Multi-ancestry polygenic risk scores for the prediction of type 2 diabetes and complications in diverse ancestries
Authors' conclusion
These validated, publicly available PRSs can improve risk stratification for type 2 diabetes onset and complications across diverse ancestries
Study context
- Population
- Individuals with T2D and controls from genome-wide association studies (GWAS) cohorts of five global ancestries409,959 T2D, 1,983,345 controls
- Comparison
- Multi-ancestry polygenic risk scores (PRSs) vs single-ancestry PRSs; PRS percentiles vs IQR
- Outcome
- Association of PRS with T2D: OR per SD of the PRS
Methodological detailsfive ancestries, single-ancestry PRS-CS and multi-ancestry PRS-CSx, ancestry-specific linkage disequilibrium panels, validation in ≥4 independent cohorts per ancestry, GWAS dataset: 359 819 T2D and 1 825 729 controls; training: 10 992 and 31 792; validation: 39 148 and 125 824
- Five ancestries
- Single-ancestry PRS-CS and multi-ancestry PRS-CSx
- Ancestry-specific linkage disequilibrium panels
- Validation in ≥4 independent cohorts per ancestry
- GWAS dataset: 359 819 T2D and 1 825 729 controls; training: 10 992 and 31 792; validation: 39 148 and 125 824
Pooled result
As reported in the abstract
T2D, multi-ancestry PRS, OR per SD
| Group | OR per SD |
|---|---|
| African or African American | 1.73 (95% CI 1.67-1.80) |
| Admixed American | 2.82 (2.67-2.97) |
| East Asian | 2.45 (2.36-2.54) |
| European | 2.36 (2.32-2.41) |
| South Asian | 2.23 (2.05-2.42) |
OR per SD
African or African American1.73 (95% CI 1.67-1.80)
Admixed American2.82 (2.67-2.97)
East Asian2.45 (2.36-2.54)
European2.36 (2.32-2.41)
South Asian2.23 (2.05-2.42)
Further pooled results
| Outcome | Values | OR (95% CI) | |
|---|---|---|---|
| T2D, 97.5th PRS percentile vs IQR97.5th percentile (African or African American)vs IQR | 3.43 (2.80–4.21) | ||
| T2D, 97.5th PRS percentile vs IQR97.5th percentile (Admixed American)vs IQR | 7.47 (5.64–9.89) | ||
| T2D, 97.5th PRS percentile vs IQR97.5th percentile (East Asian)vs IQR | 6.62 (5.58–7.85) | ||
| T2D, 97.5th PRS percentile vs IQR97.5th percentile (European)vs IQR | 6.25 (5.72–6.82) | ||
| T2D, 97.5th PRS percentile vs IQR97.5th percentile (South Asian)vs IQR | 4.50 (2.70–7.53) | ||
| 0.10.31310Odds ratio (log scale) |
| Outcome | Values |
|---|---|
| Single-ancestry PRS, incremental AUCCorrelating with GWAS sample sizes | European 0.07-0.14 · East Asian 0.02-0.16 · African or African American 0.02-0.03 · Admixed American 0.02-0.04 · South Asian 0.02-0.04 |
T2D, 97.5th PRS percentile vs IQR
- 3-7 times increased risk of T2D
Onset, complications and added value beyond clinical factors
- PRSs associated with earlier T2D onset, higher risk of microvascular complications, and additional predictive value beyond clinical factors
Microvascular complications and comorbidity in T2D (African, Admixed American, European ancestries), OR per SD
- Diabetic retinopathy
- 1.28-1.57
- Diabetic nephropathy
- 1.25-1.58
- Proliferative diabetic retinopathy
- 1.39-2.08
- End-stage diabetic nephropathy
- 1.44-1.87
- Coronary artery disease, Admixed American only
- 1.16 [95% CI 1.08-1.25]
- Significant in all three ancestry groups for retinopathy and nephropathy outcomes
Original publication
Multi-ancestry polygenic risk scores for the prediction of type 2 diabetes and complications in diverse ancestries. Lancet Diabetes Endocrinol. Published 27 April 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →