Assessing Global Endocrine Disrupting Activity in Drinking Water with Bioassays: A Possible Tool for Population-Level Exposure Assessment

Abstract

Background/Aim: Some natural and anthropogenic substances in drinking water sources are known or suspected endocrine disrupting chemicals (EDC), but few are regulated or routinely measured in U.S. public water supplies (PWS). Compared to measuring single analytes, bioassays that assess activity levels may better quantify population-level exposure to a variety of EDC. Methods: We conducted a pilot study of 10 PWS in Iowa, where agriculture-associated water contaminants (e.g., pesticides, nitrate) are suspected EDC. Source (untreated) water samples and finished (treated) drinking water samples were collected in spring and fall and concentrated 200x via a U.S. Geological Survey protocol. We assessed global EDC activity (aryl hydrocarbon [Ah], androgenic [AR], thyroid hormones [Ty], estrogenic [ER], and corticosteroid [CS]) with novel assays that express nuclear steroid receptor constructs in mammalian cell lines. We quantified statistically significant activity relative to negative controls and used bivariate and multivariate methods to compare activity overall, by season, and by utility/sample characteristics. Results: Among 62 samples (31 per season), 69% had significant Ah activity, 52% AR, 5% Ty, 2% ER and 2% CS. The percentage of samples with Ah activity was higher in the spring (84% versus 55%; c2p=0.01), and more common in utilities with ≥ median (≥ 3.19 mg/L) nitrate levels (p=0.04) or that were served by surface versus groundwater sources (p=0.02). AR activity was more common in spring versus fall samples (p<0.001), and we found non-significant within-season contrasts in activity across strata of treatment and regulated contaminants. Multivariable analyses were imprecise, but suggested that surface waters and untreated samples had greater levels of Ah and AR activity after adjusting for other factors. Conclusions: Our pilot results indicate that specific types of EDC activities in PWS vary by season and utility characteristics. This holistic approach to measuring EDC may hold promise for characterizing population-level EDC exposures through drinking water.