Objective:

To evaluate the effectiveness of multi-ancestry polygenic risk scores (PRSs) in predicting type 2 diabetes across diverse ancestry groups, highlighting their potential to improve risk assessment.

Key Findings:

• Multi-ancestry PRSs showed larger effect sizes than single-ancestry PRSs across all ancestry groups, with specific odds ratios provided.
• Odds of type 2 diabetes increased significantly with higher PRS values, particularly in the 97.5th percentile.
• Predictive performance varied by ancestry, with European and East Asian populations showing the strongest discrimination.
• Higher PRS values were associated with earlier onset of type 2 diabetes and increased odds of diabetes-related complications.

Interpretation:

Multi-ancestry PRSs can enhance risk stratification for type 2 diabetes and its complications, particularly in diverse populations, but disparities in predictive performance exist due to overrepresentation of European ancestry in GWAS data, which may limit applicability.

Limitations:

• Single-nucleotide polymorphism effect estimates are still heavily influenced by large European cohorts, potentially skewing results.
• Discrete ancestry categories may not fully capture genetic diversity, especially in admixed populations, which could affect the generalizability of findings.
• The study focused on predictive performance rather than clinical implementation, leaving the impact on patient outcomes uncertain.

Conclusion:

Validated multi-ancestry PRSs can improve risk stratification for type 2 diabetes onset and complications across diverse ancestries, with ongoing trials assessing their integration into routine care to ensure practical application.