uGMI: A More Accurate Alternative to HbA1c for Diabetes Management

Background

Hemoglobin A1c (HbA1c) has long been the gold standard for assessing long-term glucose control in diabetes management. However, continuous glucose monitoring (CGM)-derived glucose management indicator (GMI) was introduced as a complementary metric. Despite its utility, the current GMI often over- or underestimates HbA1c, leading to clinical challenges in interpretation and treatment adjustments.

Study Design

The study evaluated a revised GMI model, termed uGMI, using data from 18,860 individuals with 26,647 AG-HbA1c pairs. The analysis spanned clinical trial and real-world datasets. Key comparisons included the alignment between AG, HbA1c, and the proposed uGMI, assessed across 100 equal-sized data bins ranked by their combined positions.

Key Findings

The uGMI demonstrated superior alignment with HbA1c compared to the traditional GMI. Specifically, the regression slope between HbA1c and GMI improved from 1.4 to 1.0, indicating a reduction in proportional bias (p<0.0001). Absolute bias was notably reduced at both low (75 mmol/mol or 9.0% NGSP) HbA1c levels, decreasing from >4.4 mmol/mol (0.4% NGSP) to ≤1.1 mmol/mol (0.1% NGSP).

Expert Commentary

The uGMI addresses a critical gap in diabetes care by providing a more accurate reflection of glycemic control. Its consistency across different CGM devices (Abbott Freestyle Libre, Dexcom) enhances its clinical utility, suggesting broad applicability in diverse patient populations.

Conclusion

The uGMI represents a significant advancement in diabetes management, offering a more reliable metric for assessing glycemic control. Its adoption could enhance clinical decision-making and patient outcomes, particularly in cases where traditional GMI and HbA1c discordance has posed challenges.