Hemoglobin A1c test: what it measures, normal levels, prediabetes and diabetes cut-points, and what can skew the result

Glucose in the blood sticks to proteins it meets, including the hemoglobin inside red blood cells. The attachment is irreversible for the life of the cell, so the higher the average glucose a red cell has been bathed in, the more of its hemoglobin carries a sugar. The A1c test measures what percentage of hemoglobin has been glycated in that way. Because red cells survive roughly 120 days and are constantly being replaced, the percentage is a rolling average of glucose exposure over the previous two to three months, with the most recent month counting for about half of the result.

That averaging is the test's strength. A fasting glucose describes one moment and can be thrown off by a bad night's sleep or a stressful morning; an A1c smooths those out and does not require fasting or timing. Laboratories also report an estimated average glucose (eAG) alongside the percentage, converting it into the mg/dL units of a glucose meter so the two can be compared. An A1c of 6.0% corresponds to an average glucose of about 126 mg/dL; 7.0% to about 154; 8.0% to about 183.

The same averaging is also its limitation. A1c cannot show how much glucose swings during the day, cannot detect episodes of low blood sugar, and depends on red blood cells living a normal lifespan, so anything that shortens or lengthens that lifespan, or changes the hemoglobin itself, changes the result without changing the glucose. The section on what can skew A1c covers those situations.

A1c itself has no symptoms — it is a three-month average of blood glucose — but the high glucose it reflects does, and they build so gradually that many people with type 2 diabetes have had it for years before diagnosis:

In the prediabetes range there are usually no symptoms at all — which is exactly why the test, not symptoms, is how prediabetes is found, and why screening is recommended for adults from 35 and earlier with risk factors.

A1c can read falsely low in conditions that shorten red-cell life (hemolysis, recent blood loss, some anemias) and falsely high in iron deficiency or when red cells live longer; hemoglobin variants can also interfere. A fasting glucose or glucose tolerance test is used when A1c may be unreliable.

The American Diabetes Association recommends screening every adult from age 35, repeating at least every three years when the result is normal, and starting earlier and testing more often in adults with overweight or obesity plus any additional risk factor: a first-degree relative with diabetes, high blood pressure, low HDL or high triglycerides, polycystic ovary syndrome, a history of gestational diabetes, cardiovascular disease, physical inactivity, or belonging to a higher-risk population. Anyone who has ever had prediabetes should be tested yearly.

Symptoms that call for a test regardless of age are the classic ones of high glucose: unusual thirst, frequent urination, unexplained weight loss, blurred vision, recurrent infections, slow-healing sores, and tingling or numbness in the feet. People taking medicines that raise glucose, such as long-term steroids or some antipsychotics, and people on GLP-1 medicines or other weight or metabolic programs who want to see the metabolic effect, are common reasons for testing as well.

For people already diagnosed with diabetes, A1c is the standard measure of long-term control: at least twice a year when glucose is stable and at goal, and every three months when treatment has changed or goals are not being met. It is not the right tool for pregnancy after the first trimester, for suspected type 1 diabetes with rapid onset, or for anyone with a condition that distorts red cell turnover; those situations use glucose measurements directly.

A1c is one of the most widely available blood tests. It is part of most diabetes and metabolic panels and can be ordered on its own by your doctor, or online through TestWell for a draw at a Quest or Labcorp patient service center, with a licensed physician reviewing the order and results in your portal within a few business days. Laboratory A1c is run on methods certified by the NGSP, the standardization program that keeps results comparable across labs and aligned with the trials the cut-points come from.

No fasting is required and time of day does not matter, which is why A1c is often the first test when someone wants a single convenient draw. Eating, exercise or a stressful week before the test will not move it meaningfully; the result reflects months, not days. What does matter is the red-cell context: tell the ordering physician about anemia, a recent transfusion or significant blood loss, kidney disease, a known hemoglobin variant such as sickle cell trait, iron or B12 treatment, pregnancy, and high-dose vitamin C or E supplements.

Fingerstick A1c meters used in clinics and pharmacies are convenient for monitoring but, unless the device is certified for the purpose, results from them are not used to diagnose diabetes; a laboratory test is. A diagnosis also requires either two abnormal results, from the same sample or two separate samples, or one abnormal result together with unmistakable symptoms and a very high glucose.

Below 5.7% is normal. From 5.7% to 6.4% is prediabetes, a range in which the risk of progressing to diabetes and of cardiovascular disease is already raised and in which lifestyle change has its strongest evidence. An A1c of 6.5% or higher, confirmed by a second test unless glucose is unequivocally high, is diabetes. These are the American Diabetes Association thresholds, adopted from the level at which diabetic eye disease begins to appear in population studies, and they apply to laboratory tests on NGSP-certified methods.

