HbA1c measures the proportion of hemoglobin that has glucose attached to it, which reflects average blood glucose over roughly three months. Averages hide shape. Two people with an identical A1C of 6.0 percent can have completely different glucose patterns, one flat and one swinging from 70 to 220 mg/dL after meals, and the A1C cannot tell them apart.

What an average hides

The clinical consequence is real. An A1C in the normal range does not exclude impaired glucose tolerance, which is defined on a two hour value of 140 to 199 mg/dL after a 75 g glucose load. Early type 2 diabetes typically shows up first in post meal glucose, and someone with a normal fasting glucose and a normal A1C can still have a two hour value above 200 mg/dL. If the pretest suspicion is high, the oral glucose tolerance test answers the question that A1C cannot.

When A1C is simply wrong

A1C depends on red blood cells living a predictable length of time. NIDDK notes that conditions changing the life span of red blood cells, including recent blood loss, sickle cell disease and transfusion, can change A1C levels, that iron deficiency anemia can produce a falsely high result, and that kidney failure or liver disease can also cause false results. NIDDK adds that in people of African, Mediterranean or Southeast Asian descent, or with a family history of sickle cell anemia or thalassemia, an A1C test can be unreliable for diagnosing or monitoring diabetes.

Situation	Effect on A1C	Better test
Iron deficiency anemia	Falsely high	Fasting glucose or OGTT
Recent blood loss or transfusion	Unreliable	Fasting glucose, glucose monitoring
Hemoglobin variant or thalassemia	May be invalid	Fasting glucose or OGTT
Pregnancy	Reflects pre-pregnancy months	OGTT per obstetric protocol
Advanced kidney or liver disease	Unreliable	Fasting glucose, glucose monitoring

What A1C is still good for

A1C remains a diagnostic test in the 2026 ADA Standards of Care at 6.5 percent or above, and the practical tool for tracking treatment over months. It needs no fasting, is stable in the tube, and predicts complication risk in large populations. The problem is not the test. The problem is treating one number as the whole picture of glucose control.

If your A1C is normal but you have symptoms after meals, a strong family history, PCOS or previous gestational diabetes, ask for a two hour oral glucose tolerance test. If you have anemia, kidney disease, a hemoglobin variant or a recent transfusion, ask your clinician whether A1C is even valid for you before it is used to make decisions.

A1C used for diagnosis should be measured with a laboratory method certified by the National Glycohemoglobin Standardization Program and standardized to the DCCT assay. Where the clinical picture and the A1C disagree, the glucose criteria decide.

Where a hemoglobin variant is present, some assay methods are affected and others are not, so the reporting laboratory should be asked whether its method is valid for that variant before the result is used. NIDDK notes that conditions changing the life span of red blood cells, including recent blood loss, sickle cell disease and transfusion, can change A1C levels, that iron deficiency anemia can give a falsely high result, and that kidney failure or liver disease can also produce false results. In any of those settings, diagnose and monitor with plasma glucose instead.

What does HbA1c miss?

What an average hides

When A1C is simply wrong

What A1C is still good for

Sources

What an average hides

When A1C is simply wrong

What A1C is still good for

Sources

Related