Quantitative Associations of Polluting Chemicals and Endogenous Biomolecules in Hair: An Observational Perspective from a Population-Based Study

Abstract

Hair is a valuable, non-invasive material for human biomonitoring. However, little is known about polluting contaminants in hair, particularly regarding the relationship between biocomponents and contaminant levels in hair. We measured the concentrations of 42 contaminants, including 11 phosphorus flame retardants (PFRs), 13 bisphenols (BPs), and 18 perfluoroalkyl and polyfluoroalkyl substances (PFAS), while simultaneously measuring the levels of keratin, melanin, and eight sphingolipids in hair samples. Long-chain sphingolipids (C20CER) were negatively associated with levels of PFRs, PFAS, and BPs, while C12CER and C14CER (short-chain) were positively associated with levels of PFRs and BPs. Furthermore, we observed an overall negative association between ∑7PFRs and endogenous biocomponents but a positive dose–effect relationship with ∑5BPs and biocomponents using Bayesian kernel machine regression models. Among the biocomponents, C20CER and C14CER contributed the most to the negative and positive associations, respectively. Specifically, a change in Ln C20CER (Z-score) concentration from the 25th to 75th percentile was associated with a decrease in ∑7PFRs of 47.0%-SD (−61.8%, −32.3%) when other biocomponents were at their median values. These findings provide new insights into the relationships between biocomponents and contaminants in hair, which is an essential step for the advancement of hair as a biomonitoring material.