News Research Glucose Monitoring & Insulin Delivery Research and Evidence Predictive Risk Models

AACE 2026: How machine learning models predict hemoglobin A1c response 

Tree-based approaches may help identify patients at risk for poor glycemic control.   

April 28, 2026 By Matthew Solan 2 min read
Share

5 Key Takeaways

  • 1

    Machine learning models outperformed logistic regression in predicting glycemic response in type 2 diabetes patients.

  • 2

    The study evaluated patients' HbA1c improvement using routine clinical data from a tertiary diabetes center in India.

  • 3

    Tree-based models achieved AUCs of 0.69 for random forest and 0.67 for XGBoost, indicating better performance than logistic regression.

  • 4

    Baseline HbA1c and early glycemic measures were identified as the strongest predictors of HbA1c response.

  • 5

    The models may aid in identifying high-risk patients for inadequate control, though they are not yet suitable as standalone clinical tools.

AACE Endocrine AI is published by Conexiant under a license arrangement with the American Association of Clinical Endocrinology, Inc. (AACE®). The ideas and opinions expressed in AACE Endocrine AI do not necessarily reflect those of Conexiant or AACE. For more information, see Policies.

Related Content