You are here:
Evaluation of High-Throughput Chemical Exposure Models via Analysis of Matched Environmental and Biological Media Measurements
Citation:
Setzer, Woodrow, J. Wambaugh, R. Dodson, R. Rudel, K. Isaacs, AND D. Biryol. Evaluation of High-Throughput Chemical Exposure Models via Analysis of Matched Environmental and Biological Media Measurements. Society for Toxicology Annual Meeting 2017, Baltimore, MD, March 12 - 16, 2017.
Impact/Purpose:
The purpose of this research is to use real exposure data to validate reverse and forward methods being used in the ExpoCast project.
Description:
The U.S. EPA, under its ExpoCast program, is developing high-throughput near-field modeling methods to estimate human chemical exposure and to provide real-world context to high-throughput screening (HTS) hazard data. These novel modeling methods include reverse methods to infer parent chemical exposures from biomonitoring measurements and forward models to predict multi-pathway exposures from chemical use information and/or residential media concentrations. Here, both forward and reverse modeling methods are used to characterize the relationship between matched near-field environmental (air and dust) and biomarker measurements. Indoor air, house dust, and urine samples from a sample of 120 females (aged 60 to 80 years) were analyzed. In the measured data, 78% of the residential media measurements (across 80 chemicals) and 54% of the urine measurements (across 21 chemicals) were censored, i.e. below the limit of quantification (LOQ). Because of the degree of censoring, we applied a Bayesian approach to impute censored values for 69 chemicals having at least 15% of measurements above LOQ. This resulted in 10 chemicals (5 phthalates, 5 pesticides) with matched air, dust, and urine metabolite measurements. The population medians of indoor air and dust concentrations were compared to population median exposures inferred from urine metabolites concentrations using a high-throughput reverse-dosimetry approach. Median air and dust concentrations were found to be correlated with inferred exposures (dust: r=0.60, air: r=0.56). Inferred phthalate exposures were generally underestimated by exposure predictions based on media concentrations (via inhalation, dermal, nondietary ingestion pathways); combining forward food-contact exposure estimates with indirect exposures in a linear model of inferred parent exposure yielded an R2=0.67 and p=0.02. These result indicate that the forward and reverse modeling methods being developed in ExpoCast can aid in the interpretation of measured urinary biomarkers via inference of corresponding intake of parent chemicals and elucidation of contributing exposure pathways.
