Calibration and validation of a dynamic model that projects population outcomes from methylmercury exposure from local fish consumption

Contamination of fish has led to unacceptable levels of exposure to methylmercury for many regions of the US. Within highly exposed populations such as subsistence fishers, the developing fetus is at particular risk for adverse effects. Regulatory agencies continue to consider how best to protect these populations. As a tool for these agencies, a dynamic model was developed to project mercury concentrations in common biomarkers of exposure in response to mercury concentrations in predatory fish from local waters. The model projects blood methylmercury, hair mercury, and cord blood concentrations for intake rates representing the mean, 90th, 95th, and 99th percentiles of populations of interest. For calibration and validation, chi-squared techniques were used to evaluate if the values from the model were similar to predicted values from the literature. There were no significant differences between model output and the literature values (all p-values > 0.05). While still not significantly different, the model slightly overestimates biomarker concentrations at the highest ingestion rates. The model predicts risk by comparing the projected biomarkers of exposure to levels deemed protective by the EPA. Decision makers can use the model to determine the portion of a population at risk and the degree of reduction in fish tissue that is necessary for the desired level of protection.