A prediabetes-specific cardiovascular (CV) risk model demonstrated higher predictive performance for CV death than existing SCORE2 tools in adults with prediabetes and established cardiovascular disease (CVD), according to a study published in Cardiovascular Diabetology.
The analysis included 5,636 adults aged 45 years or older with prediabetes, defined as HbA1c 39-47 mmol/mol (5.7%-6.4%), along with established CVD and overweight or obesity. Using binary logistic regression with stratified threefold cross-validation, the study developed a prediction model based on seven routinely collected demographic and clinical variables.
Findings
- Over a mean follow-up duration of 3.3 years, 182 participants (3.2%) experienced CV death.
- Mean time to CV death was 2.0 years.
- The prediabetes-specific model demonstrated moderate discrimination for CV death prediction, with an ROC AUC of 0.730 (95% CI, 0.659-0.801).
- SCORE2 and SCORE2-Diabetes demonstrated lower predictive performance for CV death in this cohort, with ROC AUC values of 0.630 and 0.643, respectively.
- Existing models also showed lower performance for predicting major adverse cardiovascular events (MACE) in this population compared with their original validation cohorts.
The findings suggest that prediabetes-specific risk prediction models may improve identification of individuals at elevated CV risk, although external validation remains necessary before clinical implementation.
