HbA1c to Estimated Average Glucose Converter

HbA1c to Estimated Average Glucose Converter

Convert HbA1c to an estimated average glucose using the ADAG regression, and see how wide the scatter really is around that estimate.

Last updated 31 Aug 2026

HbA1c → Estimated Average Glucose

ADAG regression
166mg/dLExample

HbA1c 7.4%

Formula

eAG (mg/dL) = 28.7 × HbA1c (%) − 46.7
28.7, −46.7
the slope and intercept from the ADAG study's linear regression of HbA1c against measured average glucose
mmol/L
divide the mg/dL result by 18.016
result
a population-average estimate, not a patient-specific measurement — see the confidence interval below

Worked example

HbA1c 7.4%
28.7 × 7.4 = 212.38
212.38 − 46.7 = 166 mg/dL (165.68, rounded)
166 ÷ 18.016 = 9.2 mmol/L

eAG at common HbA1c values

HbA1c %eAG mg/dLeAG mmol/L
6.01267.0
6.51407.8
7.01548.6
7.51699.4
8.018310.2
9.021211.8
7.0% = 154 mg/dL = 8.6 mmol/L is the value worth memorising; the rest follow the same 28.7 mg/dL step per percentage point.

A translation, not a measurement

Estimated average glucose (eAG) comes from the ADAG (A1c-Derived Average Glucose) study, which measured HbA1c alongside roughly three months of frequent self-monitoring and continuous glucose data in around 500 participants and fitted a straight line between the two. Its purpose is purely communicative: expressing HbA1c as a number in the same units and rough magnitude as a glucose meter reading, which many patients find far more intuitive than a percentage or an mmol/mol value.

The regression describes the population well but any individual can sit well off the line. The published 95% confidence interval around an HbA1c of 7% spans roughly 123 to 185 mg/dL — a range wide enough to span two full HbA1c percentage points. The main driver is that people glycate haemoglobin at genuinely different rates for the same average glucose, so-called high and low glycators, a real biological variation rather than a measurement error in either test.

The practical consequence is that a patient whose own meter or CGM average disagrees with their calculated eAG is not necessarily testing incorrectly, and the two figures should not be reconciled by adjusting one to match the other. Use eAG to explain an HbA1c result in familiar units, not to cross-check a glucose meter.

Frequently asked questions

What is eAG?

Estimated average glucose — a mg/dL (or mmol/L) figure calculated from HbA1c using the ADAG study’s regression, meant to make HbA1c easier to relate to glucose meter readings.

How accurate is eAG for an individual patient?

Only moderately. The 95% confidence interval at an HbA1c of 7% spans roughly 123 to 185 mg/dL, because people glycate haemoglobin at different rates for the same true average glucose.

How do I convert eAG to mmol/L?

Divide the mg/dL value by 18.016. An eAG of 154 mg/dL is 8.6 mmol/L.

Why doesn't my meter average match my eAG?

A mismatch usually reflects normal individual variation in glycation rate rather than an error in either measurement. eAG is a population estimate, not a patient-specific calibration of the meter.

Related calculators

HbA1c NGSP to IFCC ConverterDiabetes, Glucose & Insulin ResistanceGlucose Management Indicator (GMI) CalculatorDiabetes, Glucose & Insulin ResistanceWhole Blood to Plasma Glucose ConverterDiabetes, Glucose & Insulin ResistanceGlucose Unit ConverterDiabetes, Glucose & Insulin ResistanceC-Peptide to Insulin Molar Ratio CalculatorDiabetes, Glucose & Insulin ResistanceBeta-Hydroxybutyrate Unit ConverterDiabetes, Glucose & Insulin ResistanceHOMA-IR CalculatorDiabetes, Glucose & Insulin ResistanceTriglyceride-Glucose (TyG) Index CalculatorDiabetes, Glucose & Insulin Resistance

References

  1. Nathan DM et al. Translating the A1C assay into estimated average glucose values. Diabetes Care. 2008;31(8):1473–8.
  2. American Diabetes Association. Standards of Care in Diabetes — 2026.