Activity Number:
|
166
- Understanding Mixtures in Environmental Epidemiology
|
Type:
|
Topic Contributed
|
Date/Time:
|
Tuesday, August 4, 2020 : 10:00 AM to 11:50 AM
|
Sponsor:
|
Section on Statistics in Epidemiology
|
Abstract #313439
|
|
Title:
|
Estimation of the Total Main and Interaction Effects of a Mixture of Pollutants.
|
Author(s):
|
Hua Yun Chen* and Xuelong Wang and Mary Turyk and Maria Argos and Vectoria Persky
|
Companies:
|
University of Illinois At Chicago and University of Illinois at Chicago and University of Illinois at Chicago and University of Illinois at Chicago and University of Illinois at Chicago
|
Keywords:
|
Bootstrap;
EigenPrism;
GCTA;
Random effects;
Random matrix;
Weak effects
|
Abstract:
|
Environmental pollutants are usually composed of mixtures of many chemicals that are highly correlated due to similar sources and/or similar chemical structures. Because of the usual low dosage levels of exposures to environmental pollutants, the effect of an individual chemical on a health outcome is often weak and difficult to detect. In addition, the health effects of the exposure to pollutants can be substantially heterogeneous despite of the structural similarity of the pollutants. It is therefore very challenging to detect the effects of a mixture of pollutants. We proposed to use historical data to account for exposure correlations in estimating the main and interaction effects of a mixture of pollutants. The proposed estimator overcomes the possible bias induced by the high correlation among the chemicals. Extensive simulation studies suggest the proposed approach can estimate the total variation of the health outcome explained by the pollutants mostly unbiasedly. The proposed approach is applied to the estimation of the total effects of PCBs on the standardized glycol-hemoglobin level in a subset of the NHANES data.
|
Authors who are presenting talks have a * after their name.