BMD Analysis of Multiple Endpoints in Human Health Risk Assessment: Chloroprene Case Study
On this page:
Human health risk assessment often evaluates multiple endpoint variables in order to form a comprehensive picture of the toxicity effects of an environmental chemical or toxicant. Benchmark dose (BMD) modeling is a flexible method that takes the shape of the dose-response curve and important measures of uncertainty and variability in the data into account. While BMD modeling is an improvement over previous (e.g., NOAEL) methods, existing BMD models are designed to evaluate single or independent endpoints, not multiple related endpoints simultaneously. Therefore, development of an advanced BMD approach for efficient analysis of multiple related endpoints could be beneficial to risk assessment. This presentation will present our research in univariate, ordinal and multivariate analysis of multiple endpoints including simultaneous analysis of multiple correlated endpoints using environmental chemical toxicity data. Our results demonstrate that multivariate BMD approaches produced comparable and stable BMDLs in a single analysis, with less variability and more robust estimation of adverse health outcomes. In addition, this presentation will also include the major approaches that are currently used in multivariate modeling of multiple endpoints for human health risk assessment, and the challenges in applying these models, such as model choice and model averaging, clustered, Bayesian versus likelihood or frequentist methods, and the analysis of aggregate versus individual data.
(The information in this Abstract has been subjected to review by the Center for Public Health and Environmental Assessment and approved for presentation. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.)