Accurate CV risk prediction is essential in T2D, where heart disease remains a leading cause of morbidity and mortality. A large population-based analysis from Hong Kong, presented at the European Association for the Study of Diabetes (EASD) 2025, validated and compared the performance of four CV risk equations in adults with T2D.

The study included 234,846 adults with T2D and no prior CV disease identified through the territory-wide electronic health record system. Risk prediction was assessed using the Predicting Risk of Cardiovascular Disease Events (PREVENT) equations (Full and glycated hemoglobin [HbA1c] models), the Systematic Coronary Risk Evaluation 2 (SCORE2)-Diabetes, and the FRS. Model discrimination and calibration were evaluated, and pairwise comparisons were conducted using DeLong tests.

The PREVENT (Full) equation achieved the highest discriminative performance, with AUC 0.734 in men and 0.792 in women. Recalibration improved the observed-to-expected ratios to 1.057 in men and 1.036 in women, maintaining consistent discrimination.

These findings identify the PREVENT (Full) equation, which incorporates UACR and HbA1c, as the most accurate model for 10-year CV risk prediction in patients with T2D.