What is the clinical evidence from this study?

Study design: Cohort. Population: type 2 diabetes (n=1363). Intervention: Glucose management indicator (GMI) and time in range (TIR) vs. HbA1c. Primary outcome: serious CVD (MACE, heart failure hospitalization, angina, revascularization) (HR 1.34, 95% CI 1.09-1.65).

What does this research mean for the field?

Continuous glucose monitoring metrics, such as Glucose Management Indicator and Time in Range, predict cardiovascular and mortality outcomes in type 2 diabetes comparably to HbA1c, but do not offer independent or additive prognostic value. Novelty: ClaimNovelty.INCREMENTAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

This study investigates whether CGM metrics can predict cardiovascular disease and mortality outcomes beyond HbA1c in individuals with type 2 diabetes.

June 7, 2026

2962-LB: CGM Metrics and Cardiovascular and Mortality Outcomes in Type 2 Diabetes: Data from the GRADE-CGM Substudy

Key Result

Higher CGM-derived glucose management indicator predicted serious cardiovascular disease (HR 1.34; 95% CI 1.09-1.65), but its prognostic value was not independent of HbA1c.

Key Points

This study investigates whether CGM metrics can predict cardiovascular disease and mortality outcomes beyond HbA1c in individuals with type 2 diabetes.
Analyzed data from 1,363 participants in the GRADE study with 10-day CGM data.
Outcomes included serious cardiovascular events and all-cause mortality.
Cox models estimated hazard ratios for TIR and GMI with HbA1c as a comparator.
Higher GMI predicted serious cardiovascular events (HR 1.34 [1.09-1.65]) and mortality (HR 1.36 [1.02-1.83]).
Higher HbA1c also predicted cardiovascular events (HR 1.30 [1.06-1.59]) and mortality (HR 1.42 [1.04-1.94]).
Higher TIR was protective for cardiovascular disease (HR 0.78 [0.62-0.98]) but not associated with death.

Study Design

Type

Cohort (n=1,363)

Structured PICO

Do CGM metrics (TIR and GMI) predict serious CVD and mortality beyond HbA1c in patients with type 2 diabetes?

Population

1,363 participants with type 2 diabetes and 10-day CGM data from the GRADE study, followed for a median of 2.3 years.

Exposure

CGM-derived time in range (TIR) and glucose management indicator (GMI)

Comparator

HbA1c

Outcome

Serious CVD (MACE, heart failure hospitalization, angina, revascularization) and all-cause mortalitycomposite

CGM-derived GMI and HbA1c comparably predict cardiovascular and mortality outcomes in type 2 diabetes, but their prognostic values overlap and are not additive.

Main Result

Hazard Ratio: 1.34 (95% CI 1.09–1.65)

Abstract

Introduction and Objective: Whether CGM metrics predict long-term cardiovascular disease (CVD) and mortality outcomes beyond HbA1c in type 2 diabetes is unclear. We examined the prognostic value of CGM-derived time in range (TIR) and glucose management indicator (GMI) vs. HbA1c for serious CVD and all-cause death. Methods: We studied 1,363 participants with 10-day CGM in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness (GRADE) study. Outcomes include serious CVD (MACE, heart failure hospitalization, angina, revascularization) and all-cause mortality. Cox models estimated hazard ratios (HR) for TIR and GMI (per 1-SD), with HbA1c as comparator. Model 1 adjusted for age, sex, race/ethnicity, treatment, and diabetes duration. Model 2 mutually adjusted a CGM metric and HbA1c. Results: Over median 2.3 years, 59 CVD events and 26 deaths occurred. In Model 1, higher GMI predicted serious CVD (HR 1.34 1.09-1.65) and death (HR 1.36 1.02-1.83); higher HbA1c similarly predicted CVD (HR 1.30 1.06-1.59) and death (HR 1.42 1.04-1.94). Higher TIR was protective for CVD (HR 0.78 0.62-0.98) but not death. In Model 2, all significant associations were attenuated after mutual adjustment (Figure). Conclusion: GMI and HbA1c comparably predicted CVD and mortality but with overlapping rather than additive prognostic value. TIR was protective for CVD but not independent of HbA1c nor associated with mortality. Disclosure M. Tang: Research Support; Ended; Dexcom, Inc., American Heart Association. D.M. Nathan: Advisory Panel; Ended; Lexicon Pharmaceuticals, Inc. D. Wexler: Other - Data Monitoring Committee; Ended; Novo Nordisk. Other - Data Monitoring Committee; Current; Amgen Inc. M.S. Putman: Advisory Panel; Current; Vertex Pharmaceuticals Incorporated. Research Support; Ended; Vertex Pharmaceuticals Incorporated, Dexcom, Inc. Advisory Panel; Ended; Sanofi. Board Member; Current; Anagram Therapeutics. Research Support; Ended; Samsung. I. de Boer: Consultant; Current; Boehringer Ingelheim International GmbH, Novo Nordisk, GlaxoSmithKline plc., Lilly. Consultant; Ended; Dexcom, Inc. Research Support; Current; Novo Nordisk, Dexcom, Inc. Consultant; Ended; AstraZeneca. Consultant; Current; Roche Diabetes Care. Consultant; Ended; Lexicon Pharmaceuticals, Inc. S. Kalim: None.

Bookmark