( In collaboration with Dr. Anne Riederer, Dr. Scott
Bartell and Dr. Barry Ryan, Emory University)
We want to find a subset of diet and non-diet predictors that will enable us to quantify the amount of PBA in a give human. We are working with Pyrethroids using the data available form the National Health and Nutrition Examination Survey (NHANES). Pyrethroids are the latest class of insecticides in global widespread use. We use a Bayesian Latent Gaussian model, similar to a tobit model to analyze the pyrethroid levels. The purpose of the investigation is to assess the relative importance of diet versus non-diet predictors. A secondary object is to explore whether or not the NHANES data could be used to identify particular foods driving urinary 3PBA levels in the population. The overall goal is to help identify major pathways of pyrethroid exposure in everyday American life, and thus inform the design of future exposure and intervention studies.
( In collaboration with Dr. Anne Riederer and Dr. Barry
Ryan, Emory University)
The aim of this projects it to use the latest EPA guidelines for cumulative risk assessment. The aim is to use Bayesian models and posterior distributions of model parameter to understand these guidelines and justify their use for looking at combined toxin intakes. The data are heavily censored and thus highly skewed. Our model will use dietary predictors for looking at the combined intake of chlorpyrifos and malathion and compare is the with new EPA guidelines to estimate proportion of the population exceeding the safe levels.