HbA1c

Glucose in the blood spontaneously attaches to hemoglobin, the red pigment in red blood cells. The higher the average blood sugar, the larger the share of 'sugar-coated' hemoglobin (HbA1c). Because a red blood cell lives for about 3 months, HbA1c reflects the average blood sugar over the preceding 8 to 12 weeks, regardless of what was eaten that morning. The result is given in mmol/mol (formerly as a percentage: 48 mmol/mol corresponds to 6.5 percent).

• Diabetes mellitus: HbA1c of 48 mmol/mol or higher (WHO criterion; the NHG makes the diagnosis with fasting glucose)
• Increased risk of diabetes with values just below that cutoff (about 42 to 47 mmol/mol)
• How well known diabetes is controlled (target usually 53 mmol/mol or lower)

With an HbA1c of 48 mmol/mol or higher, or just below it, the GP has a fasting glucose measured (twice, on different days) to diagnose or rule out diabetes, and maps out the cardiovascular risk profile. With known diabetes, the treatment is adjusted if needed and the HbA1c is repeated at set times.

HbA1c is a well-standardized test with little influence from food, stress or time of day; it is the standard monitoring value in diabetes. For detecting diabetes the picture is more nuanced: the WHO and many countries use 48 mmol/mol or higher as a diagnostic criterion, but HbA1c finds fewer people with diabetes than glucose does: in the New Hoorn Study, at a cutoff of 48 mmol/mol, only 24 percent of people with diabetes according to the glucose tolerance test had a raised HbA1c, at a specificity of 99 percent. A raised HbA1c is therefore a strong indication; a normal HbA1c does not rule out diabetes. That is why in the Netherlands the diagnosis is made with fasting glucose.

• Anything that changes the lifespan of red blood cells disturbs the result: blood loss, hemolysis, pregnancy and a recently started iron or B12 treatment (falsely low); a removed spleen and severe kidney failure (falsely high).
• Hemoglobin variants such as sickle cell and thalassemia carriership, more common in people with a non-Western background, can make the measurement unreliable; the diagnosis must then be made on glucose.
• HbA1c lags behind: a blood sugar that has only recently risen (for example due to prednisone or diabetes that developed quickly) cannot be seen in it yet.
• Age and ethnic background affect the value slightly (differences of 2 to 3 mmol/mol), people differ in the degree to which their proteins become glycated, and the measurement method allows a spread of a few percent.

In 2024, about 1.2 million people in the Netherlands were known to their GP to have diabetes, of whom just over 1.1 million had type 2 (61 per 1000 inhabitants), men more often than women and mainly at older age. According to the Dutch GP guideline (NHG), about 30 percent of people over 60 have a precursor stage of diabetes (impaired fasting glucose or impaired glucose tolerance); at least half of them never develop diabetes.

Useful for monitoring known diabetes and as an additional measure for people at increased risk of diabetes (overweight, diabetes in the family, a previously impaired fasting glucose, previous gestational diabetes) who cannot or do not want to have a fasting blood draw. As a standalone screening test in healthy people without risk factors, HbA1c says less than fasting glucose, because it misses some of the people with diabetes. Not suitable with anemia, hemoglobin variants, pregnancy or a recently developed disturbance of blood sugar.

• A normal HbA1c can give false reassurance: it misses some of the people who do have diabetes according to the glucose criteria.
• A slightly raised value with a hemoglobin variant or after blood loss can cause unnecessary worry.

• NHG-Standaard Diabetes mellitus type 2 (M01)
• NVKC, allesovertesten.nl: HbA1c
• WHO (2011): Use of glycated haemoglobin (HbA1c) in the diagnosis of diabetes mellitus
• Thuisarts.nl: Ik heb een grotere kans op diabetes type 2
• VZinfo (RIVM): Diabetes mellitus, op basis van Nivel Zorgregistraties Eerste Lijn 2024