HbA1c (glycated haemoglobin): what's normal and what it means

Most routine check-ups include a fasting blood glucose test. It is useful — but it captures only a moment: the state of your blood sugar on the particular morning you showed up after not eating overnight. If you slept poorly, were under stress, or had a bout of illness in the days before, the number shifts accordingly. A few weeks of careful eating before your appointment can also pull the result into the normal range even if your usual pattern is less healthy.

HbA1c (glycated haemoglobin, also written A1c or HbA1c) solves this problem. Rather than a snapshot, it provides a retrospective average of blood glucose over the previous 2–3 months. Because red blood cells survive approximately 90–120 days before being replaced, they carry a cumulative record of every glucose level the blood has seen during that period — and that record is what the test measures. No fasting required. No special preparation. Time of day does not matter.

That combination of convenience and stability makes HbA1c a cornerstone marker for both screening people who may have undiagnosed diabetes and monitoring treatment in those who already carry the diagnosis.

What HbA1c is and how it works

Haemoglobin is the protein inside red blood cells that carries oxygen. In the bloodstream, haemoglobin is in constant contact with glucose. A fraction of haemoglobin molecules undergo an irreversible reaction with glucose — a process called glycation, or non-enzymatic glycosylation.

The key principle: the higher the average blood glucose level, the greater the proportion of haemoglobin that becomes glycated. Since a red blood cell lives for roughly 90–120 days before being cleared and replaced, the HbA1c level integrates the glycaemic load over the entire lifespan of the cell — approximately the past 2–3 months.

Results are expressed in two equivalent scales:

DCCT/NGSP % — the percentage scale used in the United States and in most clinical guidelines worldwide.
IFCC mmol/mol — the international units scale adopted in Europe and WHO publications.

Both describe exactly the same thing. The conversion formula is: HbA1c (%) ≈ HbA1c (mmol/mol) × 0.0915 + 2.15, or in the other direction: HbA1c (mmol/mol) ≈ (HbA1c (%) − 2.15) / 0.0915. Laboratory reports often list both; that is normal.

Normal HbA1c range

The cut-offs below reflect the diagnostic criteria of the American Diabetes Association (ADA) and the WHO.

Category	HbA1c (%)	HbA1c (mmol/mol)	Interpretation
Normal	< 5.7%	< 39	Glycaemic profile within the healthy range
Prediabetes	5.7 – 6.4%	39 – 46	Elevated risk; lifestyle modification recommended
Diabetes	≥ 6.5%	≥ 48	Diagnostic threshold (confirmation required)

A note on confirmation. A single result at or above 6.5% is not, on its own, a diagnosis of diabetes — unless the person already has clear symptoms of hyperglycaemia (excessive thirst, frequent urination, unexplained weight loss). Without those symptoms, a diagnosis requires either a repeat HbA1c on a different day or a corroborating raised fasting glucose. Two concordant results establish the diagnosis; one alone does not.

The cut-offs are the same for adults of all ages. Target values for people who already have diabetes, however, are personalised — those are covered below.

HbA1c vs fasting glucose

These two tests measure different aspects of the same underlying problem, and neither replaces the other completely.

Fasting glucose is a snapshot. It can be spuriously elevated by acute illness, stress, a sleepless night, or certain medications such as corticosteroids. It can also be spuriously reassuring if a patient happened to eat unusually carefully in the days before the test. The test demands at least 8 hours without food.

HbA1c is a 2–3 month retrospective average. It is insensitive to what you ate yesterday or whether you had a stressful morning. However, it has its own limitations — covered in the next section.

OGTT (oral glucose tolerance test) is a third option: a standardised 75 g glucose load followed by a 2-hour glucose measurement. It is considered the most sensitive test for detecting impaired glucose tolerance (prediabetes) but is more demanding to perform and is rarely used as a first-line screen. For a detailed comparison of fasting glucose norms and OGTT thresholds, see the companion article blood glucose ranges.

For most screening situations, one or both of the first two tests is sufficient. When results are borderline, the OGTT adds useful information.

When HbA1c is unreliable

HbA1c reflects glucose averaged over the lifespan of a red blood cell. When that lifespan — or the structure of haemoglobin itself — is altered, the test gives inaccurate results.

Haemoglobinopathies. Sickle-cell trait or disease (HbS), haemoglobin C (HbC), haemoglobin E (HbE), and thalassaemia trait all distort HbA1c readings. Depending on the specific variant and the laboratory’s measurement method, the result can be falsely elevated or falsely low. If you have a known haemoglobin variant, tell your doctor before ordering an HbA1c test.

Iron-deficiency anaemia. Iron deficiency slows the production of new red blood cells, leaving older cells — which have had longer to accumulate glycation — in circulation for a greater fraction of time. In most assay methods this raises HbA1c by a small but clinically meaningful amount even when blood glucose itself is normal; the direction of the bias can vary by laboratory method (some HPLC variants instead produce a falsely lowered result). After correcting the iron deficiency, HbA1c typically returns to its true level.

Recent blood transfusion. Donor red blood cells have not been exposed to the recipient’s blood glucose history, so they dilute the glycated fraction and falsely lower the HbA1c. Any transfusion in the previous 2–3 months makes the result unreliable.

Haemolysis. When red cells are destroyed faster than normal — due to autoimmune haemolysis, malaria, or mechanical causes — the average cell age falls, and HbA1c is falsely low.

Pregnancy. In the second and third trimesters, red blood cell turnover increases, which can lower HbA1c independently of blood glucose. For that reason, gestational diabetes is screened using the OGTT, not HbA1c.

