HbA1c to eAG conversion — practical worked examples

HbA1c to eAG conversion is the calculation that turns your three-month glycated haemoglobin percentage into the same milligrams-per-decilitre units your glucometer shows. For the full clinical context of what A1C means for daily eating, see A1C and what it actually means. — making it immediately interpretable without mental abstraction. Most people with diabetes know their A1C targets (below 7.0% for most adults per ADA Standards of Care 2024 §6) but cannot intuitively feel what “7.0%” means in terms of their daily glucose experience. The eAG (estimated average glucose) bridges that gap: an A1C of 7.0% = eAG of 154 mg/dL. An A1C of 8.0% = eAG of 183 mg/dL. An A1C of 6.5% = eAG of 140 mg/dL. These are the daily average glucose levels that your A1C is mathematically derived from, based on the ADAG (A1C-Derived Average Glucose) study that correlated A1C values with continuous glucose monitoring data in 507 people with and without diabetes. The conversion formula is: eAG (mg/dL) = (28.7 × A1C%) − 46.7. This guide provides six worked conversion examples across clinically relevant A1C values, explains the limitations of the formula, and shows how to use eAG alongside time-in-range data for a complete picture of glucose exposure.

The ADAG Study — How the Conversion Formula Was Derived

The A1C-Derived Average Glucose study was published by Nathan et al. in 2008 in Diabetes Care and remains the definitive source for the eAG conversion. The study enrolled 507 participants across 10 international sites: 268 with Type 1 diabetes, 159 with Type 2 diabetes, and 80 without diabetes. Each participant wore a continuous glucose monitor for 3 months while laboratory A1C was measured simultaneously. The resulting dataset linked laboratory-measured A1C to CGM-measured mean glucose across the full glycaemic spectrum.¹

The regression equation derived from that dataset is: eAG (mg/dL) = 28.7 × A1C − 46.7. The equivalent formula in mmol/L is: eAG (mmol/L) = 1.59 × A1C − 2.59. The R² of the regression is 0.84, meaning A1C explains 84% of the variance in mean glucose — strong enough for clinical utility but imperfect enough to require the caveats described below.

The 16% of variance not explained by A1C reflects genuine biological heterogeneity. Red blood cells in different people have different average lifespans (70–120 days, not a fixed 90 days as often assumed). A person whose red cells turn over faster than average will have an A1C that underestimates their mean glucose, because haemoglobin has had less time for glycation. A person with slower red cell turnover will have an A1C that overestimates mean glucose. This is not error in the measurement — it is real biological variation that A1C alone cannot capture.¹

The ADA adopted eAG as a standard reporting format in 2010, recommending that laboratories report both A1C and eAG simultaneously so patients can interpret their result in glucometer-familiar units. Despite this, clinical adoption has been inconsistent — many labs still report A1C percentage only, leaving patients to do the conversion themselves or not at all.

Six Worked Conversion Examples — from A1C 5.0% to 9.0%

The following calculations apply the ADAG formula directly. These are the same values published in Nathan et al. 2008 and endorsed by the ADA and EASD.

A1C 5.0% → eAG 97 mg/dL Calculation: (28.7 × 5.0) − 46.7 = 143.5 − 46.7 = 96.8, rounded to 97 mg/dL. This falls within the normal (non-diabetic) range. For context, 97 mg/dL is a typical fasting glucose value for a person without insulin resistance. An A1C of 5.0% indicates very low cumulative glucose exposure and negligible glycation of other proteins.

A1C 6.0% → eAG 126 mg/dL Calculation: (28.7 × 6.0) − 46.7 = 172.2 − 46.7 = 125.5, rounded to 126 mg/dL. An eAG of 126 mg/dL is the threshold value for diabetes diagnosis by fasting plasma glucose. An A1C of 6.0% is in the prediabetes range (5.7–6.4% per ADA criteria), but the eAG it implies matches the glucose threshold for diabetes by a different metric. This apparent contradiction reflects the sampling differences between A1C (3-month average, all hours) and a single fasting measurement.

