Key Takeaways
- An HbA1c test reflects average glucose over about 2 to 3 months, with the most recent weeks counting for more.[1][2]
- The US prediabetes range starts at 5.7% (39 mmol/mol), while the UK high-risk band starts at 42 mmol/mol (6.0%).[1][3]
- You do not need to fast for an HbA1c test.[2]
- Testing is usually every 3 to 6 months for people with diabetes, and at least once a year for the UK high-risk band.[4][5][3]
- Anaemia, kidney disease, pregnancy and some haemoglobin variants can make the result unreliable.[6]
The HbA1c test shows your average blood glucose over the past 2 to 3 months.[1][2] It needs one blood sample and no fasting. The cut-offs depend on where you live. In the US the prediabetes range starts at 5.7% (39 mmol/mol). In the UK the high-risk band starts at 42 mmol/mol (6.0%).[1][3] This guide covers the numbers, the testing schedule and the cases where a result can mislead.
This guide is general education. If you use insulin or other glucose medicines, follow the plan your diabetes team gave you.
What is an HbA1c test?
It is a blood test that reports average glucose over roughly the last 3 months.[7] Red blood cells live about 120 days, and the result is a weighted average that leans on recent weeks.[1] NIDDK says the past 30 days count for more than earlier months.[2]
The test is also less likely to be thrown off by short-term changes than a fasting glucose test.[2] The ADA lists fewer day-to-day swings from stress, food or illness as one advantage.[1] Our guide to high blood sugar symptoms covers what to watch for between tests.
What is the normal HbA1c range?
In the US a normal HbA1c is below 5.7% (39 mmol/mol). NIDDK uses the word normal and the ADA tables do not.[2][1] In the UK the cut-offs differ. NICE classes 42 to 47 mmol/mol (6.0 to 6.4%) as high risk.[3]
HbA1c cut-offs for adults without a diabetes diagnosis, US and UK
| Result. | What the evidence says. | Practical point. |
|---|---|---|
| Below 5.7% (39 mmol/mol, US).[2][1] | NIDDK calls this normal.[2] | The ADA says risk is continuous and rises toward the top of each band.[1] |
| 5.7 to 6.4% (39 to 47 mmol/mol, US).[1] | The ADA calls this the prediabetes range.[1] | Ask your GP or primary care team when to repeat the test.[1] |
| 42 to 47 mmol/mol (6.0 to 6.4%, UK).[3] | NICE says this indicates high risk of type 2 diabetes.[3] | NICE advises a blood test at least once a year for this group.[3] |
| Below 42 mmol/mol (6.0%, UK).[3] | This sits outside the NICE high-risk band.[3] | Ask your GP or primary care team how to read your own result.[3] |
| 6.5% or above (48 mmol/mol or above, US and UK).[1][3] | The ADA sets 6.5% (48 mmol/mol) as the diabetes threshold and NICE flags it as possible type 2 diabetes.[1][3] | Without clear symptoms a second test is needed.[1] |
These are screening and diagnosis bands. A single result is not a diagnosis.[2] The HbA1c normal range also does not apply in pregnancy or with some blood conditions, which we cover below.
Targets are different from the normal range
Targets apply only to people who already have diabetes. The ADA says an A1C goal below 7% (53 mmol/mol) suits many nonpregnant adults.[6] NICE suggests 48 mmol/mol (6.5%) for adults with type 2, or 53 mmol/mol (7.0%) if a medicine carries a hypoglycaemia risk.[4] Agree your own target with your care team.
How is HbA1c different from a blood sugar reading?
HbA1c is an average and a finger-prick reading is a snapshot. Diabetes UK describes the first as 2 to 3 months of glucose and the second as a single moment.[5] A1C also does not show glycemic variability, real-time glucose or hypoglycemia, the ADA adds.[6]
HbA1c vs blood sugar: how they connect
The link between the two comes from the ADAG study. It followed 507 adults with and without diabetes and gave a formula that turns A1C into an estimated average glucose.[8] By the ADA table 7% (53 mmol/mol) matches 154 mg/dL (8.6 mmol/L) and 6% (42 mmol/mol) matches 126 mg/dL (7.0 mmol/L).[6]
Take that as a rough guide. The 7% (53 mmol/mol) figure carries a range of 123 to 185 mg/dL (6.8 to 10.3 mmol/L).[6] The data come from one study published in 2008, the ADA notes.[6] Meter, sensor and drug makers helped fund that study.[8]
Sensor users may see a glucose management indicator, or GMI. It estimates A1C from continuous glucose monitoring, yet the authors say people with the same GMI will not all have the same lab value.[9] The authors disclose ties to device makers. The ADA says not to use GMI estimates in pregnancy.[10] Read more in our technology and diabetes guide.
