HbA1c Converter

The two scales, and why they exist

NGSP / DCCT, reported as a percentage. Used in the United States and in Peru. This is the scale most people recognise: 5.6%, 7.2%, and so on.

IFCC, reported in mmol/mol. Used across most of Europe, the UK, Australia and much of Asia. The same result reads as 38 mmol/mol or 55 mmol/mol.

Neither is more accurate. The IFCC scale is more precisely standardised; the NGSP scale is the one all the landmark clinical trials were published in, which is why the thresholds everyone quotes are percentages.

If you have moved between countries, this is worth understanding properly — otherwise a perfectly stable HbA1c can look like a dramatic change simply because the units switched.

The diagnostic thresholds

HbA1c (%)	HbA1c (mmol/mol)	Interpretation
Below 5.7	Below 39	Normal
5.7 – 6.4	39 – 46	Prediabetes
6.5 or above	48 or above	Diabetes range

A result in the diabetes range should be confirmed with a second test before the diagnosis is made — a single value is not enough.

Estimated average glucose

The converter also shows estimated average glucose, in mg/dL (used in Peru and the US) and mmol/L (used in the UK and much of Europe). To convert between those two directly, divide mg/dL by 18.

Treat the average as exactly that: an average. Two people with an identical HbA1c of 7.0% can have very different daily patterns, one steady and one swinging between highs and lows. That variability matters clinically, and HbA1c alone cannot show it — which is what continuous glucose monitoring adds.

When HbA1c gives a misleading answer

HbA1c measures glycated haemoglobin, so anything that changes the lifespan of your red blood cells distorts it.

It reads falsely low in haemolytic anaemia, after recent blood loss or transfusion, in advanced liver disease, and in pregnancy.

It reads falsely high in iron deficiency anaemia, in vitamin B12 deficiency, and in kidney failure.

It can be unreliable altogether with haemoglobin variants such as HbS or HbC traits, depending on the assay the laboratory uses.

If your HbA1c does not match your glucose readings, one of these is usually the reason, and it is worth investigating rather than assuming the meter is wrong.

What to do with the number

If you are in the prediabetes range, that is genuinely good news arriving early: progression to type 2 diabetes is preventable in a substantial proportion of cases, and the interventions are most effective at exactly this stage.

If you are in the diabetes range, the goal is not a specific universal number. Targets are individualised by age, duration, complications and hypoglycaemia risk — a target of under 7% suits many people, while others are better served by something tighter or looser.