|
Activity Number:
|
32
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 3, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
ENAR
|
| Abstract - #301062 |
|
Title:
|
Assessing Gene-Environment Interaction for Case-Control Study: Without Independence and Rare Disease Assumptions
|
|
Author(s):
|
Bin Huang*+ and Altaye Mekibib and Siva Sivaganisan
|
|
Companies:
|
Cincinnati Children's Hospital Medical Center and Cincinnati Children's Hospital Medical Center and University Cincinnati
|
|
Address:
|
Center for Epi/Bio, MLC 5041, Cincinnati, OH, 45229,
|
|
Keywords:
|
Gene-Environmetal Interaction ; Rare Disease ; Relative Risk ; Case-Control Study ; Bias ; Efficiency
|
|
Abstract:
|
The Gene-Environment (GE) interaction refers to "a different effect of an environmental exposure (genotype) on disease risk in persons with different genotypes (environmental exposure)" (Ottman 96). Estimating GE interaction in a case-control study setting has been a subject of many resent studies. Existing methods often assume rare disease or gene-environmental independence assumption. Previous studies focused on relaxing independence assumption. In this study, We suggest that the commonly used ordinal logistic regression approach is sensitive to the rare disease assumption, even when prevalence of disease is small. A new measure is proposed that is able to incorporate external knowledge of the P(D), is simple to use, and is consistent with the existing measure under the rare disease or independence assumption. Simulation and case studies demostrated better bias and efficiency.
|