If you use insulin or a sulfonylurea such as gliclazide or glimepiride, an HbA1c result cannot warn you about day-to-day lows.[6][11] Low blood sugar usually means below 4 mmol/L (72 mg/dL) in the UK or below 70 mg/dL (3.9 mmol/L) in the US.[11][12] Our low blood sugar guide explains the signs. In the UK, the NHS says to call 999 if someone may have very low blood sugar and is not responding normally or is unconscious, and a glucagon injection is not available or you do not know how to use it, or they have not recovered within 10 minutes of a glucagon injection, or they have been drinking alcohol.[11] In the US, NIDDK says to call 911 right away if someone is unconscious or cannot swallow safely, or right after giving glucagon.[12]
Write down each result with its date and unit. NICE prints both, for example 42 to 47 mmol/mol (6.0 to 6.4%), so note both if your lab gives them.[3]
How often should HbA1c be checked?
For people with diabetes the ADA advises testing at least twice a year, and about every 3 months if goals are not met or treatment has changed.[6] Diabetes UK says most people have the test every three to six months and are entitled to one at least once a year.[5]
NICE gives more detail for the UK. Adults with type 2 diabetes are tested every 3 to 6 months until the result is stable, then every 6 months.[4] Adults with type 1 diabetes are tested every 3 to 6 months.[13] The schedule you get may differ, so follow your own team.
People in the UK high-risk band are offered a blood test at least once a year, ideally using the same type of test.[3] If a result is close to the diagnostic threshold the ADA advises repeating it in 3 to 6 months.[1] Our guide to self-monitoring covers what to do between tests.
Do you need to fast for an HbA1c test?
No. NIDDK says you do not have to fast, so blood can be drawn at any time of day.[2] You can eat and drink before it.[7] The ADA lists the test as unaffected by recent food intake, stress, illness or activity.[1]
Fasting belongs to other tests. The fasting plasma glucose test is given in the morning after at least 8 hours without food.[7] If your clinician orders more than one test on the same day, ask which ones need fasting.
- Ask whether any other blood test ordered for the same visit needs fasting.[7]
- Eat and drink as usual before an HbA1c test, because no preparation is needed for it, and do not change any medicine before a blood test unless your clinician tells you to.[1][7]
- Record your result with both units (mmol/mol and %) and the date.[3]
- Ask your GP or diabetes team whether a second lab test is needed to confirm a new diagnosis.[1][2]
Can anything make an HbA1c test unreliable?
Yes. Per the ADA, anything that changes red blood cell turnover can interfere with accuracy. Examples are anaemias, recent transfusion, kidney failure and pregnancy.[6] Some haemoglobin variants also affect the result.[7] If your result and your readings disagree, your clinician may rely on other tests.[7] NIDDK adds that the test should not be used to diagnose type 1 diabetes or gestational diabetes.[2]
Iron deficiency is a good example. It can shift the result, but studies disagree on which way. A 2026 meta-analysis found no conclusive overall difference and called the direction and size of any effect inconsistent.[14]
Pregnancy changes the picture. The ADA says A1C runs slightly low in pregnancy because of increased red blood cell turnover.[10] NICE advises against routine HbA1c use to judge control in the second and third trimesters.[15] Follow your maternity team and do not use the ranges above.
Questions to ask your diabetes team
- What personal HbA1c target fits my age, medicines and hypoglycaemia risk?[4][6]
- Do I need a second laboratory test to confirm my result?[1][2]
- Could anaemia, iron levels, kidney function or a haemoglobin variant affect my HbA1c?[1][6][7]
- How often should I have my HbA1c checked over the next year?[4][6]
Frequently asked questions
Is 5.7% a normal HbA1c?
It depends on the country's cut-offs. In the US, 5.7% (39 mmol/mol) is the first value in the ADA prediabetes range, which runs to 6.4% (47 mmol/mol), so NIDDK does not call it normal. In the UK the NICE high-risk band starts higher, at 42 mmol/mol (6.0%), so 5.7% (39 mmol/mol) sits below it. A single result is not a diagnosis, so ask your GP or primary care team what your number means for you.[2][1][3]
Can one HbA1c result diagnose diabetes?
No. Without clear symptoms the ADA says a diagnosis needs two abnormal results, from different tests or from the same test at two time points. Your GP or primary care team will decide the next step.[1]
Do I need to fast before an HbA1c blood test?
No. NIDDK says you can eat and drink as usual before an HbA1c test because it measures a 2 to 3 month average rather than a single fasting morning reading. Ask your clinic whether any other blood test ordered for the same visit requires fasting.[2][7]
When can an HbA1c test give a misleading result?
Conditions that change red blood cell turnover, such as anaemia, kidney failure, recent blood transfusion, pregnancy and some haemoglobin variants, can make HbA1c less reliable. Tell your clinician if your finger-prick or sensor readings do not match your lab result.[6][7][10]
Where to go next
Your HbA1c result is a helpful 2 to 3 month guidepost when read alongside your daily readings and your care team's advice.
- Self-monitoring glucose, for what finger-prick checks show between lab tests.
- Technology and diabetes, for how sensors and GMI fit alongside HbA1c.
- Is Diabec legit?, for what is in our supplement and what it does not do.