Advanced kidney disease. Renal failure is associated with both anaemia and chemical modifications to haemoglobin that can distort HbA1c in either direction.

In all of these situations, alternatives such as fructosamine (reflecting glycaemic control over the previous 2–3 weeks) or continuous glucose monitoring provide more reliable information. The choice of method is always made by the clinician.

Understanding why red cell parameters matter for HbA1c interpretation is another reason to read it alongside a full blood count (CBC): abnormalities in MCV, MCH, or haemoglobin concentration immediately explain why an HbA1c result might be atypical.

Target values for people with diabetes

For most adults with type 2 diabetes, the treatment target is an HbA1c below 7.0% (< 53 mmol/mol). This threshold is grounded in data from the UK Prospective Diabetes Study (UKPDS, Lancet 1998): maintaining HbA1c at around 7% significantly reduced the incidence of microvascular complications — retinopathy, nephropathy, and neuropathy — compared with less intensive control.

Targets are not one-size-fits-all, however. Most guidelines recommend individualising the goal based on the patient’s overall situation:

Patient profile	HbA1c target
Younger adults without significant comorbidities	< 6.5% (< 48 mmol/mol)
Most adults with type 2 diabetes	< 7.0% (< 53 mmol/mol)
Older adults, or those with serious comorbidities or hypoglycaemia risk	< 8.0% (< 64 mmol/mol)

A less stringent target in older or frailer patients is not a lower standard of care — it is protection against the real danger of hypoglycaemia, which in that population can cause falls, cardiac events, and loss of consciousness. The right balance is a clinical decision made together with the treating physician.

For type 1 diabetes, targets are set individually, accounting for hypoglycaemia frequency and the patient’s ability to monitor glucose.

How to prepare for the test

HbA1c is one of the rare blood tests for which fasting is not required. You can eat, drink coffee, and take your usual medications before the test. Time of day is not clinically significant.

A few caveats:

Recent acute illness or surgery? If possible, wait until you have recovered before testing — not because acute illness distorts HbA1c directly (unlike fasting glucose), but because your clinician will want a stable baseline rather than a reading taken during physiological stress.
Blood transfusion in the last 2–3 months? Inform your doctor. The result may be unreliable.
Pregnant? HbA1c is not used to screen for gestational diabetes. An OGTT is performed at 24–28 weeks.

The test uses a standard venous blood sample. Some clinics offer point-of-care HbA1c measurement from a fingerstick capillary sample (POCT device); the result may differ slightly from a laboratory venous result. For formal monitoring, a certified laboratory venous sample is preferred.

If you want to understand how to read your full laboratory report — not just HbA1c — the how to read lab results guide explains the structure of a lab report and what the reference ranges mean.

How HealthLab helps you track HbA1c

A single HbA1c result tells you which category you fall into. A trend over time answers the more important question: is your glycaemic control improving, staying stable, or creeping upward? For people with prediabetes, following HbA1c every 6–12 months turns an abstract risk into a concrete, actionable signal. For those managing diabetes, it is the primary objective measure of whether therapy is working.

Unexplained fatigue is one of the most common reasons people discover elevated glucose markers — HbA1c is a standard part of the fatigue workup panel precisely because early hyperglycaemia often presents as exhaustion before any other symptom appears.

HealthLab automatically recognises HbA1c and blood glucose from PDF lab reports issued by any laboratory and plots a trend chart over time. You see how the marker moves across visits — without manual data entry or spreadsheets.

Download HealthLab on the App Store

Frequently Asked Questions

Do I need to fast before an HbA1c test?

No. Unlike fasting glucose, HbA1c is not affected by what you ate before the test, the time of day, or yesterday’s meals. You can eat and drink normally. The only exception: if your doctor has ordered HbA1c together with fasting glucose or a lipid panel as part of a combined test panel, follow the preparation instructions for the full panel — usually an early-morning draw after at least 8 hours without food.

How often should I test HbA1c if I have diabetes?

If your blood glucose is stable and you are meeting your HbA1c target, testing twice a year is typically sufficient. If control is unstable or you have recently changed therapy, testing every 3 months gives a more timely picture — because 3 months is the full window HbA1c reflects. For people with prediabetes, testing every 6–12 months is generally appropriate; your doctor will advise based on your specific risk level.

Can HbA1c be normal even if someone has diabetes?

Yes, in several situations. First, in conditions that falsely lower HbA1c — haemolysis, recent blood transfusion, or certain haemoglobin variants. Second, in people whose blood sugar swings widely: alternating episodes of low and high glucose can average out to a deceptively normal HbA1c, while the peaks and troughs still cause harm. This is one reason HbA1c complements rather than replaces self-monitored blood glucose or continuous glucose monitoring.

What does an HbA1c of 6.0% mean if I have no symptoms?

6.0% falls in the prediabetes range (5.7–6.4%). It is not a diagnosis of diabetes and does not, on its own, require medication. It is a signal: your risk of developing type 2 diabetes is elevated above background. Most guidelines recommend addressing it through lifestyle changes — dietary adjustments, regular aerobic activity, and weight loss of 5–7% of body weight if you have excess weight. With consistent effort, a meaningful proportion of people with prediabetes return to the normal range. Discuss a specific action plan with your primary care doctor or endocrinologist, and repeat the test in 6–12 months to see whether the intervention is working.

This material does not replace clinical advice. Interpreting laboratory results always requires clinical context.