Polygenic Risk Score Portability Across Ancestries for Coronary Artery Disease: Validation in African, South Asian, and East Asian Biobank Cohorts and Ancestry-Specific Score Recalibration

This study investigates portability of European-trained coronary artery disease polygenic risk scores in African, South Asian, and East Asian biobank cohorts, and ancestry-specific recalibration to restore predictive performance within the context of statistical genetics and cardiovascular precision medicine, an area of growing scientific importance given its implications for equitable genomic medicine implementation, ancestry-diverse GWAS investment strategy, and CAD risk stratification in non-European clinical populations. Using GWAS summary statistic PRS derivation (LDpred2, PRSice-2), C-statistic and OR per SD comparison across ancestry cohorts, and ancestry-specific PRS recalibration using local LD reference panels and within-ancestry GWAS fine-mapping, we examine EUR LD reference underpinning PRS derivation poorly representing non-EUR LD blocks, combined with ancestry-specific effect size heterogeneity and causal variant tagging differences, reducing cross-ancestry PRS transferability in 69,600 participants across 3 non-EUR cohorts: 18,400 African Ancestry (4,284 CAD cases), 22,800 South Asian (5,840 cases), 28,400 East Asian (6,840 cases) with genome-wide array genotyping and electronic health record CAD ascertainment drawn from African Ancestry cohort from Million Veteran Program; South Asian from UK Biobank + PMBB; East Asian from BioBank Japan + KCHIP; all with linked EHR CAD outcomes. Results indicate that EUR CAD PRS AUC drops from 0.74 (EUR validation) to 0.58 in African Ancestry, 0.64 in South Asian, 0.66 in East Asian; ancestry-specific recalibration restores AUC to 0.68, 0.70, and 0.71 respectively, partially closing but not eliminating the gap (p < 0.001), with AUC 0.58/0.64/0.66 non-EUR vs. 0.74 EUR; recalibration restores to 0.68/0.70/0.71 as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to statistical genetics and cardiovascular precision medicine and carry actionable implications for the design of programs and policies targeting equitable genomic medicine implementation, ancestry-diverse GWAS investment strategy, and CAD risk stratification in non-European clinical populations.