GMI (Glucose Management Indicator) is a formula developed to estimate A1C from CGM (continuous glucose monitor) average glucose data. It was developed by researchers at Jaeb Center for Health Research and published in Diabetes Care in 2018.
The key advantage of GMI over eAG: it uses a formula specifically validated on CGM data, which captures 24-hour glucose patterns more completely than home meter readings. Traditional eAG (from lab A1C) was developed before CGMs were widely used. GMI gives CGM users a more accurate A1C estimate without waiting for a lab test.
GMI is now reported automatically by many CGM platforms including Dexia Clarity, Libre View, and Medtronic CareLink.
GMI Formula
The GMI formula, developed by Bergenstal et al. (2018) and validated against lab A1C in people with type 1 and type 2 diabetes:
GMI (%) = 3.31 + (0.02392 × mean glucose in mg/dL)To convert mmol/L → mg/dL first: multiply by 18.016Source: Bergenstal et al., Diabetes Care 2018
Example: A CGM average of 154 mg/dL → GMI = 3.31 + (0.02392 × 154) = 3.31 + 3.68 = 6.99% ≈ 7.0%
GMI vs eAG Formula
Formula
Equation
Best for
GMI
3.31 + (0.02392 × avg mg/dL)
CGM users
eAG → A1C
(eAG + 46.7) ÷ 28.7
Lab-based eAG
Why GMI and Lab A1C May Differ
GMI and lab A1C are correlated but often don't match exactly. Research shows that GMI tends to be lower than lab A1C in many individuals — sometimes by 0.3%–0.8%. This is expected and doesn't mean one is "wrong." Key reasons for the difference:
Red blood cell lifespan: Lab A1C reflects glucose attached to all red blood cells over their entire 2–3 month lifespan. CGM only captures data during the sensor wear period.
Hemoglobin variants: Certain hemoglobin types affect lab A1C but not CGM readings — making GMI more reliable for people with conditions like sickle cell trait.
Glucose variability: CGM captures time-in-range and glucose swings that a single average can't represent — but GMI still uses just the average.
Sensor accuracy: CGM sensors have calibration variability that can shift the average slightly.
The ADA recommends using GMI as a complementary metric alongside Time-in-Range (TIR), not as a direct replacement for lab A1C in clinical decision-making.
Frequently Asked Questions
What is a good GMI?
The same targets used for lab A1C apply to GMI. A GMI below 7.0% is the ADA's standard target for most adults with diabetes — this corresponds to a CGM average glucose of approximately 154 mg/dL (8.6 mmol/L). For people without diabetes, a GMI below 5.7% (average glucose below 117 mg/dL) is normal. Always discuss your GMI target with your healthcare provider, as individual goals vary.
Why is my GMI lower than my lab A1C?
GMI tends to run 0.3%–0.8% lower than lab A1C for many people — this is a known and expected difference, not an error. Lab A1C reflects glucose attached over the full red blood cell lifespan (up to 3 months), while GMI is based on your CGM average over the sensor wear period. Additionally, some people have red blood cell characteristics (like longer lifespan) that cause lab A1C to read higher relative to actual average glucose.
Can GMI replace the A1C blood test?
Not for clinical diagnosis or official monitoring. GMI is an estimate and is not FDA-approved as a diagnostic tool. Lab A1C remains the gold standard for diabetes diagnosis and many clinical decisions. However, GMI provides valuable real-time feedback between lab tests — especially for CGM users who want to understand how their daily glucose patterns translate to an A1C-like estimate without waiting for a blood draw.
How many days of CGM data do I need for an accurate GMI?
The ADA and CGM manufacturers recommend at least 14 days of CGM data for GMI to be meaningful. More data (30+ days) improves accuracy. GMI calculated from fewer than 7 days of data is generally not reliable. Most CGM apps will only display GMI automatically once you have sufficient data and adequate sensor coverage (typically above 70% time active).
Is GMI the same as eAG?
No — they use different formulas and have different purposes. eAG (estimated Average Glucose) is calculated from a lab A1C test using the ADA ADAG formula: eAG = (28.7 × A1C%) − 46.7. GMI goes the other direction: it estimates an A1C-like percentage from CGM average glucose using a formula specifically validated for CGM data. Both express similar information but are calculated from different starting points and use different formulas.