What Is GMI on Libre 3? A1C Estimate From Your CGM

GMI, or Glucose Management Indicator, is the number on your FreeStyle Libre 3 app that estimates what your lab A1C would be based on your average sensor glucose readings. It uses a straightforward formula that converts your CGM-measured mean glucose into a percentage on the same scale as a traditional HbA1c blood test. The number gives you a real-time window into your glucose management without waiting for a lab draw, but it comes with an important caveat: GMI and lab A1C measure fundamentally different things, so the two numbers often don’t match perfectly.

How GMI Gets Calculated

Your Libre 3 sensor reads glucose in the fluid just under your skin roughly every minute, generating a massive stream of data. GMI takes the average of all those glucose readings over a reporting period and converts it to a percentage using a published formula. The equation was developed from clinical trials that paired the most accurate CGM systems available with simultaneous lab A1C measurements to build a reliable conversion curve.

The formula itself is simple: GMI (%) = 3.31 + 0.02392 × mean glucose (in mg/dL). So if your average sensor glucose over the past two weeks is 154 mg/dL, the math works out to about 7.0%. The concept originally went by “estimated A1C” (eA1C), but the U.S. Food and Drug Administration pushed for a name change because the old term implied the number should match your lab result, which it frequently doesn’t. The rebranding to GMI was meant to signal that this is a glucose-derived indicator, not a blood-test prediction.¹

What GMI Is Actually Showing You

When you see GMI on your Libre 3 report, you’re looking at a snapshot of glucose exposure based on sensor data alone. A lab HbA1c, on the other hand, measures something physical: the percentage of your hemoglobin proteins that have glucose molecules stuck to them. That glycation process happens over the roughly two- to three-month lifespan of a red blood cell, so A1C reflects a weighted average of blood glucose over that window, with more recent weeks counting more heavily.

GMI skips the biology entirely. It doesn’t know anything about your red blood cells, your hemoglobin, or how quickly sugar attaches to proteins. It just knows what your sensor glucose readings were, and it plugs the average into a formula. This is both its strength and its limitation. The strength is immediacy: you can see GMI update on your phone without visiting a lab. The limitation is that anything affecting the biological link between glucose and hemoglobin glycation will cause the two numbers to diverge.

How Much Data You Need for a Reliable GMI

Consensus guidelines recommend at least 14 consecutive days of CGM data with at least 70% completeness to represent a 90-day window of glucose exposure.² That 70% threshold means you need to be wearing the sensor and getting valid readings for roughly 10 out of every 14 days. Gaps from sensor changes, compression lows during sleep, or days you forgot to scan can eat into that completeness.

In practice, most Libre 3 users comfortably exceed this threshold because the sensor streams data automatically to a paired phone without scanning. But if you’ve recently started CGM or had a stretch of sensor errors, your GMI may be calculated from too few readings to be trustworthy. Most reporting apps will flag this or simply not display GMI until enough data has accumulated. A study of young children with type 1 diabetes found that the standard GMI formula performed well with 21 to 27 days of CGM data, though a population-specific equation was needed when less data was available.³

Why GMI and Lab A1C Often Disagree

If you’ve ever compared your GMI to a fresh lab A1C and found them frustratingly different, you’re not alone. One clinic-based study of patients using CGM found that the measured HbA1c was, on average, about 0.34% higher than the CGM-derived GMI.⁴ That might sound small, but a gap like that can land you on different sides of a treatment threshold and change clinical decisions.

The mismatch has multiple sources, and they generally fall into two buckets: things that affect your red blood cells and things that affect your sensor readings.

Red Blood Cell Factors

Lab A1C depends on how long your red blood cells live and how readily your hemoglobin binds glucose. Anything that shortens red blood cell lifespan, like iron-deficiency anemia, hemolytic conditions, or recent significant blood loss, will artificially lower your A1C because cells don’t survive long enough to accumulate as much sugar. Conversely, conditions that extend red blood cell life, like certain hemoglobin variants, can push A1C higher relative to your actual glucose levels.

Kidney disease has a particularly notable effect. In people with chronic kidney disease (CKD), the discordance between A1C and GMI is larger: one study found an average gap of about 0.78% in the CKD group, compared to roughly 0.59% in people without CKD. About two-thirds of individuals with CKD had a difference exceeding 0.5%, versus roughly four in ten of those without CKD.⁵ That study excluded patients with conditions like liver cirrhosis, polycythemia, and metabolic dysfunction-associated steatohepatitis that are also known to skew A1C results.⁴

Sensor-Side Factors

Your Libre 3 measures glucose in interstitial fluid, the thin layer of liquid between cells just under your skin, not directly in your blood. There’s a natural physiological delay as glucose moves from your bloodstream into that interstitial space. Research in healthy fasting adults puts this delay at about five to six minutes.⁶ During stable glucose periods, this barely matters. But when glucose is rapidly rising or falling (after a meal or during exercise), the sensor lags behind what your blood glucose is actually doing. Over thousands of readings, these small timing mismatches can nudge the average glucose, and therefore GMI, in one direction or another.

