Testing Gene-environment Interactions without Measuring the Environmental Factor

Funding Source: Center for Demography of Health and Aging, 2019–2020 Pilot Project  Read more

 Principal Investigator: Qiongshi Lu

Summary
Gene-environment interaction studies for late-life health outcomes often have limited sample sizes. While the size of genome-wide association studies (GWAS) grows rapidly for almost all complex diseases and traits, most of these studies have very limited measurements on epidemiological risk factors that are of interest in gene-environment interaction analysis. In this research, we explore the possibility of using PGS as a proxy of the epidemiological/ environmental risk factor (we refer to it as E-PGS) in gene-environment interaction analysis.

We will conduct simulations to investigate the validity of using E-PGS as a proxy for the ‘E component’ in gene-environment interaction analysis. Then we will conduct a pilot analysis focusing on three prevalent late-life diseases—coronary artery disease, stroke, and breast cancer. The Wisconsin Longitudinal Study (WLS) will be used as the primary discovery cohort, and we will replicate our findings in the Health and Retirement Study (HRS).

The main goal of the proposed work is to determine the feasibility of identifying gene-environment interactions in late-life health outcomes using E-PGS as a proxy. The successful identification of interactions would provide a strong basis for expanding our analysis to broader phenotypes.

We will conduct simulations to investigate the validity of using E-PGS as a proxy for the ‘E component’ in gene-environment interaction analysis. Then we will conduct a pilot analysis focusing on three prevalent late-life diseases—coronary artery disease, stroke, and breast cancer. The Wisconsin Longitudinal Study (WLS) will be used as the primary discovery cohort, and we will replicate our findings in the Health and Retirement Study (HRS).

The main goal of the proposed work is to determine the feasibility of identifying gene-environment interactions in late-life health outcomes using E-PGS as a proxy. The successful identification of interactions would provide a strong basis for expanding our analysis to broader phenotypes.