A1C 7.0% → eAG 154 mg/dL Calculation: (28.7 × 7.0) − 46.7 = 200.9 − 46.7 = 154.2, rounded to 154 mg/dL. This is the standard glycaemic target for most adults with Type 1 or Type 2 diabetes per ADA Standards of Care 2024. An eAG of 154 mg/dL means that across all hours of the day — pre-meal, post-meal, overnight — the average glucose reading was 154. Pre-meal readings in a well-managed patient are typically 80–130 mg/dL, meaning post-meal spikes into the 180–220 mg/dL range are pulling the average up to 154.²

A1C 8.0% → eAG 183 mg/dL Calculation: (28.7 × 8.0) − 46.7 = 229.6 − 46.7 = 182.9, rounded to 183 mg/dL. An eAG of 183 mg/dL corresponds to the ADA’s 2-hour post-meal glucose threshold for diagnosing gestational diabetes. At an average of 183 mg/dL across all hours, post-meal peaks are likely reaching 220–260 mg/dL regularly. This is the glycaemic level at which early microvascular complications — retinopathy, nephropathy, neuropathy — become clinically measurable over multi-year exposure.²

A1C 8.5% → eAG 197 mg/dL Calculation: (28.7 × 8.5) − 46.7 = 243.95 − 46.7 = 197.25, rounded to 197 mg/dL. This is a commonly encountered clinical value in patients who have not achieved target control. An eAG approaching 200 mg/dL means post-meal excursions are regularly exceeding 240–260 mg/dL. At this level, osmotic symptoms (increased thirst, frequent urination) may be intermittently present.

A1C 9.0% → eAG 212 mg/dL Calculation: (28.7 × 9.0) − 46.7 = 258.3 − 46.7 = 211.6, rounded to 212 mg/dL. An eAG of 212 mg/dL represents significantly elevated chronic glucose exposure. The DCCT trial established that A1C values persistently above 9% carry substantially higher risk of microvascular complications compared to values at 7%. Each 1% reduction in A1C from this level reduces retinopathy risk by approximately 35% and nephropathy risk by 24%.³

Why eAG and Your Glucometer Average Often Differ

Your glucometer calculates a simple average of all readings taken — most of which are pre-meal, when glucose is relatively controlled. The eAG from A1C includes all glucose values including the post-meal spikes and overnight lows that are rarely captured by finger-stick monitoring. The CGM-based average is almost always higher than the glucometer average, and eAG almost always exceeds the glucometer’s stored average by 15–30 mg/dL in typical finger-stick monitoring patterns.¹

This is not a calibration error in your glucometer. It is a systematic sampling bias inherent in finger-stick testing patterns. Most people test before meals (fasting or pre-prandial readings) and before bed — times when glucose is relatively stable and often near its daily minimum. Post-meal testing, which would capture glucose peaks of 40–80 mg/dL above the pre-meal value, is less consistently done. The glucometer’s stored average therefore underrepresents the glucose peaks that A1C fully integrates.

If your glucometer’s 90-day average is reading 125 mg/dL but your A1C implies an eAG of 154 mg/dL, the gap is almost entirely explained by the post-meal hours you are not testing. This discordance is diagnostic: it tells you that your post-meal control is significantly worse than your pre-meal readings suggest. The corrective action is either to add post-meal testing at 1- and 2-hour intervals, or to use a CGM that captures the full glucose curve.

eAG Limitations — the Conditions That Invalidate the Conversion

The ADAG formula requires that A1C accurately reflects mean glucose exposure over 90 days. Several conditions alter this relationship sufficiently to make eAG unreliable as a clinical tool.

Haemoglobin variants — sickle cell trait (HbS), haemoglobin C (HbC), haemoglobin E (HbE) — alter the glycation kinetics of haemoglobin and may cause A1C to be falsely low or falsely high depending on the specific variant and the assay method used. Sickle cell trait is common: approximately 8% of African Americans carry it. In these patients, A1C-derived eAG can significantly misrepresent true mean glucose.⁴

Haemolytic anaemia shortens the lifespan of red blood cells, reducing the time available for haemoglobin glycation. This produces a falsely low A1C relative to actual mean glucose — meaning the eAG calculated from A1C underestimates true average glucose exposure. Conversely, iron deficiency anaemia and vitamin B12 deficiency, which increase red cell lifespan, produce falsely elevated A1C and eAG overestimation of mean glucose.