Sensor accuracy also varies by device model. Research comparing different CGM systems found that the standard GMI formula underestimated glycemia at higher A1C levels, and the point where underestimation became clinically significant differed by sensor model.⁷ The researchers raised the question of whether device-specific GMI equations might be needed rather than the one-size-fits-all formula currently used across all platforms.

The Glycation Gap and Why It Matters

Researchers have long studied what’s called the “glycation gap,” which is the difference between someone’s measured A1C and what their A1C should be based on other markers of glucose exposure. This concept predates CGM: earlier work used fructosamine (a blood test reflecting shorter-term glucose levels) to predict what someone’s A1C ought to be, then measured the gap. That gap turned out to be remarkably stable within individuals over time, suggesting it reflects a genuine biological trait rather than random noise.⁸

The glycation gap has been linked to the risk of diabetes complications. Japanese researchers studying the gap in people with type 2 diabetes have examined its association with hypoglycemia, complications, and quality of life.⁹ The implication is interesting: two people with identical average glucose levels may face different complication risks because one of them glycates hemoglobin more aggressively than the other. For GMI users, this means that even a perfectly accurate GMI reading won’t necessarily predict your personal complication risk the way a lab A1C might.

This is an area where the biology of glycation itself gets relevant. Glucose molecules attach to proteins through a series of chemical reactions that eventually produce permanent modifications. That process isn’t uniform across all people. Genetic variation in hemoglobin structure, differences in red blood cell turnover, and other individual quirks mean that two patients can walk around with the same blood glucose profile and end up with meaningfully different A1C values.¹⁰

Racial and Ethnic Disparities in the Gap

One of the more consequential findings in this area involves race. For a given average glucose level, Black individuals tend to have higher A1C values compared to white individuals. This isn’t a CGM problem; it has been documented with standard glucose testing as well, and it appears to reflect biological differences in hemoglobin glycation rates. The clinical concern is that the standard GMI formula, built on population-average data, can underestimate what a Black patient’s lab A1C will actually read.

Emerging research on “personalized A1C” (pA1C) attempts to address this. One study found that the correlation between GMI and A1C improved substantially when a personalized adjustment was applied: from a moderate correlation to a much tighter one. The improvement was largest in the Black population.¹¹ This work is still relatively early, and personalized A1C is not yet standard practice, but it points toward a future where CGM data and individual biology are combined more effectively than the current one-formula-for-all approach allows.

What to Do When Your GMI and A1C Don’t Match

Seeing a gap between GMI and your lab A1C doesn’t mean one number is “wrong” and the other is “right.” They’re measuring different things. But here’s how to use the discrepancy productively:

A1C higher than GMI: This is the more common direction. It could mean your hemoglobin glycates more readily than average, or it could reflect sensor readings that skew slightly low. Talk to your care team about whether your treatment targets should be based on A1C, GMI, or both.
A1C lower than GMI: Less common, but it happens. Conditions that shorten red blood cell lifespan (anemia, recent blood transfusion, heavy menstrual bleeding) can drive A1C down without actually improving glucose levels. In these cases, GMI may be the more trustworthy number.
Small gap (under 0.3%): Probably not clinically meaningful. Normal biological variability accounts for some discordance, and the lab assay itself has a margin of error.
Large gap (over 0.5%): Worth investigating. Bring it up with your doctor. It could point to an underlying condition affecting red blood cells or hemoglobin, especially if the gap is consistent across multiple comparisons.

Your endocrinologist or primary care provider should know about the gap and factor it into medication decisions. Some providers have started recording both numbers in patient charts and making treatment adjustments based on whichever measure they judge most reliable for that individual patient’s biology.

GMI Alongside Other CGM Metrics

GMI is just one number on your Libre 3 report, and increasingly, diabetes specialists argue it shouldn’t be the star of the show. Time in Range (TIR), the percentage of time your glucose stays between 70 and 180 mg/dL, has become the preferred CGM metric for day-to-day management because it captures not just your average but also how much you swing above and below target.

A related and tighter metric called Time in Tight Range (TITR), which uses a 70-140 mg/dL window, is gaining attention as well. Research has shown that when glucose variability (measured by the coefficient of variation, or CV) is high, TIR can vary quite a lot even at the same GMI. TITR tends to be more stable across different levels of glucose variability, which makes it a potentially more consistent predictor of outcomes.¹²

The practical takeaway: don’t fixate on GMI to the exclusion of your other CGM data. A GMI of 7.0% with 80% TIR and low glucose variability tells a very different story than a GMI of 7.0% with 55% TIR and wild swings between highs and lows. The first scenario reflects genuinely stable glucose management; the second reflects an average that happens to land near 7.0% because highs and lows are canceling each other out.

