|
Activity Number:
|
383
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Risk Analysis
|
| Abstract - #309621 |
|
Title:
|
Bayesian Exposure Assessment for Pesticide Intake from Multiple Food Products
|
|
Author(s):
|
Ayona Chatterjee*+ and Graham Horgan and Chris Theobald
|
|
Companies:
|
University of West Georgia and Biomathematics & Statistics of Scotland and University of Edinburgh
|
|
Address:
|
3201 Post Woods Dr Apt A, Atlanta, GA, 30339,
|
|
Keywords:
|
exposure assessment ; latent variable models ; bayesian approach ; correlated intakes
|
|
Abstract:
|
Pesticide risk assessment involves combining information from consumption and concentration data sets to obtain a distribution for the pesticide intake in a human population. In this work we present a probabilistic, Bayesian approach to modeling the intake of the pesticide Iprodione though multiple food products. We develop latent-variable models that allow for both skewness and large numbers of zeros in the consumption and concentration data. Our consumption model also accounts for correlated intakes. We combine predicted intakes and concentrations from these models to obtain a distribution for individual Iprodione exposure. The use of our probabilistic approach is intended to yield more robust estimates of high percentiles of the exposure distribution than an empirical approach. Bayesian inference is used to facilitate the treatment of data with a complex structure.
|
