Abstract Purpose Estimates of human exposure are a major component of chemical risk assessment. Studies of bisphenol A (BPA) have raised concern that exposure has been underestimated because the lack of standards for the measurement of the major BPA metabolites has necessitated the use of flawed analytical tools to indirectly estimate them. Because other endocrine disrupting chemicals (EDCs) are measured using similar indirect methods, we evaluated the accuracy of indirect analysis for representatives from three different classes of non-persistent EDCs that undergo rapid phase II metabolism; bisphenols, parabens, and phthalates. Methods A direct LC–MS/MS method that simultaneously measures bisphenol S (BPS), propyl paraben (PrP), and monobutyl phthalate (MBP), and their major metabolites in urine, was used to quantify these EDCs in sixty second-trimester human urine samples. The same samples also were analyzed with a widely used indirect method that requires enzymatic hydrolysis prior to estimating metabolite levels. Results Marked discrepancies were evident when maternal urine samples were analyzed by both direct and indirect methods. Indirect analysis underestimated levels of all three EDCs and BPA, with the magnitude of underestimation varying by analyte. Conclusions The accuracy of widely used “indirect” analytical methods that estimate metabolite levels in human urine is neither predictable nor consistent. Greater precision and accuracy is attained using authentic standards for metabolites. Given the importance of biomonitoring data in estimating human EDC exposure, analytical accuracy is critical. Availability of standards for both the parent compound and its major metabolites should be required before a chemical enters the marketplace.
Gerona et al. (Mon,) studied this question.