Why the study?
How pathway-specific polygenic risk score-based stratification strategies inform therapeutic selection and comorbidity risk in type 2 diabetes remains unclear.
Does a cluster-based polygenic risk score approach better stratify clinical traits and comorbidity risks in patients with Type 2 Diabetes compared to an extreme-value approach?
Population
19,734 East Asian patients with T2D
Comparison
Cluster-based K-means strategy vs extreme-value (top 20%) pPRS strategy
Design
Cohort study
Key result
A cluster-based polygenic risk score approach better stratified 19,734 T2D patients into 7 clinical subgroups with distinct traits and medication associations (e.g., OR 1.98 for injectables; P<.0001).
Authors
Loading...
May refine T2D subtyping in East Asians; hypothesis-generating and should not yet change practice.
Cohort (n=19,734)
Does a cluster-based polygenic risk score approach better stratify clinical traits and comorbidity risks in patients with Type 2 Diabetes compared to an extreme-value approach?
p-value: p=<.0001
A cluster-based approach using multiple pathway-specific polygenic risk scores better stratifies Type 2 Diabetes patients by clinical traits and comorbidity risks than a simple extreme-value cutoff.
SHEU et al. (2026) conducted a cohort in Type 2 diabetes (n=19,734). Cluster-based K-means polygenic risk score stratification vs. Extreme-value strategy (top 20% of each pPRS) was evaluated on Separation in clinical traits and associations with medication use and comorbidities (p=<.0001). A cluster-based polygenic risk score approach better stratified 19,734 T2D patients into 7 clinical subgroups with distinct traits and medication associations (e.g., OR 1.98 for injectables; P<.0001).