Insurance and Clinical Record-Keeping

As CGM use grows, questions about how GMI fits into formal medical documentation are becoming more pressing. Reinsurance industry analysis has noted that when both GMI and HbA1c are available, HbA1c is still the preferred measure for underwriting decisions, primarily because the long-term complication data underpinning risk models was built on A1C, not CGM metrics. However, the same guidance acknowledges that GMI may be a reasonable substitute when a recent lab A1C isn’t available.¹³

This matters if you’re applying for life insurance or disability coverage and your medical records contain CGM reports but no recent lab A1C. Some insurers may request a blood draw to get an A1C even if your CGM data looks stellar. It’s worth keeping a recent A1C on file, particularly if you expect an insurance evaluation. Over time, as outcome studies using CGM data accumulate, this bias toward lab A1C in insurance and clinical guidelines will likely soften, but the field isn’t there yet.

When GMI Is Especially Useful

Despite its limitations, GMI fills real gaps in diabetes management. Between quarterly lab visits, it gives you and your care team a running estimate you can act on. If you make a major change to your diet, medication, or exercise routine, you don’t have to wait three months to see whether it’s working. Within two or three weeks, your GMI will start reflecting the impact.

GMI also shines for people whose lab A1C is unreliable due to medical conditions. If you have a hemoglobin variant, chronic anemia, or you’ve recently received a blood transfusion, your lab A1C may not reflect your actual glucose control. In those situations, GMI, which is entirely glucose-based and doesn’t care what your hemoglobin is doing, becomes the more trustworthy guide. The same applies to pregnancy, where rapid changes in blood volume and red blood cell turnover can make A1C less meaningful.

For people using insulin, GMI combined with TIR can be particularly powerful for dosing discussions. A rising GMI alongside declining TIR might prompt a basal rate adjustment, while a stable GMI with increasing time below range could signal overtreatment. These conversations happen faster and with better data when CGM metrics are part of the picture, rather than relying on a single lab number every few months.

Device-Specific Accuracy Considerations

The GMI formula currently used across most CGM platforms, including the Libre 3, was derived from a pooled dataset of clinical trial participants wearing various sensor models. But not all sensors read glucose identically. Research examining whether device-specific GMI formulas are needed found that the standard equation tended to underestimate A1C in the higher ranges, with the point where the underestimation became clinically important varying by sensor hardware.⁷

This is an active area of investigation. If your A1C consistently runs higher than your GMI by a meaningful margin, part of the explanation may simply be that the generic formula doesn’t perfectly match the specific way your sensor reads glucose. Future software updates may incorporate device-specific or even user-specific corrections, but for now, the universal formula is what you’ll see on your Libre 3 reports. Knowing this limitation exists helps you interpret the number with appropriate perspective rather than treating it as gospel.

1
PubMed Central. Glucose Management Indicator (GMI): A New Term for Estimating A1C From Continuous Glucose Monitoring
2
PubMed Central. Impact of Missing Data and Monitoring Duration on Downstream Analyses in Continuous Glucose Monitoring
3
SHARE @ Children’s Mercy. An Examination of the Glucose Management Indicator in Young Children with Type 1 Diabetes
4
PubMed Central. The Discrepancy Between Hemoglobin A1c and Glucose Management Indicators in 26 Patients Treated With Continuous Glucose Monitoring in an Internal Medicine Residency Clinic
5
PubMed Central. Discordance Between Glycated Hemoglobin A1c and the Glucose Management Indicator in People With Diabetes and Chronic Kidney Disease
6
PubMed Central. Time lag of glucose from intravascular to interstitial compartment in humans
7
ScienceDirect. Glucose management indicator: Do we need device-specific equations?
8
PubMed Central. Estimation of the glycation gap in diabetic patients with stable glycemic control
9
PubMed Central. Association of glycation gap with hypoglycemia CGM indices, diabetic complications, and quality of life in Japanese type 2 diabetes patients
10
PubMed Central. Glycation at the Crossroads of Disease Pathogenesis: Mechanistic Insights and Therapeutic Frontiers
11
American Diabetes Association. 167-OR: Personalized A1C Enhances Clinical Decision Making by Improving Agreement between A1C and GMI
12
PubMed. Time in tight range and time in range for predicting the achievement of typical glucose management indicator and HbA1c targets
13
RGA (Reinsurance Group of America). Continuous Glucose Monitoring: Insurance Implications