Note: If you are pregnant, breastfeeding, or considering giving any supplement or herbal product to a child, consult your healthcare provider before making any changes to your diet, supplement routine, or prescribed medications.
Disclaimer: These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. Talk to your doctor before changing your diet, your activity or your medicines, and before starting any supplement.
Sources & References
- American Diabetes Association Professional Practice Committee for Diabetes (2026). 2. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes-2026. Diabetes Care. pubmed.ncbi.nlm.nih.gov/41358893 Funding: none listed on the PubMed record. COI: Duality of interest information for each contributor is available at https://doi.org/10.2337/dc26-SDIS.
- NIDDK (last reviewed April 2018). The A1C Test and Diabetes. Checked 25 October 2026. niddk.nih.gov Funding: not applicable, this is official or charity guidance, not a funded study.
- NICE (public health guideline PH38, last updated 15 September 2017). Type 2 diabetes: prevention in people at high risk. Checked 25 October 2026. nice.org.uk Funding: not applicable, this is official or charity guidance, not a funded study.
- NICE (guideline NG28, last updated 18 February 2026). Type 2 diabetes in adults: management, section 1.5 HbA1c measurement and targets. Checked 25 October 2026. nice.org.uk Funding: not applicable, this is official or charity guidance, not a funded study.
- Diabetes UK. What is HbA1c? Blood test and target levels. Checked 25 October 2026. diabetes.org.uk Funding: not applicable, this is official or charity guidance, not a funded study.
- American Diabetes Association Professional Practice Committee for Diabetes (2026). 6. Glycemic Goals, Hypoglycemia, and Hyperglycemic Crises: Standards of Care in Diabetes-2026. Diabetes Care. pubmed.ncbi.nlm.nih.gov/41358894 Funding: none listed on the PubMed record. COI: Duality of interest information for each contributor is available at https://doi.org/10.2337/dc26-SDIS.
- NIDDK (last reviewed July 2022). Diagnosis of Diabetes and Prediabetes. Checked 25 October 2026. niddk.nih.gov Funding: not applicable, this is official or charity guidance, not a funded study.
- Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ (2008). Translating the A1C assay into estimated average glucose values. Diabetes Care. pubmed.ncbi.nlm.nih.gov/18540046 Funding: American Diabetes Association, European Association for the Study of Diabetes, International Diabetes Federation, Abbott Diabetes Care, Bayer Healthcare, GlaxoSmithKline, Lifescan, Medtronic Minimed, Novartis, Novo Nordisk, Roche Diagnostics and Sanofi-Aventis. COI: study sponsors provided funding and blood glucose monitoring supplies as listed in the Diabetes Care report.
- Bergenstal RM, Beck RW, Close KL, Grunberger G, Sacks DB, Kowalski A, Brown AS, Heinemann L, Aleppo G, Ryan DB, Riddlesworth TD, Cefalu WT (2018). Glucose Management Indicator (GMI): A New Term for Estimating A1C From Continuous Glucose Monitoring. Diabetes Care. pubmed.ncbi.nlm.nih.gov/30224348 Funding: none listed on the PubMed record. COI: authors reported research support, consulting or advisory board ties with Abbott Diabetes Care, Dexcom, Medtronic, Novo Nordisk, Onduo, Roche, Sanofi and Eli Lilly.
- American Diabetes Association Professional Practice Committee for Diabetes (2026). 15. Management of Diabetes in Pregnancy: Standards of Care in Diabetes-2026. Diabetes Care. pubmed.ncbi.nlm.nih.gov/41358885 Funding: none listed on the PubMed record. COI: Duality of interest information for each contributor is available at https://doi.org/10.2337/dc26-SDIS.
- NHS (page last reviewed 3 August 2023). Low blood sugar (hypoglycaemia). Checked 25 October 2026. nhs.uk Funding: not applicable, this is official or charity guidance, not a funded study.
- NIDDK (last reviewed July 2021). Low Blood Glucose (Hypoglycemia). Checked 25 October 2026. niddk.nih.gov Funding: not applicable, this is official or charity guidance, not a funded study.
- NICE (guideline NG17, last updated 17 August 2022). Type 1 diabetes in adults: diagnosis and management, section 1.6 blood glucose management. Checked 25 October 2026. nice.org.uk Funding: not applicable, this is official or charity guidance, not a funded study.
- Hag Elagib MA, Ahmed RA, Eltom H, Abdelrahman MF, Mohammed MMEA, Ahmed A, Ahmed SAM (2026). Iron Deficiency, Iron Deficiency Anemia, and HbA1c Levels in Nondiabetic Adults: A Systematic Review and Meta-Analysis. Cureus. pubmed.ncbi.nlm.nih.gov/42798745 Funding: none listed on the PubMed record. COI: the authors declared no conflicts of interest.
- NICE (guideline NG3, last updated 16 December 2020). Diabetes in pregnancy: management from preconception to the postnatal period. Checked 25 October 2026. nice.org.uk Funding: not applicable, this is official or charity guidance, not a funded study.