Glycated hemoglobin shows your average glucose over 3 months

Glycated hemoglobin (HbA1c) shows how much sugar has "stuck" to the hemoglobin in your red blood cells over the last 2–3 months [1]. It is like a "black box" that continuously records all your blood glucose levels, not just the ones you measure yourself. The higher and longer your blood glucose has been, the more hemoglobin becomes irreversibly glycated.

HbA1c is expressed in percentages (%) or in mmol/mol. These are two scales for the same test, not two different tests. The mmol/mol unit comes from the international IFCC standard and appears on the reports of some laboratories, next to the percentage. You can convert one into the other with the HbA1c calculator. In people without diabetes, HbA1c is below 5.7% (39 mmol/mol) [2]. For you, with type 1 diabetes, the target is below 7% (53 mmol/mol), individualized (higher or lower) by your doctor according to your age and other conditions [1]. An HbA1c of 7% (53 mmol/mol) corresponds to an average glucose over 2–3 months of approximately 154 mg/dL (8.6 mmol/L) [3]. Each 1% (11 mmol/mol) increase in HbA1c means roughly 29 mg/dL (1.6 mmol/L) more in your average glucose [3].

HbA1c is tested every three months if you are not yet at target or if you have made treatment changes [1]. Once your treatment is stable and you are at your HbA1c target, you can reduce to two tests per year. If your sensor data reach your doctor regularly, the test can be done even less often [4].

There is no point in testing more frequently than every three months. Red blood cells live for about 120 days, but HbA1c depends most on glucose from recent weeks, so it reflects the last 2–3 months [2]. In special situations (major treatment changes, preparation for surgery), your doctor may request more frequent tests. The cost of the test and whether it is reimbursed depend on the health system in your country and on the type of health insurance you have.

The standard target for most adults with type 1 diabetes is HbA1c below 7% (53 mmol/mol) [1]. This threshold generally offers the best balance between preventing complications and the risk of hypoglycemia. In children and adolescents, the target is 6.5% (48 mmol/mol) or less if they use a sensor and a closed-loop pump and can reach it safely [5]. Without these technologies, the target is 7% (53 mmol/mol) or less [5]. A target below 7.5% (58 mmol/mol) may be chosen for a child who cannot report hypoglycemia or does not feel it [6]. It may also be chosen for a child who has no access to a sensor or an advanced pump [6].

Your personal target may differ in either direction. It may be below 6.5% (48 mmol/mol) if you are healthy, without complications and without a significant risk of hypoglycemia [1] [5]. A more relaxed target, below 8% (64 mmol/mol), may be recommended if you have severe or frequent hypoglycemia, other serious illnesses or advanced complications [1]. In pregnancy you cannot rely on HbA1c alone. It remains useful, with its own targets, though it can be falsely low, so it is interpreted alongside your daily blood glucose readings. Discuss your individualized HbA1c target with your doctor.

HbA1c reflects the overall average over 2–3 months, while daily blood glucose readings are snapshots of the moment [1]. Be careful. You can have a good HbA1c but very high glycemic variability if you often swing from 40 mg/dL (2.2 mmol/L) to 300 mg/dL (16.7 mmol/L) and back. You can also have good readings on the glucometer but a high HbA1c because of hyperglycemia you do not catch (for example, at night or between tests).

HbA1c is a weighted average in which glucose from recent weeks counts more than glucose from three months ago [2]. HbA1c does not detect hypoglycemia and is in fact lowered by it, so you can end up with a good HbA1c alongside many dangerous low episodes [7]. That is why interpretation needs as much context as possible. HbA1c is like the overall average, and your daily readings are the individual grades.

A number of conditions can affect HbA1c, independently of blood glucose [8]:

• Iron deficiency anemia — falsely raises the result;
• Anemia caused by the breakdown of red blood cells (hemolytic) — lowers it;
• Red blood cell disorders, also called hemoglobinopathies (thalassemia, sickle cell disease) — interfere with the measurement, generally leading to falsely low values;
• Chronic kidney failure — can raise HbA1c through uremia, but dialysis falsely lowers it;
• Blood transfusions — radically change the result (a false decrease) [8];
• Pregnancy and the rapid turnover of red blood cells (red cells are renewed faster than normal, so they have less time to become glycated) — falsely low values;
• Some genetic hemoglobin variants — also give falsely low HbA1c values [2].

