A Glucose Management Indicator (GMI) of 6.3% is a CGM-derived estimate of your lab A1C. It means your 14-day average sensor glucose was roughly 125 mg/dL — firmly in the prediabetes zone. But GMI and lab A1C are not interchangeable, and differences of 0.3-0.5% between them are common and usually not a sign that either test is wrong.
What GMI Actually Is
GMI was introduced by Bergenstal and colleagues in 2018 as a replacement for the older “estimated A1C” calculation in CGM reports. It uses the formula:
GMI (%) = 3.31 + 0.02392 x mean CGM glucose (mg/dL)
The formula comes from a regression of simultaneous CGM and lab A1C data in hundreds of people with type 1 and type 2 diabetes. GMI shows up in Dexcom Clarity, FreeStyle LibreView, Tidepool, and similar platforms. It requires at least 10-14 days of CGM wear with over 70% data capture to be reliable.
Where 6.3% Falls
| GMI (%) | Mean CGM glucose (mg/dL) | Interpretation |
|---|---|---|
| 5.7 | 100 | Normal |
| 6.0 | 112 | Borderline |
| 6.3 | 125 | Prediabetes range |
| 6.5 | 133 | Diabetes range |
| 7.0 | 154 | Typical diabetes target |
| 8.0 | 183 | Elevated |
A GMI of 6.3% sits in what the ADA considers the prediabetes window (A1C 5.7-6.4%). See our primer on A1C levels for how this range fits into diagnostic thresholds.
Why GMI and Lab A1C Differ
Both metrics estimate average glucose, but they use different biology. Lab A1C measures glucose bound to hemoglobin over a red blood cell’s 120-day lifespan. GMI uses direct interstitial fluid readings averaged over 10-14 days. Here are the main reasons they diverge:
- Red blood cell lifespan varies: Longer-lived cells accumulate more glucose, raising lab A1C relative to GMI. Shorter-lived cells (hemolysis, recent blood loss) lower A1C.
- Time window: GMI is a 14-day snapshot; A1C spans 3 months. If glucose changed recently, the two will reflect different averages.
- CGM bias: Sensors can drift slightly, particularly at the low and high ends.
- Hemoglobin variants: Sickle trait, HbF, and certain thalassemias interfere with some A1C assays.
- Iron status and pregnancy: Iron deficiency raises A1C; pregnancy lowers it.
Bergenstal’s original paper reported that about 28% of participants had a GMI-A1C difference greater than 0.5%, and about 8% had gaps greater than 1.0%. Your individual relationship tends to be consistent — track the gap over multiple tests.
What to Do with a 6.3 GMI
A 6.3% GMI is a wake-up signal, not a diagnosis. The next steps:
- Get a lab A1C to confirm. Pair it with fasting plasma glucose.
- Review your CGM time-in-range. Aim for more than 70% of readings in 70-140 mg/dL for prediabetes-level goals.
- Identify the biggest post-meal spikes. Work on those meals first.
- Start the core lifestyle playbook: 150 min/week of movement, 5-7% weight loss if overweight, post-meal walks, less ultra-processed food.
- See our guide to reversing prediabetes for a longer treatment path.
Clinical Implications of GMI-A1C Discordance
When your GMI and lab A1C disagree by more than 0.3%, the direction of the gap carries treatment implications.
GMI Lower Than Lab A1C (High Glycators)
If your CGM says 6.3% but the lab says 6.8%, your red blood cells glycate faster or live longer than average. Clinically, this means:
- Treatment targets tied to A1C (such as the 6.5% or 7.0% thresholds) may be harder to hit even with excellent day-to-day glucose control.
- Post-meal spikes may be more consequential for you than the numbers suggest — chasing A1C by adding medication can push time-in-range into hypoglycemic territory.
- Clinicians increasingly individualize targets by combining CGM time-in-range (goal >70% at 70-180 mg/dL) with A1C rather than treating A1C in isolation.
GMI Higher Than Lab A1C (Low Glycators)
If your CGM says 6.3% but the lab says 5.9%, your red blood cells glycate more slowly or turn over faster. Implications:
- A “normal” lab A1C may understate real hyperglycemia, delaying diagnosis or treatment intensification.
- CGM time-in-range and average glucose become more reliable than A1C for tracking progress.
- Evaluate for causes of red blood cell turnover: iron deficiency recovery, reticulocytosis, hemolysis, hemoglobinopathies, or pregnancy.
Measurement Error vs Biological Gap
Before accepting a gap as biological, rule out the measurable sources of error:
- Insufficient CGM data: GMI requires at least 10-14 days of wear with above 70% sensor capture. Gaps, warm-up periods, and sensor failures distort the average.
- Calibration drift: Factory-calibrated sensors (Dexcom G6/G7, Libre 2/3) have mean absolute relative differences of 8-10%. Check your sensor against a fingerstick if the GMI seems off.
- Compression lows: Sleeping on a sensor can drop readings 20-40 mg/dL and pull GMI downward.
- Lab assay variability: NGSP-certified A1C methods have a coefficient of variation of 2-3%. Two labs can legitimately differ by 0.2%.
- Timing mismatch: A 14-day CGM average compared to an A1C drawn 6 weeks later reflects different metabolic windows — not a biological gap.
A consistent gap that persists across multiple lab A1Cs and CGM reports is likely biological and should guide how you and your clinician interpret future numbers.
GMI Is Not Always Better Than A1C
CGM bias, sensor dropouts, or short wear periods (less than 10 days) can make GMI unreliable. ADA still recommends lab A1C for diagnosis. Use GMI for real-time feedback between lab visits.
The Bottom Line
A 6.3% GMI points to average glucose around 125 mg/dL — prediabetes territory. Your lab A1C may be slightly higher or lower depending on red blood cell biology and assay factors. Use GMI to guide daily decisions and track progress between lab tests, and use lab A1C for diagnosis and 3-month monitoring.
