To convert mg/dL to A1C, use the ADAG formula from the American Diabetes Association: A1C (%) = (mean glucose in mg/dL + 46.7) / 28.7. Going the other way: eAG (mg/dL) = 28.7 x A1C – 46.7. For example, an average glucose of 120 mg/dL equals an A1C of about 5.8%. The formula works best when applied to 10+ days of CGM or multiple daily fingerstick data — not to a single reading.
Where the Formula Comes From
The A1C-Derived Average Glucose (ADAG) study, published by Nathan and colleagues in Diabetes Care (2008), enrolled 507 participants (non-diabetic, type 1, type 2). Each wore a CGM continuously plus checked 7-point fingerstick profiles for 12 weeks and had lab A1C measured. The regression produced:
eAG (mg/dL) = 28.7 x A1C – 46.7
Solving for A1C gives:
A1C (%) = (mean glucose + 46.7) / 28.7
The correlation between calculated eAG and true mean glucose was strong (r = 0.92). The ADA endorsed this formula in 2008, and it’s the basis for the eAG reported on many lab reports and CGM downloads.
Conversion Chart
| A1C (%) | eAG (mg/dL) | eAG (mmol/L) | Category |
|---|---|---|---|
| 4.5 | 83 | 4.6 | Low-normal |
| 5.0 | 97 | 5.4 | Normal |
| 5.4 | 108 | 6.0 | Normal |
| 5.7 | 117 | 6.5 | Prediabetes |
| 6.0 | 126 | 7.0 | Prediabetes |
| 6.4 | 137 | 7.6 | Prediabetes (upper) |
| 6.5 | 140 | 7.8 | Diabetes threshold |
| 7.0 | 154 | 8.6 | Typical target |
| 7.5 | 169 | 9.4 | Elevated |
| 8.0 | 183 | 10.2 | Elevated |
| 8.5 | 197 | 10.9 | Poor control |
| 9.0 | 212 | 11.8 | Poor control |
| 10.0 | 240 | 13.3 | Very poor |
| 12.0 | 298 | 16.5 | Severely elevated |
Worked Examples
Example 1: CGM 14-day average = 110 mg/dL
A1C = (110 + 46.7) / 28.7 = 156.7 / 28.7 = 5.5%. That’s normal-range territory. Keep current habits.
Example 2: Fingerstick average (4 readings/day x 14 days) = 165 mg/dL
A1C = (165 + 46.7) / 28.7 = 211.7 / 28.7 = 7.4%. Consistent with moderately controlled diabetes. Consider reviewing meals and medications with your doctor.
Example 3: Reverse calculation — A1C of 6.3%
eAG = 28.7 x 6.3 – 46.7 = 180.81 – 46.7 = 134 mg/dL. Firm prediabetes range.
How to Collect the Right Data
The conversion assumes your “mean glucose” represents a real average over time. A single morning reading won’t tell you your A1C. Good inputs:
- CGM data (14 days or more): best input. Most CGMs compute GMI, which uses the slightly different Bergenstal formula (3.31 + 0.02392 x mean glucose) but produces similar numbers.
- Structured fingerstick profiling: 7 readings daily (pre-meals, 2 hours post-meals, bedtime) for 1-2 weeks gives a solid average.
- Less reliable: fasting-only readings. They miss post-meal excursions that often dominate A1C.
Why CGM GMI and ADAG Differ Slightly
The ADA eAG formula (ADAG 2008) and the newer GMI formula (Bergenstal 2018) use different regressions. For the same mean glucose:
| Mean glucose (mg/dL) | ADAG A1C (%) | GMI (%) |
|---|---|---|
| 100 | 5.1 | 5.7 |
| 125 | 6.0 | 6.3 |
| 150 | 6.9 | 6.9 |
| 175 | 7.7 | 7.5 |
| 200 | 8.6 | 8.1 |
At lower glucose averages, GMI tends to be higher than the ADAG-derived A1C; at higher averages, they converge or GMI runs slightly lower. Both are valid estimates. Lab A1C is still the reference for diagnosis.
Common Pitfalls
- Using a single reading: a morning fasting of 140 does not equal A1C 6.5%. You need an average over days to weeks.
- Skipping post-meal readings: if you only check fasting, you’ll underestimate average glucose and A1C.
- Anemia, hemoglobinopathy, pregnancy: these widen the gap between calculated and lab A1C. See our piece on A1C levels.
- Treating the estimate as a diagnosis: ADA requires lab A1C (or fasting plasma glucose, OGTT) for formal diagnosis.
Putting the Conversion to Work
Use the calculator as a real-time progress check. If your CGM shows a 14-day average of 135 mg/dL, your eAG-derived A1C is about 6.3%. Three weeks later, if it drops to 120 mg/dL (A1C ~5.8%), you’ve made meaningful progress. Track the trend monthly and get a lab A1C every 3-6 months. See our treatment hub for what to do with the trend.
The Bottom Line
The mg/dL to A1C conversion formula — A1C = (mean glucose + 46.7) / 28.7 — is a useful estimator of how your average glucose maps to hemoglobin A1C. Use it with CGM data or structured fingerstick averages over 10+ days for best results. Confirm with lab A1C before diagnosing or making major treatment decisions. The formula is a window into your glucose control, not a replacement for your clinician’s assessment.