Recent blood transfusions replace the patient’s glycated red cells with donor cells carrying the donor’s glycation pattern, producing an A1C that reflects a mixture of the patient’s and donor’s glucose exposure history. In patients who received a transfusion within 90 days, A1C is unreliable for any purpose.

In all of these populations, fructosamine (reflecting 2–3 weeks of average glucose) or CGM-derived time-in-range (TIR) and Glucose Management Indicator (GMI) are more reliable than A1C-based eAG.

Glucose Management Indicator — The CGM Alternative to A1C

The Glucose Management Indicator (GMI) is the CGM-era equivalent of A1C. Where A1C measures actual haemoglobin glycation in a blood sample, GMI is calculated directly from the CGM’s mean glucose reading using the ADAG formula in reverse: GMI (%) = 3.31 + (0.02392 × mean glucose in mg/dL). GMI represents what A1C would be if it perfectly tracked CGM mean glucose — a standardised estimate of A1C from continuous data rather than from a blood draw.⁴

For most patients, GMI and laboratory A1C correlate within 0.5%. A discordance greater than 0.5% between GMI and A1C is the signal that indicates one of the conditions described above: a haemoglobin variant, altered red cell lifespan, recent transfusion, or another factor that decouples A1C from true mean glucose. In clinical practice, discordance greater than 0.5% should prompt a conversation with your endocrinologist about alternative glycaemic monitoring metrics.

GMI has a further practical advantage: it updates continuously as CGM data accumulates, providing a near-real-time equivalent of A1C rather than a 90-day lagged value. A patient who makes a dietary change or medication adjustment can see the GMI impact within 2 weeks rather than waiting for the next laboratory appointment.

Putting eAG Into Daily Practice — the Glucometer Average Target

Knowing your eAG target gives you a glucometer average goal that directly predicts your A1C at the next clinic visit. If your A1C target is 7.0% (eAG 154 mg/dL) and you are testing primarily pre-meal, your pre-meal average target should be approximately 120–130 mg/dL — lower than 154 — to leave room for the post-meal peaks that A1C captures but your glucometer testing misses.²

Working backwards from common A1C targets:

• A1C target 6.5% (eAG 140 mg/dL): pre-meal glucometer target approximately 110–120 mg/dL
• A1C target 7.0% (eAG 154 mg/dL): pre-meal glucometer target approximately 120–130 mg/dL
• A1C target 8.0% (eAG 183 mg/dL): pre-meal glucometer target approximately 140–155 mg/dL (common target for elderly patients or those with hypoglycaemia risk)

These figures assume a typical post-meal excursion of 30–40 mg/dL above the pre-meal baseline. If your post-meal excursions are consistently larger (due to high-GI foods, missed mealtime insulin, or delayed gastric emptying), the required pre-meal average is lower. If your diet and insulin timing produce smaller excursions, the pre-meal target can be higher.

CalEye’s glucose log integration allows you to track pre-meal and post-meal readings alongside food entries. The pattern that emerges — which meals produce large excursions, which produce smaller ones — is the dietary self-management data that makes eAG targets actionable rather than abstract. An A1C of 7.5% is a three-month summary. The individual meal logs that produced it are the intervention points.

References

1. Nathan DM, Kuenen J, Borg R, et al. “Translating the A1C Assay into Estimated Average Glucose Values.” Diabetes Care 31, no. 8 (2008): 1473–1478. (The ADAG study — source of the eAG formula and all reference conversion values.)

2. American Diabetes Association Professional Practice Committee. “Glycemic Goals and Hypoglycemia: Standards of Care in Diabetes—2024.” Diabetes Care 47, Supplement 1 (2024): S111–S125. Section 6.

3. The Diabetes Control and Complications Trial Research Group. “The Effect of Intensive Treatment of Diabetes on the Development and Progression of Long-Term Complications in Insulin-Dependent Diabetes Mellitus.” New England Journal of Medicine 329, no. 14 (1993): 977–986.

4. Bergenstal RM, Beck RW, Close KL, et al. “Glucose Management Indicator (GMI): A New Term for Estimating A1C from Continuous Glucose Monitoring.” Diabetes Care 41, no. 11 (2018): 2275–2280.