Quantifying levels of per- and polyfluoroalkyl Substances (PFAS) in water and serum after contamination from agricultural biosolid application

Abstract

The National Academies of Sciences, Engineering, and Medicine recommends per- and polyfluoroalkyl substance (PFAS) blood testing for patients with risk of elevated exposure, and the Agency for Toxic Substances and Disease Registry (ATSDR) suggests PFAS blood testing based on exposure. Barriers to PFAS blood testing include cost, access to labs, and evolving laboratory methods. We quantify water and serum PFAS levels among a highly-exposed cohort in an area with groundwater contaminated by historical agricultural biosolid application. We compare the gold standard PFAS serum test with a commercial test and results from a one-compartment toxicokinetic model. Participants were adults (n = 30) whose household (n = 19) water had levels of the sum of six PFAS > 500 ng/L. Serum PFAS were measured using liquid chromatography-tandem mass spectrometry. Demographic and water consumption data were collected via telephone. Serum PFAS results from the commercial test were accessed via medical record. Statistical analysis included descriptive statistics and bivariate plots of serum levels. Perfluorohexanoic acid, perfluoroheptanoic acid (PFHpA), perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), perfluorobutanesulfonic acid, perfluorohexanesulfonic acid (PFHxS), and perfluorooctanesulfonic acid (PFOS) were detected in 19 wells, and PFHpA, PFOA, PFNA, perfluorodecanoic acid, perfluoroundecanoic acid, PFHxS, and PFOS were detected in at least 19 participants’ serum. In well water, PFOA and PFOS levels had geometric means (GMs) of 1749 ng/L (geometric standard deviation [GSD] 2.4) and 887 ng/L (GSD 19.7), respectively. In serum, PFOA and PFOS had GMs of 116.2 µg/L (GSD 13.5) and 58.3 µg/L (GSD 13.8), respectively. Our results are comparable with and had a wider mix of PFAS than other high-exposure cohorts. There was good agreement between the commercial and gold standard tests for PFOA, PFNA, and PFHxS, and mixed agreement between the gold standard test and modeled predictions, suggesting water-based toxicokinetic models of serum PFAS may be inadequate for assessing exposure in this population.