A1c to Estimated Average Glucose (eAG)

About A1c to Estimated Average Glucose (eAG)

Hemoglobin A1c (HbA1c) reflects the average blood glucose over the preceding 2-3 months, weighted toward the most recent 30 days. For decades, clinicians and patients struggled to translate A1c values into a more intuitive metric, until the A1c-Derived Average Glucose (ADAG) study, published by Nathan and colleagues in Diabetes Care in 2008, established a linear relationship between A1c and estimated average glucose (eAG). The ADAG study combined continuous glucose monitoring and frequent fingerstick glucose measurements in 507 adults with type 1 and type 2 diabetes (and 80 non-diabetic controls) across 10 international centers, deriving the equation eAG (mg/dL) = 28.7 × A1c − 46.7.

What the calculator measures

The calculator returns two values: the eAG in mg/dL and the eAG in mmol/L (the international unit). The eAG can be displayed alongside the A1c on lab reports and glucose meters, helping patients understand that an A1c of 7% corresponds to an average glucose of 154 mg/dL (8.6 mmol/L), a level that is meaningfully above the normal range.

How the formula works

The ADAG formula is a simple linear regression: eAG (mg/dL) = 28.7 × A1c − 46.7. For example:

• A1c 5.0% → eAG 97 mg/dL (5.4 mmol/L)
• A1c 6.0% → eAG 126 mg/dL (7.0 mmol/L)
• A1c 7.0% → eAG 154 mg/dL (8.6 mmol/L)
• A1c 8.0% → eAG 183 mg/dL (10.2 mmol/L)
• A1c 9.0% → eAG 212 mg/dL (11.8 mmol/L)
• A1c 10.0% → eAG 240 mg/dL (13.4 mmol/L)

The relationship is log-linear at the lower end (corresponding to the physiology of glycation) but is well-approximated by the linear formula across the clinically relevant range (A1c 4-12%). The 95% prediction interval for an individual patient is approximately ±30 mg/dL around the eAG value, meaning that an A1c of 7% could correspond to an average glucose anywhere from 124 to 184 mg/dL in an individual patient.

Interpretation

The American Diabetes Association recommends the following A1c targets:

• General adult target: <7.0% (eAG <154 mg/dL)
• Tight control (selected patients): <6.5% (eAG <140 mg/dL)
• Looser control (older/high-risk patients): <8.0% (eAG <183 mg/dL)
• Diagnosis of prediabetes: 5.7-6.4% (eAG 117-137 mg/dL)
• Diagnosis of diabetes: ≥6.5% (eAG ≥137 mg/dL)

The eAG helps patients understand that an A1c of 9% means their average glucose is 212 mg/dL, well above the postprandial target of <180 mg/dL, and motivates behavior change. The eAG is also useful for interpreting continuous glucose monitor (CGM) data, which provides a direct measure of average glucose that can be compared to the eAG implied by the A1c.

Clinical use

The eAG is reported alongside the A1c on most modern laboratory reports and on most glucose meters, allowing clinicians and patients to interpret the A1c in the same units as fingerstick readings. The eAG is particularly useful for patients who use glucose meters or CGMs, because they can compare their measured average glucose to the eAG implied by their A1c. A significant discrepancy between measured average glucose and eAG should prompt investigation of conditions that affect A1c accuracy (see Limitations below).

Limitations

The A1c-to-eAG conversion has several limitations. First, the A1c is inaccurate in patients with altered red blood cell turnover, anemia (iron deficiency, hemolytic anemia, B12/folate deficiency), recent transfusion, erythropoietin therapy, hemoglobinopathies (sickle cell trait, thalassemia), or pregnancy. In these patients, alternative measures of glycemic control (fructosamine, glycated albumin, CGM) should be used. Second, the A1c may be falsely low in patients with shortened red blood cell lifespan (hemolysis, recent transfusion, erythropoietin therapy) and falsely high in patients with lengthened red blood cell lifespan (iron deficiency, B12/folate deficiency, splenectomy). Third, the A1c reflects the average glucose over 2-3 months but does not capture glucose variability or hypoglycemia, a patient with frequent hypoglycemia and hyperglycemia can have the same A1c as a patient with stable glucose. CGM metrics (time in Range, GMI, coefficient of variation) capture this information that A1c misses. Fourth, the ADAG formula was derived in a predominantly white population and may slightly underestimate eAG in African American patients (the A1c tends to be ~0.4 percentage points higher in African Americans at the same average glucose).

Comparison with other glycemic markers

• Fasting plasma glucose (FPG): Reflects overnight glucose; requires fasting.
• Oral glucose tolerance test (OGTT): Reflects postprandial glucose handling; requires fasting and a glucose drink.
• Fructosamine: Reflects average glucose over 2-3 weeks (based on glycated serum proteins); useful when A1c is unreliable.
• Glycated albumin: Reflects average glucose over 2-3 weeks; less commonly used in the U.S.
• Continuous glucose monitor (CGM): Provides real-time glucose data; metrics include Time in Range, GMI, and coefficient of variation.
• A1c: Reflects average glucose over 2-3 months; most widely used but inaccurate in altered RBC turnover.

How to use the result

A patient with an A1c of 8.5% has an eAG of 197 mg/dL (10.9 mmol/L). This is above the postprandial target of <180 mg/dL, indicating that glycemic control is suboptimal. The clinical conversation should address lifestyle modification (Mediterranean-style diet, exercise, weight loss), medication intensification (consider adding a GLP-1 receptor agonist or SGLT2 inhibitor to existing metformin), and CGM use to identify patterns of hyperglycemia. The eAG helps the patient understand that an A1c of 8.5% means their average glucose is nearly 200 mg/dL, well above the targets they may have been told to aim for with fingerstick readings.