Abstract 10666: Epigenetic Risk Score Capturing Alcohol Consumption is Not Associated with Cardiac Structure and Function: The Framingham Heart Study

Abstract

Habitual, heavy alcohol consumption increases the risk of detrimental outcomes such alcoholic dilated cardiomyopathy; however, alcohol’s complex effects on cardiac structure and function are incompletely understood. We previously identified 144 differentially methylated CpGs associated with alcohol consumption and sought to create an epigenetic risk score (ERS) derived from these CpGs to examine 1) the cross-sectional association of the ERS with cardiac structure and function, and 2) longitudinal association of the ERS with incident heart failure and atrial fibrillation in Framingham Heart Study (FHS) participants (n=1,843). We first conducted analyses examining the cross-sectional association between each of the 144 alcohol-associated CpGs and the outcomes listed above. After application of Bonferroni correction (p=3.47E-04), two CpGs (cg07504977, cg16246545) were significantly associated with left ventricular posterior wall thickness (p=2.69E-05 and p=1.75E-04, respectively) and one CpG (cg11682350) was significantly associated with left ventricular global longitudinal strain (p=2.15E-04). cg07504977 annotates to the promoter of a lcRNA LINC00623, while cg16246545 maps to the PHGDH , a protein-coding gene implicated in serine synthesis. cg11682350 is not annotated to a gene. Cross-sectional analyses examining the association between the ERS and left ventricular mass (BSA-indexed; p=0.15) systolic diameter (p=0.92), diastolic diameter (p=0.87), fractional shortening (p=0.95), and global longitudinal strain (p=0.28), were not statistically significant, while ERS association with LV posterior wall thickness was nominally significant (p=0.03). Longitudinal associations for HF (p=0.18) and AF (p=0.65) were not significant. Though most of our results are not statistically significant, we plan to explore other means of generating alcohol ERS including by LASSO penalty coefficient thresholds, Gene Ontology biological pathways, and weighted gene expression network analyses. In conclusion, such alcohol-related ERS may be examined in relation to other phenotypes and serve as a predictive or diagnostic tool to monitor and elucidate the epigenetic pathways connecting environmental exposures and disease.