Title tag (58 chars): A1C to average glucose: conversion chart and what it means
Reviewed by Mukul Mittal, MD, Medical Director at Ultrahuman
An A1C test (also written HbA1c) is one of the most common things your doctor will run on a blood panel. It’s a single number that estimates what your average blood sugar has been over the past two to three months. A high number can be the first sign of prediabetes or type 2 diabetes.
The number itself is a percentage, which is confusing for most people. An A1C of 6.5% doesn’t obviously mean anything on its own.
The conversion chart below translates every A1C reading into a real blood sugar number, which can be easier to understand.
This guide covers the chart, explains what each range means, flags where the reading can mislead, and lists the everyday changes required to move it.
The conversion chart
Every A1C percentage converts to an average blood sugar value using a formula built from a study of 507 people at 10 international research centres (Nathan DM et al., Diabetes Care 2008 ). The formula produces something called estimated average glucose (eAG) – the same average as A1C, just expressed in the units a home glucose meter uses.
eAG (mg/dL) = 28.7 × A1C − 46.7 eAG (mmol/L) = 1.59 × A1C − 2.59
The table below runs the maths for the values that matter most:
| A1C (%) | eAG (mg/dL) | eAG (mmol/L) | What it means |
|---|---|---|---|
| 5.0 | 97 | 5.4 | Normal |
| 5.5 | 111 | 6.2 | Normal |
| 5.7 | 117 | 6.5 | Prediabetes – lower edge |
| 6.0 | 126 | 7.0 | Prediabetes |
| 6.4 | 137 | 7.6 | Prediabetes – upper edge |
| 6.5 | 140 | 7.8 | Diabetes (diagnostic threshold) |
| 7.0 | 154 | 8.6 | Diabetes – typical treatment target |
| 7.5 | 169 | 9.4 | Diabetes – above target |
| 8.0 | 183 | 10.1 | Poor glucose control |
| 9.0 | 212 | 11.8 | Poor glucose control |
| 10.0 | 240 | 13.4 | Severe – needs clinical intervention |
What’s a healthy A1C?
The American Diabetes Association groups A1C readings into three bands (ADA Standards of Care 2025, Diabetes Care ):
- Normal: below 5.7%
- Prediabetes: 5.7% to 6.4%
- Diabetes: 6.5% or above
The prediabetes band is where most people learn something new. Roughly one in three US adults sits in this range, and most don’t know it (CDC National Diabetes Statistics Report). It’s the band where lifestyle changes have the biggest impact – the further along someone is toward the diabetes threshold, the harder it becomes to reverse.
For someone already diagnosed with diabetes, doctors typically aim for an A1C below 7.0%, with the exact target set individually. Younger, healthier adults might aim for 6.5%; older adults with other health conditions might aim for 8.0%. The tighter the target, the higher the risk of dropping too low from medication – so the goal balances long-term risk against short-term safety.
Where A1C can mislead
Because A1C measures glucose stuck to red blood cells rather than glucose itself, anything that changes how long red blood cells live can throw the reading off. A few common conditions do this without changing actual blood sugar at all.
Conditions that can make A1C look lower than it really is:
- Recent blood loss or haemolytic anaemia (fast red-blood-cell turnover means newer cells on average, with less glucose attached)
- Sickle cell trait or other haemoglobin variants like thalassaemia
- Chronic liver disease or advanced cirrhosis
- Pregnancy, especially in the second and third trimester
Conditions that can make A1C look higher than it really is:
- Iron-deficiency anaemia (older red blood cells accumulate more glucose)
- B12 or folate deficiency
- Advanced chronic kidney disease
- Splenectomy (spleen removal)
There’s also a population-level factor. Studies have found that the relationship between A1C and actual blood glucose varies across racial and ethnic groups, with implications for how accurate the test is as a screening tool (Herman WH, Cohen RM, Journal of Clinical Endocrinology & Metabolism 2012 ).
Researchers are still working out why – differences in red blood cell lifespan and glucose-binding chemistry are the likely factors. It’s one reason continuous glucose monitoring (CGM) is increasingly used alongside A1C rather than after it.
What A1C can’t show you
A1C gives an average. It doesn’t say anything about how steady or how erratic your blood sugar actually is. Consider two people, both with an A1C of 6.2%:
- Person A sits comfortably between 100 and 140 mg/dL most of the day
- Person B swings between 65 mg/dL (low) and 220 mg/dL (high), averaging out to the same number
Their A1C reads the same. Their actual experience of blood sugar – and the strain on their body – is very different. This is where continuous glucose monitoring fills the gap A1C leaves.
