Beyond the Mean: A Machine Learning-Based Trend Analysis of Continuous Glucose Monitoring (CGM) Metrics for Improved HbA 1C Estimation in Type 2 Diabetes
Authors' conclusion
Continuous glucose monitoring-based HbA1C estimation improves when variability and temporal patterns are included. Nighttime hyperglycemia adds notable predictive value, though further validation is needed.
Study context
- Population
- Participants with T2D from a three-month RCT, with ≥70% CGM data coverage and valid end-of-trial HbA1c; 51 metrics from a standardized 90-day CGM window
- Participants
- 159 participants
- Comparison
- Multi-metric CGM models vs benchmark models (mean glucose and glucose management indicator (GMI)) for estimating laboratory HbA1c
- Primary endpoint
- Estimation of laboratory-measured HbA1c (R-squared)
- Follow-up
- End of trial
Methodological detailsmachine learning modelling study, threefold cross-validated multiple linear regression, forward and exhaustive feature selection
- Machine learning modelling study
- Threefold cross-validated multiple linear regression
- Forward and exhaustive feature selection
Primary result
As reported in the abstract
R-squared for HbA1c estimation
Benchmark models
0.53
R-squared
Forward selection model
0.60
R-squared
Best-performing model
0.61
R-squared
Original publication
Beyond the Mean: A Machine Learning-Based Trend Analysis of CGM Metrics for Improved HbA 1C Estimation in Type 2 Diabetes. J Diabetes Sci Technol. Published 8 May 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →