Estimated Average Glucose (eAG) Calculator

What estimated average glucose means

Estimated average glucose, or eAG, translates a laboratory A1C percentage into the units used by glucose meters and continuous glucose monitors. In the United States, eAG is usually reported in mg/dL. Many other settings use mmol/L. The number answers a narrow question: which mean glucose was associated with this A1C in the study used to derive the equation?

eAG is calculated rather than measured. A result of 154 mg/dL does not mean glucose stayed near 154.

One person may have a fairly narrow glucose profile while another alternates between lows and large post-meal rises. Both patterns can produce the same arithmetic mean and similar A1C. Some laboratories report eAG beside A1C because familiar glucose units can make a percentage easier to discuss. A1C remains the measured laboratory value. eAG does not replace fasting plasma glucose, an oral glucose tolerance test, current meter readings, or continuous glucose monitoring.

[1][2][5][6]

The eAG equation and unit conversion

The calculator uses the A1C-Derived Average Glucose equation: eAG in mg/dL = 28.7 × A1C − 46.7. To express the result in mmol/L, divide mg/dL by 18.0182. The reverse conversion is A1C (%) = (eAG in mg/dL + 46.7) / 28.7. Keep A1C as a percentage, not the IFCC value in mmol/mol, when using this equation.

The ADAG study collected about 2,700 glucose observations per A1C measurement over three months in 507 adults with type 1 diabetes, type 2 diabetes, or no diabetes. Its correlation coefficient was 0.92. The published ranges around each estimate show that an individual's actual mean can differ substantially.

Reporting 154 mg/dL is useful. Reporting 154.2 implies precision the equation cannot supply. A small calculated change may reflect normal A1C analytical and biological variation rather than a true change in average glucose.

[1][2][3]

Worked eAG examples

For A1C 7.0%, multiply 28.7 by 7.0 to get 200.9, then subtract 46.7. The result is 154.2 mg/dL, reported as about 154 mg/dL. Dividing by 18.0182 gives 8.56 mmol/L, reported as about 8.6 mmol/L.

For A1C 6.4%, the calculation is 28.7 × 6.4 − 46.7 = 137.0 mg/dL. That converts to 7.6 mmol/L.

The A1C value lies in the laboratory prediabetes range. The derived eAG has no independent diagnostic cutoff. Diagnosis rests on the measured A1C or a qualifying plasma glucose test.

[1][2][3]

Diagnosis, goals, and monitoring

A1C and plasma glucose have defined diagnostic roles. eAG does not.

For nonpregnant people, A1C of 5.7% to 6.4% meets the A1C criterion for prediabetes and 6.5% or higher meets a diabetes criterion, usually requiring confirmation. An eAG produced from those values is an explanatory conversion, not another test result.

For monitoring, the ADA considers A1C below 7% appropriate for many nonpregnant adults without severe or harmful hypoglycemia. That converts to an eAG near 154 mg/dL. Personal goals may be lower or higher based on health, function, diabetes duration, complications, treatment burden, life expectancy, and hypoglycemia risk. The calculator cannot select that goal. People with stable glucose within their goal range may have A1C checked about twice per year. Recent treatment changes, unstable readings, or values outside the agreed goal commonly call for assessment about every three months. eAG can translate each result into familiar units. Treatment decisions also require current glucose patterns and safety information.

[2][4][5]

eAG compared with meter and CGM averages

A meter average depends on when a person checks. Testing mainly before breakfast can miss post-meal highs and overnight lows, so the displayed mean may not represent the full day. A CGM average uses many more observations, but it still depends on sensor accuracy, adequate wear, calibration when required, and whether the selected date range overlaps the A1C period.

A1C reflects roughly two to three months and gives greater weight to recent weeks. A 14-day or 30-day device average can differ after a recent illness, medication change, steroid course, or improvement in glucose management. Before treating a mismatch as an error, compare the date ranges and ask whether the recent period was typical.

CGM adds measures that eAG cannot infer: time in the usual range of 70 to 180 mg/dL, time below 70 mg/dL, time above range, and glucose variability.

Glucose management indicator is also calculated from CGM mean glucose, but it uses a different evidence base and should not be labeled a laboratory A1C.

[2][6][8]

A1C assay and preanalytic limitations

The eAG equation inherits every limitation of the A1C entered.

Diagnostic A1C should come from an NGSP-certified method traceable to the Diabetes Control and Complications Trial assay. Point-of-care methods can be useful in monitoring, but diagnostic use is restricted to approved devices operated by trained personnel in qualifying laboratory settings.

Conditions that change red-cell lifespan can separate A1C from actual glucose: blood loss, transfusion, hemolytic anemia, iron deficiency, erythropoietin treatment, kidney failure, dialysis, pregnancy, and some liver disease. Some hemoglobin variants cause method-specific interference. NGSP publishes a method table for HbC, HbS, HbE, HbD, and elevated fetal hemoglobin.

Plasma glucose has its own preanalytic weakness because blood cells continue to consume glucose after collection unless the sample is processed promptly or preserved correctly. A1C is more stable after collection, but biological and assay effects remain. Discordance should prompt review of both measurements.

[4][5][7]

Population and personal factors

The original ADAG sample included adults and was 83% non-Hispanic White. Researchers found a strong overall relationship, yet the published intervals remained wide.

Pediatric goals, pregnancy targets, and goals for older adults require population-specific clinical guidance even though a calculator applies the same arithmetic to every entry. Self-identified race should not be used to adjust eAG. The ADA states that race and ethnicity are poor proxies for genetic factors that may alter A1C. A known G6PD variant, hemoglobin variant, anemia, or altered red-cell turnover provides more useful evidence than a race-based correction.

Kidney disease can affect A1C through anemia, erythropoietin exposure, dialysis, and shortened red-cell survival. Pregnancy changes red-cell turnover and has distinct glucose targets. In these settings, clinicians may emphasize plasma glucose, CGM, fructosamine, or glycated albumin according to the clinical question.

[2][4][5][7]

Interpreting a mismatch

Start with arithmetic. A laboratory report showing A1C 8.0% should produce an eAG near 183 mg/dL. A difference of one unit from rounding is unimportant. A report that uses IFCC A1C units in mmol/mol must first be converted to an A1C percentage; entering 64 as though it were 64% produces a meaningless result.

Compare eAG with a device mean from an overlapping period. Review missing CGM days, selective meter checks, recent changes, sensor alerts, and meter quality. A mismatch can reflect incomplete glucose sampling, a device problem, A1C interference, or stable personal variation in the A1C-glucose relationship.

A consistent personal gap can be clinically useful once a care team has excluded correctable causes. Treat population eAG as a group translation, not a precise personal prediction. Laboratory A1C, device patterns, hypoglycemia exposure, symptoms, and other validated markers can then be considered together.

[2][3][7][8]

How to use this eAG calculator

Enter the measured A1C percentage from a dated laboratory report. Confirm that the input is a percentage, such as 7.2, rather than an IFCC value in mmol/mol. Review the output in the same unit as your meter or CGM. Save the date and avoid comparing it with a device average drawn from a nonoverlapping week.

Use eAG to translate, not to diagnose or dose.

Bring the A1C report and device summary to a visit and ask whether the values agree closely enough for your clinical context. Discuss repeated discrepancies, recent transfusion, anemia, kidney or liver disease, pregnancy, hemoglobin variants, or medications that alter red-cell production. Arrange formal testing if you have no diagnosis and an A1C result is near or above a diagnostic threshold. Seek prompt assessment for excessive thirst, frequent urination, unexplained weight loss, vomiting, confusion, difficult breathing, or very high glucose with illness. eAG is a long-term summary and cannot rule out an acute glucose emergency.

[2][4][5][6]

How it works

eAG (mg/dL) = 28.7 × A1c − 46.7. Validated in the ADAG study.

Frequently asked questions

• Why convert A1c to eAG?eAG uses familiar glucometer units, helping patients understand what their A1c means in daily readings.

Related calculators

References

1. Nathan et al., ADAG Study Group. Translating the A1C assay into estimated average glucose values
2. American Diabetes Association Professional Practice Committee. Glycemic Goals, Hypoglycemia, and Hyperglycemic Crises: Standards of Care in Diabetes 2026
3. Bergenstal et al.. Relationship between A1C and glucose levels in the ADAG study
4. American Diabetes Association Professional Practice Committee. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes 2026
5. National Institute of Diabetes and Digestive and Kidney Diseases. The A1C Test and Diabetes
6. NGSP. NGSP: HbA1c and Estimated Average Glucose
7. NGSP. NGSP: Factors that Interfere with HbA1c Test Results
8. American Diabetes Association Professional Practice Committee. Diabetes Technology: Standards of Care in Diabetes 2026