A Scalable Method for Quantitation of Per-and Polyfluoroalkyl Substances in Human Milk

Abstract Breastfeeding is a major route of early-life exposure to per- and polyfluoroalkyl substances (PFAS), yet population-level characterization of this exposure remains limited by the lack of sensitive, high-throughput methods to quantitate PFAS in human milk. To date, PFAS analysis in human milk has required large volumes (1−5 mL) and labor-intensive workflows that are unsuitable for large-scale, population-level biomonitoring. We developed and validated a semi-automated, high-throughput workflow for quantitating 23 legacy and replacement PFAS in 300 μL of human milk, enabling analysis of 72 samples per 20 h batch. Human milk and quality control (QC) samples are processed in 96-well format using an Opentrons Flex automated liquid-handling system with enhanced matrix removal (EMR) mixed-mode matrix cleanup and analyzed by liquid chromatography with high-resolution mass spectrometry (LC-HRMS). Method detection limits ranged from 0.53 to 9.19 pg/mL, with 22 compounds achieving limits of quantitation ≤5 pg/mL. Average accuracy was 93 ± 10% at 5 pg/mL and 90−98% at higher spiking levels. Applied to paired human milk samples collected at 1 and 6 months postpartum from 36 mothers, the method detected 4 PFAS above the LOQ in ≥50% of mothers at one or both time points. Their concentrations showed compound-specific changes, yet their summed concentration remained unchanged. Short-chain PFAS were quantifiable less consistently, with detection frequency shifting notably between time points. This scalable, low-volume workflow enables high-throughput quantitation of PFAS in human milk and supports large-scale biomonitoring of early-life exposure, with implications for exposure assessment, epidemiologic studies, and public health guidance.