Two features of the scale are worth knowing. It is continuous, so 5.6% is not safe and 5.8% is not sick; risk rises smoothly through the prediabetes band, and a result that has climbed from 5.2% to 5.6% in two years is telling you something even though both numbers are normal. And the cut-points are for diagnosis, not treatment goals: someone treated for diabetes with an A1c of 6.8% is above the diagnostic threshold but comfortably at goal.

The estimated average glucose in the table is the standard conversion laboratories print beside the percentage. It is an average, not a fasting value, so it will usually sit above the fasting glucose numbers a home meter shows in the morning.

For most non-pregnant adults with diabetes the ADA recommends an A1c under 7%, which corresponds to an average glucose of about 154 mg/dL and is the level at which the long-term complication trials showed clear benefit. A tighter goal, under 6.5%, is reasonable when it can be reached without significant hypoglycemia or medication burden, typically early in the disease, with lifestyle change or metformin alone, or with medicines that do not cause low blood sugar. A looser goal, under 8%, is appropriate for people with a history of severe hypoglycemia, limited life expectancy, advanced complications or extensive other illness, where the risks of intensive treatment outweigh the long-term gains.

A1c moves slowly by design. Because it averages three months, a change in treatment takes about that long to show fully, and roughly half of the eventual change is visible at six weeks. Testing more often than every three months rarely adds information. Where day-to-day detail matters, in insulin-treated diabetes especially, continuous glucose monitoring supplies the variability and hypoglycemia data that A1c cannot, and its glucose management indicator (GMI) is read alongside the laboratory A1c rather than instead of it.

Discrepancies between A1c and glucose readings are common enough to expect. A meter or CGM average that is clearly lower than the A1c implies, or an A1c that falls faster than the glucose data can explain, usually means one of the red-cell factors below is at work, and the clinician will lean on the glucose data.

Anything that changes how long red cells live changes A1c independently of glucose. Cells that live longer than normal accumulate more sugar, so iron-deficiency anemia, vitamin B12 or folate deficiency, a spleen removed, and some cases of chronic kidney disease read falsely high. Cells that are replaced faster carry less, so hemolytic anemias, recent significant blood loss, a recent transfusion, treatment with erythropoietin or iron or B12 (which floods the blood with young cells), the second and third trimesters of pregnancy, advanced liver disease and some HIV medicines read falsely low. Very high doses of vitamin C or E, aspirin and alcohol can interfere with certain assays.

Inherited hemoglobin variants are the other large group. Sickle cell trait, hemoglobin C, D and E traits and thalassemias can push results up or down depending on the laboratory method; modern NGSP-certified assays handle most common variants, but a known variant should be flagged when the test is ordered so the lab can choose a method that is not affected. Studies have also found that, at the same measured glucose, average A1c runs modestly higher in some groups, including Black Americans, which is one reason the ADA advises interpreting a borderline result together with glucose rather than in isolation.

When A1c cannot be trusted, glucose measurements take over: fasting plasma glucose, a two-hour oral glucose tolerance test, or continuous glucose monitoring. Fructosamine, which reflects glycated blood proteins over the previous two to three weeks and does not depend on red cells, is the usual laboratory alternative for monitoring in these situations.

A first result in the prediabetes range is confirmed with a repeat and then rechecked yearly. The intervention with the strongest evidence is structured lifestyle change: the Diabetes Prevention Program showed a roughly 58% reduction in progression to diabetes over three years with a program aimed at about 7% weight loss and 150 minutes of weekly activity, and CDC-recognized versions of that program are widely available. Blood pressure, lipids and, where relevant, sleep apnea and liver fat are assessed at the same time, because prediabetes rarely travels alone.

A first result of 6.5% or higher is confirmed with a second test (a repeat A1c, a fasting glucose or an oral glucose tolerance test) unless symptoms and a very high glucose make the diagnosis obvious. Once confirmed, the initial visit covers kidney function and urine albumin, a lipid panel, blood pressure, a foot and eye examination, and a conversation about goals and medication. An A1c well above 10%, or any high result with heavy thirst, weight loss, vomiting or confusion, needs same-day medical attention rather than a scheduled follow-up.

For someone already treated, an A1c drifting above goal triggers a review of medication, adherence, diet and activity, and a recheck in three months; a result unexpectedly far below goal in a person on insulin or a sulfonylurea prompts a look for unrecognized hypoglycemia. An A1c that does not match the glucose data is handled by checking for the red-cell factors above and, if needed, switching the monitoring to glucose-based measures.