Cardiovascular disease (CVD) is the leading cause of mortality in the United States. Twin and family-based studies have demonstrated a strong genetic contribution to a wide-array of CVD-related traits. Genome-wide association studies (GWAS) have identified thousands of genetic variants associated with CVD-related traits such as body mass index, lipid levels and hypertension. In aggregate these associated genetic variants explain only a small proportion of the overall genetic contribution to these traits. The interplay between genes and environmental exposures is likely to play a substantial role in explaining away some of these missing trait heritabilities. GWAS studies have largely ignored this consideration, despite the growing empirical evidence that such interactions are important. One major limitation has been that statistical power to detect these interactions is severely limited by the added complexity and dimensionality of studying such interactions. High multiple-test corrected significance thresholds from studying pair-wise interactions require interaction effects to be considerable and the sample size of the study cohort to be very large. In this study, we propose to apply a novel approach to reduce the dimensionality of this interaction problem. Prior to testing specific interactions, we propose to first identify a reduced set of variants that demonstrate some evidence, based on heteroscedasticity of genotype effects, for being subject to interaction. We also propose to use our unique knowledge of the public NHGRI genetic database dbGaP to identify, harmonize and combine genetic, phenotypic and environmental exposure data across large relevant genetic-epidemiological studies to increase our statistical power. The goal of this study is to identify important GxE interactions that impact CVD in order to provide greater insight into the molecular mechanisms of the disease and facilitate more targeted and more effective intervention strategies.

R21HL126045

2014-12-01

2018-11-30