Discuss with your doctor if you have such conditions, so that other ways of assessing metabolic control can be used.

Yes. You can have an HbA1c of 7% (53 mmol/mol) both with stable glucose, around 154 mg/dL (8.6 mmol/L), and with extreme variability of 40–300 mg/dL (2.2–16.7 mmol/L) [3]. The average can be the same, but the risk of complications differs greatly. High glycemic variability is associated with increased cardiovascular risk for the same HbA1c.

That is why, alongside HbA1c, time in range also matters: the time spent between 70 and 180 mg/dL (3.9–10 mmol/L), measured by a glucose sensor [9]. The target is generally >70% time in range, below 4% of time spent below 70 mg/dL (3.9 mmol/L) and below 25% above 180 mg/dL (10 mmol/L). The sensor app also calculates the coefficient of variation, which shows how far your glucose readings stray from their average. A value below 36% indicates acceptable glycemic variability [9]. HbA1c remains important, but as you can see it does not tell the whole story.

The simple formula for estimating average glucose is: Average (mg/dL) = 28.7 × HbA1c(%) − 46.7 [3]. As a rough guide:

• HbA1c 6% (42 mmol/mol) → 126 mg/dL (7 mmol/L);
• HbA1c 7% (53 mmol/mol) → 154 mg/dL (8.6 mmol/L);
• HbA1c 8% (64 mmol/mol) → 183 mg/dL (10.2 mmol/L);
• HbA1c 9% (75 mmol/mol) → 212 mg/dL (11.8 mmol/L);
• HbA1c 10% (86 mmol/mol) → 240 mg/dL (13.4 mmol/L).

Remember that this is the overall 24/7 average, including the periods when you do not test (at night or between meals). It is not the average of your tests. The average of glucometer tests can be lower than your true average glucose, because you miss some of the peaks after meals. For maximum accuracy, a glucose sensor gives you the real average, calculated from thousands of small measurements, as the figure below shows [9].

The target below 7% (53 mmol/mol) comes from the classic DCCT study, carried out in 1993 [10]. It showed that intensive treatment, with a median HbA1c of 7% (53 mmol/mol) versus 9% (75 mmol/mol), reduces early microvascular complications (retinopathy, nephropathy, neuropathy) by 35–76% [11]. The benefit persisted long after metabolic control was relaxed once the study ended ("metabolic memory"). Patients in the intensive group had fewer complications even 30 years after the intervention [11].

The lower the HbA1c, the lower the risk of complications, and the difference in HbA1c explained almost all of the difference in complications between the groups [11]. The difference between 9% (75 mmol/mol) and 7% (53 mmol/mol) means roughly halving the risk. There is an additional benefit for HbA1c values below 7% (53 mmol/mol), but it is smaller, and the risk of hypoglycemia rises sharply without a glucose sensor [7]. That is why an HbA1c of 7% (53 mmol/mol) offers the best risk–benefit ratio for most patients.

Yes, an HbA1c below 6% (42 mmol/mol) is very low. In a patient on treatment that carries a risk of hypoglycemia (in type 1 diabetes, insulin), it may indicate frequent low episodes, some possibly unnoticed [7]. The ACCORD study was stopped early because aggressive targeting below 6% (42 mmol/mol) increased mortality in patients with type 2 diabetes on such treatment [12]. That finding does not transfer automatically to type 1 diabetes. For people with type 1 diabetes, it matters whether you use a glucose sensor. Without modern insulins and technologies, the lower the HbA1c, the higher the risk of severe hypoglycemia [7]. The risk rises by about 27% for every 10% relative decrease in HbA1c [7]. For example, a decrease in HbA1c from 8% (64 mmol/mol) to 7.2% (55 mmol/mol) is a 10% relative decrease.

There are also some exceptions. In the honeymoon period, or near the onset of the LADA (latent autoimmune diabetes in adults) form of type 1 diabetes, your HbA1c can safely be nearly normal. The reason is good residual insulin production. With access to modern technology (sensors, advanced closed-loop pumps), some patients reach an HbA1c below 6.5% (48 mmol/mol) without significant hypoglycemia [5]. Assessment together with your doctor is essential for setting a treatment plan that gives you the best possible HbA1c.