high HbA1c · 9 min read
HbA1c: what it is, what counts as normal, and how to lower it
If you have just been handed a blood test with an HbA1c of 6.8 and you went home with the word "diabetes" spinning in your head, this article is for you. I am going to explain exactly what that number means, without drama and without frightening you.
What HbA1c actually is
Haemoglobin is the protein your red blood cells carry to transport oxygen. When there is glucose circulating in your blood, some of it sticks to that haemoglobin. And it does not come off.
Here is the key: your red blood cells live for about 120 days. So if we measure what percentage of your haemoglobin has glucose stuck to it, we get a summary of how your glucose has been over the last two to three months. Not this morning: this quarter.
The idea in one sentence: fasting glucose is a photograph of a single moment. HbA1c is the three-month film.
That is why it is so useful, and why it is so hard to fool. You can behave impeccably for the three days before the test and your fasting glucose will come out decent. HbA1c will not play along: it also reflects the dinners of the two months before.
One nuance worth knowing: not every day counts equally. The most recent month weighs more than the two before it, because the older red cells are disappearing. That is good news, actually — it means whatever you do from today starts counting sooner than you think.
What counts as normal
| Category | HbA1c (NGSP) | HbA1c (IFCC) |
|---|---|---|
| Normal | Below 5.7% | Below 39 mmol/mol |
| Increased risk | 5.7 – 6.4% | 39 – 47 mmol/mol |
| Diabetes | 6.5% or above | 48 mmol/mol or above |
The two units that confuse everyone
Notice there are two columns. Laboratories report HbA1c in two different units, and they often print both on the same sheet:
- NGSP (or DCCT): the percentage. The number we are used to: 6.5%, 7.2%…
- IFCC: millimoles per mole. The international unit, which produces much bigger numbers: 48, 55, 64…
This causes genuine alarm. Somebody sees a 48 on their report, mentally compares it with the 6.5 they read online, and concludes they have a value ten times worse than they do. 48 mmol/mol and 6.5% are the same value in two units.
Quick conversion: mmol/mol = (% − 2.15) × 10.929. And back again: % = mmol/mol ÷ 10.929 + 2.15. If you would rather not do the sums, the calculator gives you both.
And now the full table: every test
HbA1c is not the only way to diagnose. These are all of them, and it is worth knowing them because you will not always be given the same one:
| Test | Normal | Increased risk | Diabetes |
|---|---|---|---|
| Fasting glucose | <100 mg/dL | 100 – 125 mg/dL | ≥126 mg/dL |
| HbA1c (NGSP) | <5.7% | 5.7 – 6.4% | ≥6.5% |
| HbA1c (IFCC) | <39 mmol/mol | 39 – 47 mmol/mol | ≥48 mmol/mol |
| 75 g oral glucose tolerance test, at 2 h | <140 mg/dL | 140 – 199 mg/dL | ≥200 mg/dL |
| Random glucose + typical symptoms | — | — | ≥200 mg/dL |
These are the American Diabetes Association criteria. In Spain many doctors use the WHO and Spanish Diabetes Society criteria, which are stricter about what counts as prediabetes: fasting glucose of 110 to 125 mg/dL and HbA1c of 6.0 to 6.4%. If your report uses different ones, your doctor's criteria are the ones that count.
Two points that get missed and matter:
A single abnormal result diagnoses nothing. Diagnosis requires confirmation, usually a second measurement on a different day. The exception is a random glucose of 200 or more accompanied by typical symptoms — heavy thirst, passing a lot of urine, unintended weight loss — which is enough on its own.
The tests may disagree with each other. It is common to have a normal fasting glucose and an HbA1c of 5.9, or the other way round. That is not a lab error: they measure different things. Your doctor interprets them together — not a website.
Translate your HbA1c into average glucose
The percentage does not say much on its own. What genuinely helps is knowing what average glucose it corresponds to, because that number you do understand: it is the one on your meter.
There is an internationally validated formula for that translation, published in the ADAG study in Diabetes Care and endorsed by the American Diabetes Association:
Average glucose (mg/dL) = 28.7 × HbA1c − 46.7
| HbA1c | Approximate average glucose | What it means |
|---|---|---|
| 5.5% | 111 mg/dL | Normal |
| 5.7% | 117 mg/dL | Threshold of increased risk |
| 6.0% | 126 mg/dL | Increased risk |
| 6.5% | 140 mg/dL | Diabetes threshold |
| 7.0% | 154 mg/dL | Usual target in type 2 diabetes |
| 8.0% | 183 mg/dL | Insufficient control |
| 9.0% | 212 mg/dL | Clearly insufficient control |
| 10.0% | 240 mg/dL | High risk of complications |
Notice something important: going from 8.0 to 7.0 is not "dropping a little point". It is dropping your average glucose from 183 to 154 mg/dL, nearly 30 mg/dL less every hour of every day. That is the real size of the change.
You can work out yours here: HbA1c calculator.
Is this happening to you?
If you have been trying for a while and the number will not move, tell me. The first conversation is free and carries no commitment.
First call free · No commitment
How much it can drop, and how long it takes
The question everybody asks. With published data, not promises.
First, time: because HbA1c reflects three months, there is no point repeating it sooner than three months. If you test four weeks after changing your habits, the result still carries the previous weeks and you will get a disappointment you did not earn. The first blood test that can prove you right comes around the third month.
