A1C Test Explained: Ranges, Accuracy Limitations, and CGM Correlation
A1C measures glycated hemoglobin over 2-3 months. Normal below 5.7%, prediabetes 5.7-6.4%, diabetes 6.5%+. Limitations, eAG conversion, GMI from CGMs, and when A1C disagrees with CGM data.
What Does the A1C Test Measure?
The A1C test — also called hemoglobin A1C, HbA1c, or glycated hemoglobin — measures the percentage of hemoglobin proteins in red blood cells that have glucose permanently attached to them through a non-enzymatic process called glycation. Because red blood cells have an average lifespan of 90-120 days, the A1C value reflects the average blood glucose concentration over the preceding 2-3 months. Higher average glucose means more hemoglobin molecules become glycated, producing a higher A1C percentage. The test requires a simple blood draw (venous or capillary) and does not require fasting — a significant practical advantage over fasting glucose tests. The A1C test was standardized in 2007 through the National Glycohemoglobin Standardization Program (NGSP) to ensure consistent results across laboratories. It is the primary diagnostic and monitoring test for diabetes worldwide, used by the American Diabetes Association (ADA), World Health Organization (WHO), and International Diabetes Federation (IDF) as the criterion for diagnosing diabetes and assessing long-term glucose control.
A1C Ranges: Normal, Prediabetes, and Diabetes
Three clinical ranges define A1C interpretation. Normal: an A1C below 5.7% corresponds to an estimated average glucose (eAG) below 117 mg/dL. This range indicates that glucose metabolism is functioning within healthy parameters, though it does not guarantee the absence of insulin resistance — a person can have significant insulin resistance with compensatory hyperinsulinemia and still maintain an A1C below 5.7%. Prediabetes: an A1C between 5.7% and 6.4% corresponds to an eAG of 117-137 mg/dL. Approximately 96 million American adults (38% of the adult population) fall in this range according to CDC 2022 data. Without intervention, 15-30% of people with prediabetic A1C will progress to type 2 diabetes within 5 years. An A1C of 6.0% carries approximately 10x the diabetes risk compared to an A1C of 5.0%. Diabetes: an A1C of 6.5% or higher (eAG above 140 mg/dL) on 2 separate tests constitutes a diabetes diagnosis. The ADA treatment target for most adults with diabetes is an A1C below 7.0%, corresponding to an eAG of approximately 154 mg/dL. Tighter targets of below 6.5% are appropriate for younger patients and those using CGMs who can safely achieve lower averages without frequent hypoglycemia.
The eAG Conversion Formula
The estimated average glucose (eAG) formula converts an A1C percentage into an average glucose value in mg/dL, making A1C results more intuitive for patients who check their glucose daily. The validated formula is: eAG (mg/dL) = 28.7 x A1C - 46.7. This equation was derived from the ADAG (A1C-Derived Average Glucose) study, which simultaneously measured A1C and mean glucose from CGM data in 507 participants. Examples: an A1C of 5.0% corresponds to an eAG of 97 mg/dL. An A1C of 5.7% converts to eAG of 117 mg/dL. An A1C of 6.5% equals eAG of 140 mg/dL. An A1C of 7.0% equals eAG of 154 mg/dL. An A1C of 8.0% equals eAG of 183 mg/dL. An A1C of 9.0% equals eAG of 212 mg/dL. The formula has a standard deviation of approximately plus or minus 15 mg/dL, meaning an individual with an A1C of 7.0% could have a true average glucose anywhere between 139 and 169 mg/dL. This imprecision is one reason CGM-derived metrics are increasingly preferred for day-to-day management.
Accuracy Limitations of A1C Testing
The A1C test has 5 clinically significant limitations that can produce misleading results. First, hemoglobin variants — including hemoglobin S (sickle cell trait, present in 8% of African Americans), hemoglobin C, hemoglobin E, and hemoglobin D — interfere with certain A1C assay methods and can falsely elevate or depress results by up to 1.0%. The NGSP maintains a list of which assay methods are affected by which variants. Second, conditions that alter red blood cell lifespan skew A1C results: iron deficiency anemia falsely elevates A1C because older RBCs accumulate more glycation, while hemolytic anemia, recent blood loss, blood transfusions, and erythropoietin therapy falsely lower A1C by introducing younger, less-glycated RBCs. Third, chronic kidney disease stages 4-5 produces carbamylated hemoglobin that interferes with A1C measurement. Fourth, pregnancy alters A1C due to hemodilution and increased RBC turnover — A1C underestimates true glucose in pregnant women by approximately 0.5%. Fifth, age-related increases in A1C occur independent of glucose levels: A1C rises approximately 0.03% per decade after age 30. These limitations affect an estimated 10-15% of the population and are the primary reason the ADA recommends confirming unexpected A1C results with a fructosamine test or CGM data.
CGM Correlation: GMI vs A1C
The Glucose Management Indicator (GMI) is a CGM-derived metric that estimates what a person's A1C would be based solely on their average sensor glucose over 14+ days. The formula is: GMI (%) = 3.31 + (0.02392 x mean CGM glucose in mg/dL). GMI was introduced in 2018 as a replacement for the older "estimated A1C" metric because it avoids implying that CGM data directly measures A1C. In theory, GMI and laboratory A1C should agree closely. In practice, they diverge in 25-40% of patients — a discrepancy that reveals important physiological information. When A1C is higher than GMI by more than 0.3%, the patient may have a hemoglobin variant, iron deficiency, or naturally higher glycation rate (some individuals glycate hemoglobin faster than average independent of glucose levels). When GMI is higher than A1C by more than 0.3%, the patient may have a condition that shortens RBC lifespan or may have had more high-glucose episodes during the 14-day CGM window than is typical for their 3-month average. Clinicians who use both A1C and CGM data can identify these discrepancies and investigate the cause rather than relying on either metric alone. The 2022 ADA Standards of Care recommend reporting both A1C and GMI when CGM data is available.
When A1C and CGM Disagree: What It Means
A persistent mismatch between laboratory A1C and CGM-derived GMI — defined as a difference greater than 0.5% — occurs in approximately 30% of CGM users and requires clinical interpretation. The most common scenario is high A1C with normal GMI, which occurs when a patient has a high glycation index (HGI). High glycators produce more glycated hemoglobin per unit of glucose exposure than average, meaning their A1C overstates their actual average glucose. A 2019 Diabetes Care study found that 10-15% of the population are high glycators with a HGI above 0.5%. For these individuals, CGM metrics (Time in Range, GMI, coefficient of variation) are more accurate reflections of glucose control than A1C. The reverse scenario — low A1C with high GMI — suggests the patient may have high glucose variability with significant time spent both very high and very low, which averages to a moderate A1C but masks dangerous swings. CGM data showing a coefficient of variation above 36% with a normal A1C is a red flag for this pattern. In both cases, the resolution is to prioritize CGM-derived metrics for daily management while using A1C as a complementary long-term trend indicator. Clinicians treating to an A1C target alone may over-treat high glycators (pushing them into hypoglycemia) or under-treat patients with high variability.