Temporal Fallacies in Biomarker Based Exposure Inference. S. M. Bartell, R. P. Ponce, W. C. Griffith, and E. M. Faustman, University of Washington Institute for Risk Analysis and Risk Communication, Seattle, WA.
Biomarker measurements (e.g. hair mercury, blood lead, and urinary arsenic) obtained at single time points are often used to make steady-state inferences about toxicant exposure over longer periods. However, most toxicant exposures vary in magnitude over time and many are intermittent or episodic, rendering the steady-state assumption invalid. While it is often assumed that any error introduced by a fallacious steady-state assumption is minimal and can safely be ignored, the degree of error potentially introduced by this temporal fallacy is not typically investigated. We present a general framework for evaluating the magnitude of error introduced by this temporal fallacy under stochastic exposure conditions and demonstrate its application to mercury blood and hair biomarkers. We discuss nonparametric solutions for the expectation and variance of temporal fallacy error under simplifying conditions such as general independence of terms and describe a simulation model for exploration of more complex assumption sets. Model results can be combined with de minimus error levels to provide guidelines for determining whether or not the potential error introduced by temporal fallacies is negligible in specific situations.
Sponsored by the National Research Center for Statistics and the Environment by Environmental Protection Agency Cooperative Agreement CR-825173-01-4 and by the Consortium for Risk Evaluation with Stakeholder Participation by Department of Energy Cooperative Agreement DE-FCO1-95EW55084.
Go to . . .
1999 SRA Table of Contents
1999 SRA Author Index
Main Abstracts Menu Page
RiskWorld Home Page