Estimated Average Glucose Calculator
Turn an A1C percentage into estimated average glucose in mg/dL or mmol/L with the ADAG formula.
What it does
A1C reports the share of hemoglobin that has glucose attached to it. It is a long run average, roughly the past two to three months, but the percentage is not intuitive for most people. Estimated average glucose translates that percentage into the same units a glucose meter uses, so a person can compare an A1C result with the numbers they see every day.
The conversion comes from the ADAG study, which measured A1C and continuous glucose in 507 adults across ten centers. The study found a straight line relationship: eAG in mg/dL equals 28.7 times A1C minus 46.7. In SI units the same line is reported as eAG in mmol/L equals 1.59 times A1C minus 2.59. Dividing the mg/dL result by 18 gives the same answer, with small rounding differences.
The formula is a population fit, not a promise. The 95 percent interval around the estimate is wide, often tens of mg/dL. Two people with the same A1C can have different average glucose patterns, and a person with frequent highs and lows can have the same A1C as someone with steady readings. eAG is a communication tool, not a replacement for a meter or a continuous monitor.
A1C itself can mislead in some conditions. Anemia, recent blood loss, pregnancy, sickle cell trait, and some kidney or liver diseases change red cell survival and shift the result. Labs use standardized methods, but a result that does not fit the glucose log deserves a conversation before it drives a treatment change.
This page uses the ADAG equation and the ADA A1C categories. It returns the estimated average in the unit system you choose and labels the A1C as normal, prediabetes, or diabetes range. The categories are screening bands for adults, not a diagnosis, and they do not replace a clinician's assessment.
How to use
- Choose US units for mg/dL or metric for mmol/L.
- Enter the A1C percentage from a standardized lab report.
- Read the estimated average glucose and the A1C category.
- Compare the estimate with your meter or sensor average, keeping in mind that the 95 percent interval is wide.
- Look at the trend over several A1C results rather than one value.
- Discuss any mismatch between A1C and daily readings with your care team.
Formula
- A1C
- Glycated hemoglobin percentage from a standardized assay.
- 28.7
- Slope of the ADAG regression line in mg/dL per A1C point.
- 46.7
- Intercept of the ADAG line in mg/dL.
- 18
- Approximate milligrams per deciliter per millimole per liter of glucose.
- eAG
- Estimated average glucose in mg/dL or mmol/L.
The ADAG equation was fitted to 507 subjects, including 268 with type 1 diabetes, 159 with type 2 diabetes, and 80 without diabetes. The published 95 percent interval is wide, so eAG is best read as a range around the estimate. The SI version can be written as 1.59 x A1C - 2.59, which is the same line after unit conversion. A1C values reported in IFCC mmol/mol must be converted to NGSP percent first: NGSP percent = 0.09148 x IFCC + 2.152.
Worked example
Example: A1C 7.0 percent
A standardized A1C comes back at 7.0 percent. Convert it to estimated average glucose in both unit systems.
- Multiply the A1C by 28.7: 28.7 x 7.0 = 200.9.
- Subtract 46.7: 200.9 - 46.7 = 154.2 mg/dL.
- The US display rounds to 154 mg/dL, matching the ADA table.
- Convert to SI: 154.2 / 18 = 8.567 mmol/L.
- The metric display rounds to 8.6 mmol/L, also matching the ADA table.
- An A1C of 7.0 percent sits in the diabetes range under ADA adult categories.
- The estimate is a population average, so a meter average a little above or below 154 mg/dL is still consistent.
Reference table
| A1C | eAG mg/dL | eAG mmol/L | Category |
|---|---|---|---|
| 5.0 % | 97 | 5.4 | Normal range |
| 6.0 % | 126 | 7.0 | Prediabetes range |
| 6.5 % | 140 | 7.8 | Diabetes range |
| 7.0 % | 154 | 8.6 | Diabetes range |
| 8.0 % | 183 | 10.2 | Diabetes range |
| 9.0 % | 212 | 11.8 | Diabetes range |
| 10.0 % | 240 | 13.4 | Diabetes range |
Limits
- eAG is a population average with a wide 95 percent interval, not a personal daily target.
- It does not replace self-monitoring or continuous glucose monitoring.
- Hemoglobin variants, anemia, blood loss, pregnancy, and kidney or liver disease can shift A1C.
- A1C lags behind recent glucose changes, so a new treatment effect may not show for weeks.
Common mistakes
- Treating eAG as a number every meter reading should match.
- Entering IFCC mmol/mol as if it were an NGSP percentage.
- Using the formula for a person with a condition that changes red cell survival without noting it.
- Comparing A1C results from labs that use different methods without checking standardization.
- Ignoring a large gap between eAG and the glucose log.
Try it yourself
Quick checks. Answers sit tucked under each question.
A1C 6.0 percent. What is the estimated average glucose?
28.7 x 6.0 - 46.7 = 125.5 mg/dL, displayed as 126 mg/dL. In SI, 125.5 / 18 = 7.0 mmol/L.
A1C 8.0 percent. What is the estimated average glucose?
28.7 x 8.0 - 46.7 = 182.9 mg/dL, displayed as 183 mg/dL. In SI, 182.9 / 18 = 10.2 mmol/L.
A1C 9.5 percent. What is the estimated average glucose?
28.7 x 9.5 - 46.7 = 225.95 mg/dL, displayed as 226 mg/dL. In SI, 225.95 / 18 = 12.6 mmol/L.
Questions
What is eAG?
Estimated average glucose is an A1C result translated into the same units as a glucose meter, using the ADAG equation.
How accurate is eAG?
The equation fits a population, and the 95 percent interval around an individual estimate is wide. Use it as a communication aid, not a precise personal average.
Why 28.7 and 46.7?
They are the slope and intercept of the ADAG regression line in mg/dL. The SI line uses 1.59 and 2.59.
Can I use eAG instead of checking blood sugar?
No. eAG is a retrospective average and cannot show daily highs, lows, or variability.
What if my A1C is in mmol/mol?
Convert it to NGSP percent first with NGSP = 0.09148 x IFCC + 2.152, then apply the eAG formula.
Does eAG change with age?
The formula does not include age. A1C targets may differ by age and health, but the conversion itself does not.
Sources
- American Diabetes Association: eAG and A1C conversion
- Nathan et al 2008: Translating the A1C assay into estimated average glucose values