B-225 Translating GWAS into Clinical Practice: Real-World Utility of a Polygenic Risk Score for Diabetes Risk Stratification

Abstract Background Type 2 diabetes mellitus (T2DM) is a chronic metabolic disorder which arises from a complex interplay of genetic, environmental and behavioral factors. We evaluated the clinical utility of polygenic risk score (PRS)-based multigene panel test for predicting diabetes mellitus (DM) in a healthy population. Methods Overall, 302 individuals underwent genetic testing using the HelloGene™ DM panel, which comprises four DM-related SNPs (CDKAL1, HHEX, KCNQ1, TCF7L2). Each genotype was analyzed using a PRS algorithm based on a dataset of 70,000 Koreans, classifying participants into four risk groups (low, moderate, high, and very high). Fasting blood glucose (FBG), glycated hemoglobin (HbA1c), and body mass index were measured at baseline and after at least 3 years of follow-up. Smoking status, alcohol use, and exercise status were recorded. Results No significant differences in age, sex, or lifestyle habits were observed among the PRS groups. The ‘very high risk’ group had significantly higher 3-year follow-up FBG (p=0.001) and baseline/3-year follow-up HbA1c (p=0.025 and 0.001, respectively) levels (Table 1). The ‘very high risk’ group had a 4.5-fold higher risk of developing DM compared to other groups (Table 2A). Smoking was a significant modifier of genetic risk; smokers in the ‘very high-risk’ group had a 25% greater likelihood of developing DM. CDKAL1, HHEX, and TCF7L2 genotype carriers were prevalent in the individuals diagnosed with DM, while HHEX carriers in the ‘high risk’ group showed the greatest susceptibility, especially among current smokers (Table 2B). Conclusion PRS-based genetic testing demonstrated clinical utility in DM risk prediction. Elevated FBG and HbA1c levels in the high PRS group emphasize the importance of early identification. Smoking significantly increases genetic risk, underscoring the need for targeted interventions in genetically susceptible individuals.