GMI (Glucose Management Indicator): How It's Calculated — and Its Limits

GMI (Glucose Management Indicator) is a projection of your A1C calculated from the average of the blood glucose readings you've logged. It uses a published formula that converts an average glucose value into a percentage on the same scale as A1C. It's a helpful, at-a-glance number for spotting trends between appointments, but it is an estimate, not a lab test, and it can differ from the A1C your doctor measures.

What is GMI?

Your A1C (also written HbA1c) reflects your average blood glucose over roughly the past two to three months, because glucose attaches to hemoglobin in your red blood cells for the life of those cells. A lab measures this chemically from a blood sample. GMI works differently: it takes the average of the glucose numbers you have recorded and converts that average, using a published formula, into the A1C-equivalent percentage those numbers correspond to.

GMI was introduced in a 2018 study led by Dr. Richard Bergenstal and colleagues, published in Diabetes Care, as a way to give people using frequent glucose monitoring a same-scale approximation of A1C without a blood draw. It replaced an older, informal term, "estimated A1C" or "eA1C", with a named, published metric and a standard formula, so that the number means the same thing across apps and devices.

Because it's built from your own logged readings, GMI is only as good as the data behind it. A full, evenly spread set of readings produces a meaningful number. A handful of readings all taken at the same time of day does not.

How GMI is calculated

The published formula, from the Bergenstal 2018 study, is:

GMI (%) = 3.31 + 0.02392 × average glucose (mg/dL)

If you log in mmol/L, multiply your average by 18 to get mg/dL first, then apply the same formula.

Gludiary applies this formula to the average of the readings you have logged. For example, an average glucose of 154 mg/dL (8.6 mmol/L) works out to a GMI of about 7.0%, and an average of 126 mg/dL (7.0 mmol/L) works out to a GMI of about 6.3%.

GMI to average glucose reference table

Values are calculated from the published GMI formula and are rounded. These are population-level conversions: your personal targets are set with your care team, and individual A1C goals vary. The ADA suggests a general A1C goal of below 7% for many adults, but goals are individualized: someone older, someone with frequent low blood sugars, or someone newly diagnosed may be given a different target. Never adjust medication or treatment based on an estimated number, that's a decision for your doctor.

GMI vs. the older "estimated A1C" (eA1C)

Before GMI was published, many apps and glucose meters showed a number sometimes called "estimated A1C" or "eA1C," calculated by inverting the ADAG (A1C-Derived Average Glucose) formula: eAG (mg/dL) = 28.7 × A1C − 46.7. That older approach and GMI both convert an average glucose into an A1C-style percentage, but they come from different studies and use slightly different formulas, so the two numbers can differ a little for the same average glucose. GMI is the newer, more widely adopted term, and it's what Gludiary now shows, but you may still see "estimated A1C" used elsewhere, and it's describing a closely related idea.

It's also worth knowing that GMI was originally validated using continuous glucose monitor (CGM) data: readings taken every few minutes, day and night, from devices such as a Freestyle Libre or Dexcom. That produces a very dense, unbiased average. A GMI built from fingerstick or manually logged readings, like the one Gludiary calculates, relies on the moments you chose to log, which may cluster around meals or mornings and miss the hours in between. The formula is the same either way, but the input data is not, which is part of why your in-app GMI and a lab A1C can diverge.

How to read your GMI sensibly

A single GMI value in isolation tells you little. What's useful is watching it over time and understanding what's driving it. A few habits make the number more trustworthy:

• Log across the whole day. Mixing fasting, pre-meal, post-meal, and bedtime readings gives a balanced average rather than one weighted toward a single moment.
• Give it enough data. Because A1C reflects two to three months, a GMI from at least several weeks of consistent logging is more meaningful than one from a few days.
• Watch the direction, not just the digit. A gentle downward trend over several weeks is more informative than any one reading, and it's less sensitive to a single unusual day.
• Compare it to your last lab result. If your GMI and your most recent lab A1C are close, that's reassuring. If they diverge sharply, that's a useful thing to raise at your next appointment.

Why GMI isn't the same as a lab A1C

A lab A1C and a GMI answer slightly different questions. The lab test measures the actual glycation of your hemoglobin over the red blood cell lifespan. GMI infers a number from a formula applied to your average glucose. Even with good data, expect them to be close rather than identical.

GMI is most likely to be misleading in a few common situations:

• A short data window. A1C reflects two to three months of glucose. If you've only logged a week or two of readings, your GMI is really a reflection of that short window, not the longer period a lab test captures.
• Uneven or biased reading times. If most of your readings are fasting, or most are after meals, the average is skewed and GMI drifts with it. A representative mix across the day gives a truer picture.
• Fingerstick sampling vs. continuous data. A few daily fingersticks can miss overnight lows or post-meal peaks that a continuous monitor would capture, so the calculated average, and the GMI built from it, may not match your true average.
• Conditions that affect red blood cells. Anemia, recent blood loss or transfusion, pregnancy, kidney disease, and certain hemoglobin variants can change how A1C relates to average glucose. In these cases even a lab A1C can be less reliable, and a GMI calculated from logged glucose can differ noticeably. This is a conversation to have with your doctor.

Treat GMI as a trend indicator between visits, not a diagnostic result. Only a lab test ordered and interpreted by your healthcare provider can confirm your A1C.

Seeing your GMI in Gludiary

Once you've logged readings for a while, Gludiary calculates a GMI from the average of those readings and shows it alongside your time in range and trend charts. Tagging each entry with its meal context (fasting, before a meal, after a meal, or bedtime) helps you (and the app) build a balanced average rather than one skewed toward a single time of day. When it's time for an appointment, you can export a PDF report so your doctor sees the readings behind the number. For a step-by-step walkthrough of the calculation and its limits inside the app, see the GMI guide. Remember: the app records and analyzes your readings. It doesn't measure your glucose, and its GMI never replaces a lab test.

Frequently asked questions

How accurate is GMI?

Its accuracy depends entirely on your logged data. With a broad, representative set of readings over a couple of months, GMI tends to land close to a lab A1C. With few readings, a short window, or readings clustered at one time of day, it can be off. Use it to watch direction and trends, and rely on the lab test for the confirmed number.

Can GMI replace a lab test?

No. GMI is a calculated projection, not a diagnostic measurement. Diagnosis, treatment decisions, and confirmed A1C values come from tests your doctor orders and interprets. GMI is a between-visit companion, not a substitute.

Why is my GMI different from my doctor's A1C result?

Common reasons include a short logging window, readings taken mostly at one time of day, fingersticks missing highs or lows a continuous monitor would catch, or medical conditions such as anemia that change how A1C relates to average glucose. If the gap is large or persistent, bring your logs to your doctor and ask about it.

Is GMI the same as "estimated A1C"?

They're closely related. Both convert an average glucose into an A1C-style percentage, but GMI is a specific, published formula (Bergenstal et al., 2018), while "estimated A1C" or "eA1C" usually refers to an older, informal calculation based on inverting the ADAG eAG formula. The two can give slightly different numbers for the same average glucose. Gludiary uses the published GMI formula.

Want to understand the underlying math? See our A1C to average glucose chart, and if you track in different units, our mmol/L to mg/dL converter explains the ×18 factor.