The most useful CGM metric here is Time-in-Range – the percentage of the day glucose spends between 70 and 180 mg/dL. Research has shown that every 10-percentage-point improvement in Time-in-Range corresponds to roughly a 0.6-point drop in A1C on average (Beck RW et al., Journal of Diabetes Science and Technology 2019 ).
A second CGM metric, the Glucose Management Indicator (GMI), estimates what a lab A1C would read based purely on CGM data (Bergenstal RM et al., Diabetes Care 2018 ). When GMI and lab A1C disagree by a lot, that gap itself is useful information – it can be the first sign one of the A1C-distorting conditions listed above is at play.
How to lower A1C in three months
Three months isn’t arbitrary – it’s the lifespan of the red blood cells the test is measuring. Changes made today will show up on the next A1C reading, though most of the drop tends to come from the final four to six weeks before the test rather than the first.
The interventions that reliably move A1C are the ones that flatten the blood sugar spikes that happen after meals. Those spikes drive most of the glucose that ends up stuck to haemoglobin.
- Change the order in which you eat. Eating fibre, protein and fat before the carb portion of a meal noticeably flattens the post-meal blood sugar rise (Shukla AP et al., Diabetes Care 2015 ). Same food, different order, better result.
- Walk for 10 to 15 minutes after meals. This is the single most effective post-meal intervention. Light walking moves glucose into muscle tissue before insulin has to work as hard – and walking specifically after meals lowers post-meal blood sugar more than the same total walking spread across the day (Reynolds AN et al., Diabetologia 2016 ).
- Add regular aerobic and strength training. Zone-2 cardio (the easy, conversational pace) makes cells more efficient at using glucose. Strength training makes muscles more responsive to insulin. Both help, and they compound.
- Prioritise sleep consistency. A single bad night raises fasting glucose the next morning. Chronic short sleep is linked to worse A1C even in otherwise healthy adults.
- Deal with chronic stress. Cortisol raises blood sugar. Persistent stress – poor work-life balance, ongoing illness, sleep debt – often shows up as a slowly creeping A1C without any change in diet at all.
For anyone tracking progress in real time, an Ultrahuman M2 Live CGM shows how these changes land within days, rather than waiting three months for the next lab draw. That makes it much easier to see which changes actually work for a specific person, rather than guessing.
For a fuller walkthrough of these lifestyle levers, see Ultrahuman’s guide on how to lower A1C naturally.
When to test A1C
The ADA recommends:
- Twice a year for someone with diabetes who is stable and meeting their targets
- Every three months if therapy has changed recently or targets aren’t being met
- Once a year for adults with prediabetes
- Every three years starting at age 35 for adults without diabetes, as a screening test – more often if the first result raises questions
Home A1C tests exist and have improved in accuracy, but a lab test remains the gold standard. When a home reading sits close to one of the clinical thresholds, it’s worth confirming with a lab test before making decisions about lifestyle or medication. CGM data, where available, is the most actionable thing to add alongside – it shows not just how high the average is, but exactly when and why it climbs.
Frequently asked questions
What is a normal A1C level? Below 5.7% for adults without diabetes. That corresponds to an average blood sugar under 117 mg/dL.
What does an A1C of 5.7 mean? 5.7% is the lower edge of the prediabetes range. The estimated average blood sugar is around 117 mg/dL. It’s a signal worth acting on – sustained lifestyle changes bring many people in this band back to a normal A1C.
Can A1C be wrong? It can be misleading in people with anaemia, sickle cell or thalassaemia, advanced kidney disease, recent blood loss, or during pregnancy. In those situations, CGM-based metrics (Time-in-Range, GMI) or a fasting glucose test are usually more reliable.
How long does it take to lower A1C? Around 8 to 12 weeks, matching the lifespan of the red blood cells being measured. Most of the visible improvement tends to come from the final four to six weeks before the test, because newer red blood cells are weighted more heavily.
Is A1C the same as average glucose? Not quite. A1C is a percentage – the share of haemoglobin with glucose attached. Estimated average glucose (eAG) is the same information in mg/dL. The conversion formula is eAG = 28.7 × A1C − 46.7.
Should I use a CGM if my A1C is normal? CGM data can reveal patterns a single A1C misses – post-meal spikes, night-time swings, how specific foods affect an individual response. Many people with a normal A1C still find the variability data useful.
Can A1C be too low? Yes. A very low A1C in someone without diabetes can point to a red blood cell turnover problem (rapid breakdown, chronic blood loss) rather than excellent glucose control. When A1C drops sharply and unexpectedly, that’s worth a conversation with a doctor.
This article is for informational purposes and isn’t medical advice. People with diabetes or suspected metabolic conditions should work with a clinician for diagnosis, medication and treatment decisions. Disclosure: Ultrahuman sells the M2 Live CGM, Ring AIR and Ring PRO referenced in this guide.
THE AUTHOR
