Convert HbA1c to estimated average glucose (eAG) and back
| HbA1c | eAG (mg/dL) | eAG (mmol/L) | Category |
|---|---|---|---|
| 5% | 97 | 5.4 | Normal |
| 5.7% | 117 | 6.5 | Prediabetes begins |
| 6% | 126 | 7.0 | Prediabetes |
| 6.5% | 140 | 7.8 | Diabetes threshold |
| 7% | 154 | 8.6 | Common ADA target |
| 8% | 183 | 10.2 | Above target |
| 9% | 212 | 11.8 | Well above target |
| 10% | 240 | 13.4 | High |
| 11% | 269 | 14.9 | Very high |
| 12% | 298 | 16.5 | Very high |
Values follow the American Diabetes Association's eAG table, derived from the ADAG study formula. Because the relationship is a straight line, each additional 1% of A1c adds about 29 mg/dL to your average.
| Test | Normal | Prediabetes | Diabetes |
|---|---|---|---|
| HbA1c | Below 5.7% | 5.7% – 6.4% | 6.5% or higher |
| Fasting plasma glucose | Below 100 mg/dL | 100 – 125 mg/dL | 126 mg/dL or higher |
| 2-hr OGTT (75 g) | Below 140 mg/dL | 140 – 199 mg/dL | 200 mg/dL or higher |
| Random glucose + symptoms | — | — | 200 mg/dL or higher |
Thresholds per American Diabetes Association Standards of Care. A diabetes diagnosis needs a confirmatory second test unless the result is unambiguous. mmol/L equivalents: 100 mg/dL = 5.6, 126 = 7.0, 140 = 7.8, 200 = 11.1.
HbA1c measures the percentage of your hemoglobin that's coated in sugar. Since red blood cells live about three months, the number acts like a memory of your glucose over that span, and doctors increasingly quote it as an estimated average glucose (eAG) in the same units a meter uses.
eAG (mg/dL) = 28.7 × A1c − 46.7. Flip it around and A1c (%) = (eAG + 46.7) ÷ 28.7. To switch mg/dL into the international mmol/L, divide by 18. The equation comes from the A1c-Derived Average Glucose study, which logged thousands of continuous glucose monitor readings from about 500 people across the US and Europe and regressed them against lab A1c (Diabetes Care, 2008).
Pick a direction, type your number, and read the result in both unit systems. The category pill shows where you land on the ADA bands: normal, prediabetes, or diabetes. If you're converting a meter average to A1c, remember the caveat below about testing patterns before treating the two as identical.
Say your lab returns an A1c of 7.5%. eAG = 28.7 × 7.5 − 46.7 = 168.6, so about 169 mg/dL, which is 169 ÷ 18 = 9.4 mmol/L. Going the other way, a three-month CGM average of 140 mg/dL works out to (140 + 46.7) ÷ 28.7 = 6.5%, right on the diabetes diagnostic line. That's no coincidence: the ADA picked 6.5% partly because it maps to the fasting threshold's long-run average.
One caution from the same study: about 10% of people fall well off the regression line. Sickle cell trait, hemolytic anemia, recent transfusion, pregnancy, and kidney disease all change how long red cells survive or how they get glycate, which shifts A1c without changing actual glucose. If your A1c and meter numbers disagree wildly, that gap itself is worth bringing to your provider.
An A1c of 7% corresponds to an estimated average glucose of 154 mg/dL (8.6 mmol/L) using the ADAG formula: 28.7 × 7 − 46.7. Each 1% change in A1c moves the average by about 29 mg/dL.
eAG (mg/dL) = 28.7 × A1c − 46.7. To go the other way, A1c (%) = (eAG + 46.7) ÷ 28.7. For mmol/L, divide the mg/dL value by 18. The formula comes from the A1c-Derived Average Glucose study of roughly 500 people published in Diabetes Care in 2008.
An A1c of 6.5% or higher meets the ADA criterion for diabetes when confirmed by a second test. Prediabetes is 5.7% to 6.4%, and anything below 5.7% falls in the normal range. A1c of 6.5% works out to an eAG of 140 mg/dL.
Meter averages only include the moments you actually tested, so they skew toward fasting mornings if that's when you check. eAG is a weighted 3-month average that includes overnight and post-meal hours you never test. If you check mostly before breakfast, expect your meter average to run lower than eAG.
Roughly one full point. Since each 1% of A1c equals about 29 mg/dL of average glucose, a 30 mg/dL drop moves an A1c of 8% (183 mg/dL) down to about 7% (154 mg/dL) over the red blood cell lifespan of around 3 months.
No. eAG is a 24-hour weighted average over about 3 months, while fasting glucose is a single morning reading. Diagnostic cutoffs differ too: diabetes is eAG-equivalent 140 mg/dL at an A1c of 6.5%, but fasting glucose only needs to hit 126 mg/dL to cross the same line.