Second, how much. It depends on three things: where you start from, how much weight you lose, and what kind of exercise you do.
- Exercise alone, diet unchanged. In people with type 2 diabetes, programmes combining aerobic and resistance training lower HbA1c by around 0.5 points on average. Not spectacular, but real and with very few risks. If you use insulin or sulfonylureas, check with your doctor first: exercise can drop your glucose too far.
- With significant weight loss. Here the numbers change scale. In the DiRECT trial the relationship between kilos lost and diabetes remission was direct: 7% remission among those who lost less than 5 kg, 34% between 5 and 10 kg, 57% between 10 and 15 kg, and 86% among those who lost 15 kg or more.
- The higher you start, the more room you have. Going from 9 to 7.5 is usually faster than going from 7 to 6.5. The last few tenths always cost more.
What actually lowers it
In order of impact, based on the evidence and on what I see in practice:
1. Losing fat, especially around the abdomen
By far the most powerful lever. Not for looks: fat stored in the liver and around the pancreas is what directly interferes with how you handle glucose. I explain it in detail in belly fat and type 2 diabetes.
2. Changing the shape of your carbs before cutting them
The instinctive reaction is to drop bread and pasta. It usually works for a few weeks and is then abandoned, because it is unsustainable. Change first which carbohydrate, what you eat it with, and in what order: vegetables and protein before the carb, and the carb whole and unblended. The effect on post-meal glucose is notable and the cost of doing it is far lower.
3. Resistance training, not just walking
Walking is good and you should do it. But muscle is your body's biggest consumer of glucose — it clears around 80% of the sugar insulin moves — and to have more muscle you have to work it against resistance.
Here is the long-term part almost nobody tells you: muscle is lost over the years unless you give it a reason to stay, and walking barely slows that loss. The walk lowers this month's number; the muscle you keep is what holds it down five years from now. Guidelines recommend strength training two or three days a week, on non-consecutive days, alongside aerobic activity. I develop this in exercise and blood sugar.
4. Breaking up sitting, even when it is not "exercise"
Guidelines recommend interrupting long sitting periods every 30 minutes. Standing up for two minutes every half hour appears in no training plan and still has a measurable effect on glucose.
5. Sleep and stress
Short sleep and sustained stress raise cortisol, and cortisol raises glucose. It is the part every plan ignores, and the one that most often explains why somebody "doing everything right" does not improve.
When HbA1c misleads you
You will not find this section in most blogs, and it matters. HbA1c measures haemoglobin, so anything that alters your red blood cells alters the result.
It can read falsely low if your red cells live less than normal: anaemia from blood loss or from red cell destruction, recent transfusions, pregnancy, some spleen conditions. It can read falsely high in iron or vitamin B12 deficiency, and in some cases of kidney failure. And certain haemoglobin variants — common in people with African, Mediterranean or South-East Asian family origins — interfere directly with the measurement.
If your HbA1c does not match what you see on your meter day after day, tell your doctor. There are alternatives for assessing control, such as fructosamine or continuous glucose monitoring. Do not accept a figure that consistently contradicts your experience.
What to do now
- Write down your value and the date. Without a starting point there is no way to know whether you are improving.
- Translate it into average glucose to understand what we are talking about.
- Set a three-month target, not a three-week one.
- Pick one single lever and hold it. A new habit takes about two months on average to become automatic (18 days for some, over 250 for others), and missing a day does not undo it.
- Repeat the test after three months, not before.
And if you have been trying for a while and the number will not move, the problem is probably not your willpower — it is the plan. That is where I can help.
Pablo Berna Sierra · Nurse, dietitian and trainer
Nursing registration no. 15794 · Colegio Oficial de Enfermería de Zaragoza
Over 14 years practising as a nurse and over a decade working with people with type 2 diabetes, prediabetes and excess weight. More about me.
Sources
10
- Nathan DM et al. “Translating the A1C Assay Into Estimated Average Glucose Values”. Diabetes Care, 2008. View source
- NGSP. “HbA1c and Estimated Average Glucose (eAG)”. View source
- NGSP. “IFCC and NGSP unit conversion”. View source
- American Diabetes Association. “Diagnosis and Classification of Diabetes: Standards of Care in Diabetes”. View source
- Lean MEJ et al. “Primary care-led weight management for remission of type 2 diabetes (DiRECT)”. The Lancet, 2018. View source
- Umpierre D et al. “Physical activity advice only or structured exercise training and association with HbA1c levels in type 2 diabetes”. JAMA, 2011. View source
- Lally P et al. “How are habits formed: Modelling habit formation in the real world”. European Journal of Social Psychology, 2010. View source
- DeFronzo RA, Tripathy D. “Skeletal muscle insulin resistance is the primary defect in type 2 diabetes”. Diabetes Care, 2009. View source
- Colberg SR et al. “Physical Activity/Exercise and Diabetes: A Position Statement of the American Diabetes Association”. Diabetes Care, 2016. View source
- American Diabetes Association. “Facilitating Positive Health Behaviors and Well-being to Improve Health Outcomes: Standards of Care in Diabetes—2025”. Diabetes Care, 2025. View source
Please note. This article is for general guidance and does not replace consultation with your doctor or the treatment prescribed to you. Do not change your medication on your own.