InstalabInstalabLoading...
Instalab ResearchThe formula behind eAG was developed from a landmark study of 507 people who wore continuous glucose monitors and also did frequent fingerstick checks. Researchers measured each person's actual average glucose over two to three months and compared it to their HbA1c results. From that data, they built a simple equation: eAG (mg/dL) = 28.7 × A1c (%) - 46.7.
The relationship was strong overall, and it held up across people with type 1 diabetes, type 2 diabetes, and those without diabetes. It was also consistent across age, sex, race/ethnicity, and smoking status in that study population. Major diabetes organizations then endorsed reporting eAG alongside HbA1c so patients could link that "percentage" to the glucose numbers they actually see at home.
Here's what the basic conversion looks like:
| HbA1c (%) | Approximate eAG (mg/dL) | Approximate eAG (mmol/L) |
|---|---|---|
| 6.0 | ~126 | ~7.0 |
| 7.0 | ~154 | ~8.6 |
| 8.0 | ~183 | ~10.2 |
This is where things get more nuanced. In the original study, about 90% of people had actual averages within 15% of their eAG value. That sounds solid until you think about what it means in practice. If your eAG reads 154 mg/dL, your true average could be anywhere from about 131 to 177 mg/dL. And for the remaining 10% of people, the gap was even wider.
Several research teams have raised concerns that this error range is too large for precise individual treatment decisions. eAG assumes everyone's body glycates hemoglobin (attaches sugar to red blood cells) at the same rate. But that rate varies from person to person based on biology that has nothing to do with blood sugar levels.
Research into personalized mathematical modeling of hemoglobin glycation and red blood cell lifespan has shown that accounting for these individual differences can cut errors in average glucose estimates by more than 50%. That's promising, but these personalized approaches aren't yet part of routine clinical care.
Several factors can make eAG less reliable for specific individuals:
If you wear a continuous glucose monitor, you may also see something called the Glucose Management Indicator (GMI). This is essentially the CGM's version of eAG, working in the opposite direction: it takes your CGM average glucose over 10 to 14 days and converts it into an estimated A1c-like number.
Here's the catch. Research consistently finds that GMI and lab-measured A1c often disagree. This mismatch is common and, according to recent analysis, largely driven by stable non-glycemic factors (like individual differences in how your body glycates hemoglobin) rather than errors in the CGM itself.
The emerging expert consensus is that you're better off looking at your raw CGM mean glucose and your time-in-range rather than treating GMI as a reliable stand-in for your lab A1c. One prominent commentary in the diabetes research literature went so far as to suggest that GMI should be replaced with simple mean CGM glucose to reduce confusion.
There is one genuinely useful trick: comparing your eAG (derived from your lab A1c) with the actual average glucose from your meter or CGM. If there's a consistent, meaningful gap between these two numbers, that tells you something important.
If your eAG is significantly higher than your meter/CGM average, your body may glycate hemoglobin at a higher-than-average rate. Your A1c looks worse than your day-to-day glucose control actually is. If your eAG is consistently lower than your measured average, the opposite may be true, and your A1c might be giving you a falsely reassuring picture.
Research has shown that spotting these mismatches can flag situations where relying on A1c alone might mislead treatment decisions. The ratio between eAG and fasting glucose has also been explored as a potential indicator of insulin resistance in younger adults with diabetes, though this is still in early research stages.
The bottom line from the research is clear: eAG is a helpful translator, not a precision tool. Here's how to put that into practice.
Use eAG the way it was designed: as a quick way to understand what your A1c percentage means in the glucose units you see every day. If your doctor says your A1c is 7%, it's useful to know that roughly corresponds to an average glucose of about 154 mg/dL.
But don't treat eAG as gospel for your personal situation. If you track your glucose at home with a meter or CGM, compare that average to your eAG. A persistent gap in either direction is worth discussing with your doctor, because it may mean your A1c is painting an inaccurate picture of your real control.
Be especially cautious about relying on eAG if you are pregnant, have kidney disease or anemia, or know you carry a hemoglobin variant. In these situations, the standard conversion is more likely to be misleading.
And if you use a CGM, pay more attention to your actual mean glucose and your time-in-range (the percentage of the day your glucose stays between 70 and 180 mg/dL) than to eAG or GMI. These direct measurements tell you what's actually happening, without the assumptions baked into the conversion formulas.