The T-PDM-CVD model effectively predicted 10-year risks in prediabetes, with the isolated high-T2D subgroup having the highest risk of developing diabetes (HR 3.31; 95% CI 3.13-3.50).
Cohort (n=65,676)
Yes
Can the T-PDM-CVD model accurately predict the 10-year risks of type 2 diabetes and cardiovascular disease in individuals with prediabetes?
The T-PDM-CVD model provides accurate simultaneous prediction of 10-year risks for type 2 diabetes and cardiovascular disease in individuals with prediabetes, enabling enhanced risk stratification.
Hazard Ratio: 3.31 (95% CI 3.13–3.5)
Introduction and Objective: This study aimed to develop and validate a novel dual-endpoint model to simultaneously predict 10-year risks of type 2 diabetes (T2D) and cardiovascular disease (CVD) in individuals with prediabetes, enabling more comprehensive risk stratification. Methods: We included 18,848 participants with prediabetes from the Kailuan cohort for model derivation and internal validation, and 46,828 participants with prediabetes from the UK Biobank for external validation.The T-PDM-CVD (Transfomer model for diabetes and CVD progression) was developed using 10 baseline variables, including demographic and laboratory data. Participants were stratified into four subgroups using a 2×2 framework based on predicted risks of progression to diabetes and CVD. Results: In internal validation, T-PDM-CVD achieved AUCs of 0.72 (95% CI: 0.70-0.73) for T2D and 0.64 (95% CI: 0.63-0.65) for CVD. In external validation, AUCs improved to 0.80 (95% CI: 0.79-0.81) for T2D and 0.75 (95% CI: 0.74-0.76) for CVD.Compared to the low T2D-CVD risk group, the isolated high-T2D subgroup had the highest risk of developing diabetes (HR 3.31, 95% CI 3.13-3.50), and the isolated high-CVD subgroup had the highest cardiovascular risk (HR 1.93, 95% CI 1.77-2.10). The dual high-risk subgroup showed concurrent but attenuated risks for T2D (HR 1.54, 95% CI 1.46-1.64) and CVD (HR 1.23, 95% CI 1.15-1.31), with effect sizes lower than those observed in the corresponding isolated high-risk groups. Conclusion: The T-PDM-CVD model enables simultaneous prediction of T2D and CVD risks in prediabetes, providing enhanced risk stratification to inform personalized prevention strategies. Disclosure S. Wang: None. X. Liu: None. Q. Huang: None. K. Meng: None. S. Wu: None. L. Ji: None. W. Hu: None. X. Zou: None. Funding National Natural Science Foundation of China (T2341011), Beijing Nova Cross Program of Science and Technology (20250484806)
Wang et al. (Sat,) conducted a cohort in Prediabetes (n=65,676). Isolated high-T2D risk subgroup (T-PDM-CVD model) vs. Low T2D-CVD risk group was evaluated on Development of type 2 diabetes (HR 3.31, 95% CI 3.13-3.50). The T-PDM-CVD model effectively predicted 10-year risks in prediabetes, with the isolated high-T2D subgroup having the highest risk of developing diabetes (HR 3.31; 95% CI 3.13-3.50).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: