Factors affecting sampling strategies for design of an effects‐directed analysis for endocrine‐active chemicals

Effects‐directed analysis (EDA) is an important tool for identifying unknown bioactive components in a complex mixture. Such an analysis of endocrine‐active chemicals (EACs) from water sources has promising regulatory implications but also unique logistical challenges. We propose a conceptual EDA (framework) based on a critical review of EDA literature and concentrations of common EACs in waste and surface waters. Required water volumes for identification of EACs under this EDA framework were estimated based on bioassay performance (in vitro and in vivo bioassays), limits of quantification by mass spectrometry (MS), and EAC water concentrations. Sample volumes for EDA across the EACs showed high variation in the bioassay detectors, with genistein, bisphenol A, and androstenedione requiring very high sample volumes and ethinylestradiol and 17β‐trenbolone requiring low sample volumes. Sample volume based on the MS detector was far less variable across the EACs. The EDA framework equation was rearranged to calculate detector “thresholds,” and these thresholds were compared with the literature EAC water concentrations to evaluate the feasibility of the EDA framework. In the majority of instances, feasibility of the EDA was limited by the bioassay, not MS detection. Mixed model analysis showed that the volumes required for a successful EDA were affected by the potentially responsible EAC, detection methods, and the water source type, with detection method having the greatest effect on the EDA of estrogens and androgens. The EDA framework, equation, and model we present provide a valuable tool for designing a successful EDA.