A1c to Average Glucose Converter

What A1c actually measures

HbA1c measures the share of your haemoglobin that has glucose attached to it. Red blood cells live around three months, so the figure reflects roughly that window — but it is weighted toward the most recent few weeks, which is why a good last month can pull the number down more than you would expect.

Percent or mmol/mol — which does your clinic use?

Both describe the same measurement. The percentage scale (NGSP/DCCT) is standard in the United States. The mmol/mol scale (IFCC) is standard across most of Europe, including Sweden and the UK. A Swedish lab reporting 53 mmol/mol and an American lab reporting 7.0% are saying the same thing. The conversion is IFCC = (NGSP − 2.15) × 10.929, which this page applies in both directions.

Where estimated average glucose comes from

The eAG formula comes from the ADAG study, which compared A1c against continuous glucose data in several hundred people: eAG in mg/dL = 28.7 × A1c% − 46.7. Divide by 18.0182 for mmol/L. It is a population average, not a personal one — two people with identical A1c can have genuinely different mean glucose.

Why your CGM average may disagree

A mismatch between your CGM average and your lab A1c is common and usually not an error. The windows differ: A1c covers about three months, a CGM average covers whatever period you selected. Red blood cell lifespan varies between people. Anaemia, recent blood loss, pregnancy, kidney disease and some haemoglobin variants all move A1c away from your true mean glucose. If the gap is large and persistent, that is worth raising with your care team rather than explaining away.

A1c is not the whole picture

Two people can share an A1c of 7.0% while one runs steadily near target and the other swings between hypos and highs that average out. That is why time in range has become the more useful day-to-day measure. A1c remains the standard for long-term risk, but it says nothing about variability — and variability is what most people actually feel.

mmol/mol is not mmol/L — the most common mix-up

These two look almost identical and mean completely different things. mmol/mol is HbA1c: a three-month average, typically somewhere between 30 and 100. mmol/L is a glucose reading: what your meter or CGM shows right now, typically between 4 and 15. A value of 53 mmol/mol is a perfectly ordinary A1c. A glucose of 53 mmol/L would not be survivable. If a number seems alarming, check the unit before anything else — this page keeps the two apart deliberately, and shows glucose in both mmol/L and mg/dL so there is no ambiguity.

What the published targets actually say

The widely cited general target for non-pregnant adults is an A1c below 7% (53 mmol/mol), from the American Diabetes Association and echoed in European guidance. The diagnostic thresholds are separate: 6.5% (48 mmol/mol) and above is the cut-off used for diagnosing diabetes, and 6.0-6.4% (42-47 mmol/mol) is the range Swedish practice treats as raised but below the diagnostic line. Every one of these is a population figure. Published guidance is explicit that targets should be individualised — looser for people with a history of severe hypoglycaemia, limited life expectancy or significant comorbidity, tighter for others. Your own target is a conversation with your care team, not something a converter can tell you.

Why it is measured roughly every three months

The three-month interval is not arbitrary. It matches the lifespan of a red blood cell, so each test covers a genuinely new population of cells rather than re-measuring the same ones. Testing much more often tends to produce noise rather than signal: a change driven by the past three weeks is still competing with the two months behind it. Most clinics settle on three to six months depending on how stable things are.

How much change actually counts as change

Laboratory assays certified to NGSP standards are precise, but not infinite. Between analytical variation and ordinary biological variation, differences smaller than roughly 0.3 to 0.5 percentage points (about 3 to 5 mmol/mol) can sit inside the noise. A move from 7.4% to 7.2% may be real, or may be the same underlying value measured twice. A move from 7.4% to 6.6% is almost certainly real. Direction across three or four consecutive tests tells you far more than the gap between any two.

Things that shift A1c without your glucose changing

A1c is a measurement of haemoglobin, so anything that changes your red blood cells changes the result. Iron-deficiency anaemia tends to push it up. Recent blood loss, transfusion, haemolysis, pregnancy and chronic kidney disease tend to push it down or distort it. Haemoglobin variants — including HbS, HbC and HbE traits, which are common in parts of Africa, the Mediterranean and Southeast Asia — can interfere with some assays outright. Red cell lifespan also varies naturally between individuals, which is why two people with the same true average glucose can return A1c results a few tenths apart. None of this makes the test unreliable; it makes context necessary.