Abstract
Primary open-angle glaucoma (POAG) remains the leading cause of irreversible blindness worldwide, with early detection crucial for preventing vision loss. However, current risk assessment methods have limited predictive power. Here, we present a multi-trait polygenic probability risk score (PPRS) approach that integrates multiple glaucoma-related traits and leverages functional genomic annotations to enhance POAG prediction across diverse ancestries. We constructed PRSs for POAG, intraocular pressure (IOP), vertical cup-to-disc ratio (VCDR), and retinal nerve fiber layer (RNFL) thickness using extensive genomic coverage (>7 million variants) and 96 functional annotations through the SBayesRC method. Validation in the UK Biobank (n=324,713, European ancestry) and Mexican American Glaucoma Genetic Study (MAGGS, n=4,549, Latino ancestry) demonstrated significant improvements in predictive accuracy over conventional approaches. Our multi-trait PPRS achieved area under the curve (AUC) values of 0.814 in Europeans and 0.801 in Latinos, compared to AUC ≤0.79 for single-trait models. We identified ancestry-specific differences in genetic contributions, with IOP demonstrating the strongest association in Europeans (OR=1.63, P = 5.37 × 10(-89)), while VCDR was predominant in Latinos (OR=1.64, P = 2.04 × 10(-11)). The model achieved remarkable risk stratification, with the highest PPRS decile showing 80.2-fold and 51.1-fold increased POAG risk in Europeans and Latinos, respectively, compared to the lowest decile. Importantly, the top PPRS quintile captured 65.9% and 62.2% of POAG cases in Europeans and Latinos, substantially improving upon previous approaches. Our findings demonstrate that integrating multiple disease-relevant traits and functional annotations significantly enhances polygenic prediction of POAG across diverse populations, with significant implications for targeted screening, early intervention, and reduction of disease burden.