Differences between glucose management indicator (GMI) and hemoglobin A1c (HbA1c) are commonly observed in type 1 diabetes mellitus (T1DM). A cross-sectional analysis published in the Diabetes Care evaluated whether specific glucose excursion patterns contribute to this discordance.

The study included 611 adults with T1DM. Ninety-day continuous glucose monitoring (CGM) data were paired with HbA1c values obtained within ±15 days. Glucose excursions were quantified using the glucose rate increase detector (GRID) algorithm. Discordance was defined using GMI-to-HbA1c and updated GMI (uGMI)-to-HbA1c ratios.

Excursions with peak glucose ≥250 mg/dL and time to peak ≥90 minutes were associated with higher uGMI-to-HbA1c ratios. These associations remained consistent across CGM sensor types. For sensor type 1, β was 0.174 (95% CI, 0.147-0.201). For sensor type 2, β was 0.102 (95% CI, 0.068-0.136). All findings were statistically significant (P<0.001).

Adjustment for GRID-derived excursion metrics altered the relationships between glycemic indices and albuminuria and triglyceride-glucose (TyG) index. The effects differed by outcome and context. These adjustments enhanced the informativeness of GMI and uGMI but not HbA1c.

The findings indicate that frequent high and prolonged glucose excursions contribute to GMI–HbA1c discordance in T1DM.