AI-Driven Tool Flags High Heart Failure Risk in Diabetic Patients

Researchers at UT Southwestern Medical Center have unveiled a machine learning model designed to identify diabetic patients at elevated risk for heart failure due to a distinct heart condition known as diabetic cardiomyopathy. Published in the European Journal of Heart Failure, this tool offers promise for early detection and targeted interventions to mitigate heart failure risks among diabetic patients—a group challenging to monitor due to the condition’s silent progression and diverse impacts on heart health.

Using data from more than 1,000 individuals with diabetes and no previous cardiovascular disease, the model categorizes patients based on heart health indicators, such as NT-proBNP levels and left ventricular structural changes. The team identified a high-risk subgroup, accounting for 27% of participants, marked by heightened cardiac stress markers and structural abnormalities. This subgroup demonstrated a notably high five-year heart failure rate of 12.1%, underscoring the model’s potential for early, risk-based intervention.

Validation studies using additional patient datasets confirmed the model’s capacity to flag up to 29% of diabetic patients as high-risk, strengthening its reliability as a diagnostic aid.

This research offers clinicians a new tool for personalized care in diabetic patients, with an evidence-based approach to identifying those who might benefit from early, intensive therapies, such as SGLT2 inhibitors. Targeting preventive treatments to this high-risk phenotype could help reduce heart failure rates and improve patient outcomes, potentially refining both treatment and prevention strategies for diabetes-related heart conditions.

By leveraging machine learning, this study helps address the complexities of diabetic cardiomyopathy and its often subtle early symptoms, supporting a more refined approach to cardiovascular care. With early identification, healthcare providers may be able to take more proactive steps to prevent heart failure, reshaping clinical strategies and advancing research in diabetes-related cardiovascular health.