Abstract 560: Genetics Of Physical Activity And Risk Of Cardiovascular Disease

Abstract

Objective: Sedentary lifestyle with minimal physical activity is a well-established risk factor for poor metabolic health outcomes. Atherosclerotic cardiovascular diseases (ASCVD) are often thought to be associated with physical inactivity, although observational studies may be limited by residual confounding which makes causal inference challenging. Mendelian Randomization (MR) is a method which leverages the naturally random allocation of genetic variants as a natural experiment, permitting causal inference. To better characterize sedentary lifestyle behaviors, we performed a genome-wide association study (GWAS) of sedentary behaviors using genomic structural equation modeling (SEM) and conducted a MR study using this phenotype to characterize the relationship of physical inactivity to ASCVD. Methods: Genomic SEM was performed to integrate three separate GWAS of physical activity and sedentary behavior into a common factor model. Following this, two-sample univariate and multivariable MR testing the effect of sedentary behavior on ASCVD traits was performed. Outcome data was taken from the Million Veteran Program (MVP) for PAD and CAD, and the GIGASTROKE consortium for stroke. Results: The common factor GWAS of sedentary behavior identified 44 genomic risk loci. 32 of the 44 lead variants were not significantly associated with the original sedentary traits, pointing toward possible new genetic mechanisms. On gene-based testing, the CADM2 gene was implicated as the strongest risk locus of 156 mapped genes; this gene has been associated with impulsivity, substance use and obesity in prior studies. MR experiments revealed sedentary behaviors were associated with a significantly increased risk of CAD, PAD, and ischemic stroke; these effects were no longer significant when controlling for smoking, BMI, and diabetes as confounders. Conclusions: This study takes the largest available GWASs of PAD, CAD, and Stroke, and compares them against a novel GWAS of sedentary behavior traits. The findings suggest that the effects of sedentary behavior on ASCVD traits are likely not independent from other contributing factors (smoking; BMI), but that there is a shared causal pathway between these traits that has implications on the development of ASCVD.