maternal thyroid function disruption by prenatal exposure to perfluorinated and polyfluorinated substances (PFASs) is limited and inconsistent. The study examined the effects of PFASs exposure during early pregnancy on maternal thyroid function [free thyroxine (FT4), thyroid stimulating hormone (TSH), thyroid peroxidase antibodies (TPOAb) and FT4/TSH ratio]. The associations were evaluated using both single and mixed pollutant models, statistical analyses were further utilized in benchmark dose (BMD) estimations to offer critical references for human health risk assessment. Linear regression was used and then Bonferroni correction adjustment was set up to correct for multiple comparisons. The results revealed a significant association between PFHxS exposure and TSH (β = 0.473; 95 % CI: 0.180, 0.767). According to BKMR mixed-effects models, PFHxS was significantly positively correlated with TSH at the 25th percentile. PFASs were associated with the FT4/TSH ratio at the 25th to 40th percentile. The BMD value of the increasing FT4 effect induced by PFBA and PFPeA in pregnant women were 6.68 ng/mL and 1.37 ng/mL, respectively. The BMDs were obtained for TSH in the case of PFBA (0.33 ng/mL), PFHxS (0.28 ng/mL). Although BMDL₁₀ is higher than observed for maternal TSH elevation in animal studies, both studies agree that thyroid homeostasis is the sensitive target. The fact that BMD results at this stage are lower than current exposure levels to PFHxS underscores the urgency of prioritizing endocrine end points in PFASs risk assessment.