Ju Hee Kim, Jisu Jung, Jungmin Kwak
Per- and polyfluoroalkyl substances (PFAS) accumulate in breast milk and pose a distinct hazard to breastfed infants, whose immature renal clearance substantially prolongs internal PFAS half-lives. We developed a two-compartment physiologically based pharmacokinetic (PBPK) model incorporating a postnatal glomerular filtration rate (GFR) maturation function to estimate individual-level infant blood PFAS concentrations from breast milk ingestion, and applied it to two prospective Korean birth cohorts (Cohort A: n = 221, 2018; Cohort B: n = 204, 2023). Five PFAS─perfluorodecanoic acid (PFDeA), perfluorohexanesulfonate (PFHxS), perfluorononanoic acid (PFNA), perfluorooctanoic acid (PFOA), and perfluorooctanesulfonate (PFOS)─were quantified by LC-MS/MS. Age-adjusted ANCOVA revealed compound-specific temporal shifts: PFHxS declined 66.7% (adjusted ratio: 0.33; p < 0.001), while PFNA and PFDeA increased 161.5% and 135.1% (both p < 0.001), evidencing regrettable substitution. PFOA and PFOS remained unchanged (both p > 0.69). Global sensitivity analysis confirmed that breast milk concentration and daily intake volume dominated model output uncertainty (|PRCC| = 0.82-0.91), whereas pharmacokinetic clearance parameters (elimination half-life and GFR maturation) had negligible influence on peak concentration (|PRCC| ≤ 0.05), reflecting accumulation-dominated kinetics over the first six months. External benchmarking against Fromme et al. (2010) yielded predicted-to-observed ratios of 0.67 and 0.61, respectively, within the prespecified acceptable range of 0.5-2.0. PBPK-estimated four-compound daily intake exceeded the EFSA (2020) TWI daily equivalent of 0.63 ng/kg/day (median multiples: 61× and 77×). These findings demonstrate that compound-specific regulatory action has not reduced aggregate infant fluorinated burden and highlight the urgent need for infant-specific PFAS reference values accounting for GFR immaturity.