Markov Chain Monte Carlo Analysis for Methylmercury Exposure Estimation.* B. Allen, H. Clewell, E. Hack, and J. Yager, Environ International, Electric Power Research Institute
A Bayesian analysis, referred to as Markov Chain Monte Carlo (MCMC), was applied, in combination with a physiologically based pharmacokinetic (PBPK) model of methylmercury (MeHg) developed for humans, to evaluate exposure variability in the U.S. population. The analysis was based on the newly available 1999 National Health and Nutrition Survey (NHANES) blood and hair mercury concentration data in a cohort of children aged 1-5 years and of women of childbearing age (16-49 years). The available PBPK model for MeHg predicts hair and blood mercury concentrations resulting from specific MeHg exposure scenarios. The PBPK model includes a number of pharmacokinetic parameters (e.g., tissue volumes, partition coefficients, rate constants for metabolism and elimination) that vary from individual to individual within the subpopulations of interest.
We have previously published a variability analysis using this model. In the present case, however, a somewhat different type of analysis was needed in order to estimate the variability in exposure (daily ingestion rate) implied by the data on the variation in blood and hair concentrations of mercury in the NHANES database. The blood and hair values represent the result (output) of a process that includes exposure to MeHg and its kinetics. The extent of, and variation in, the exposure must be estimated by a population distribution; prior information yields an estimate of that population distribution. The Bayesian approach, using MCMC, allows us to combine the prior distributions (of exposure levels and of PBPK model parameters) with the NHANES data to derive posterior distributions that "update" our estimates of population exposure. The revised MeHg exposure estimate distribution could then be compared to current exposure guidelines, including the USEPA RfD and the ATSDR MRL.
*The Electric Power Research Institute provided financial support for the analyses reported herein.
Go to . . .
2002 SRA Annual Meeting Table of Contents
2002 SRA Annual Meeting Author Index
Main Abstracts Menu Page
RiskWorld Home Page