Abstract 9924: Polygenic Risk Score for PR Interval is Associated with Increased Risk of Ischemic Stroke

Abstract

Atrial cardiomyopathy (AC) is an emerging risk factor for ischemic stroke (IS). We used polygenic risk scores (PRS) to study the association of the PR interval (PRI)-- an ECG marker for AC-- with IS and IS subtypes. We studied PRI, rather than other AC traits, because a validated multi-ethnic PRI-PRS exists. Methods: We calculated PRI-PRS for patients in the NINDS Stroke Genetics Network (SiGN) database, which contains well-characterized stroke patients and in which we have found an association of a PRS for atrial fibrillation with IS subtypes. We derived PRS for 10 GWAS thresholds after a clumping procedure to eliminate SNPs in linkage disequilibrium. SNPs were weighted by their beta coefficients from the GWAS summary statistics to generate PRS. We normalized by the number of SNPs included. We used logistic regression to compare PRI-PRS in stroke cases (n=13930) and non-stroke controls (n=49473). We controlled for sex and the first 10 principal components of genetic ancestry. We stratified by genetic ancestry and genotype platforms. Logistic regression analyses for each stratum were meta-analyzed using an inverse-variance weighting-based method, assuming fixed effects. Results and Conclusions: In the overall meta-analysis, PRI-PRS was weakly associated with increased risk of IS (p<0.02, only for one GWAS threshold). PRI-PRS was associated with increased risk of IS due to small artery occlusion (SAO) (p=0.032 and 0.033, using GWAS of PRI p-value thresholds 0.05 and 0.005 respectively), but not with other IS subtypes. The SAO association may reflect the association of PRI and hypertension, a risk factor for SAO. In a secondary analysis of European strata, we observed a robust association (p<0.02 across all GWAS thresholds) of the PRI-PRS with all IS in GEOS, which consists of younger onset strokes (age of stroke onset < 50 years). PRI-PRS was also associated in GEOS (p<0.01 across all GWAS thresholds) with the subtype of strokes of undetermined origin, which includes embolic strokes of unknown source (ESUS). PRI-PRS was also associated with large artery strokes at some GWAS thresholds in GEOS. Our results suggest that genetic risk for prolonged PRI is associated with SAO and may have more robust associations with other stroke subtypes in some populations.