Why does my meter average not match my A1C?

Two reasons. The conversion is a population regression with roughly forty mg/dL of scatter around it, and your meter average only covers the moments you tested, which are rarely the moments that matter most.

What is the mmol/mol figure?

The IFCC unit, used across most of the world. It measures the same thing on a different scale, and both are reported here because international results and guidelines mix them.

Health & Fitnessclinical

A1C to Blood Glucose Converter

A1C to estimated average glucose and back, plus the IFCC unit used outside the United States. The relationship is a regression with real scatter, so two people with the same A1C can genuinely differ by twenty or thirty mg/dL in mean glucose.

Also called: a1c to average glucose, hba1c converter.

Estimated average glucose

140

An A1C of 6.5% corresponds to an estimated average glucose of 140 mg/dL, or 7.8 mmol/L. That is at or above 6.5, the diagnostic threshold for diabetes. These are diagnostic thresholds applied with other findings and repeat testing, not a diagnosis from one number. A1C measures glycation over the life of a red cell, so anaemia, haemoglobin variants, recent transfusion, pregnancy and kidney disease change the reading without any change in glucose. The published scatter around this conversion is roughly forty mg/dL.

A1C, NGSP percent

6.5

A1C, IFCC mmol/mol

Estimated average glucose, mmol/L

7.8

Reference band

at or above 6.5, the diagnostic threshold for diabetes

On the band

That is at or above 6.5, the diagnostic threshold for diabetes. These are diagnostic thresholds applied with other findings and repeat testing, not a diagnosis from one number.

When A1C misleads

A1C measures glycation over the life of a red cell, so anaemia, haemoglobin variants, recent transfusion, pregnancy and kidney disease change the reading without any change in glucose. The published scatter around this conversion is roughly forty mg/dL.

How this is calculated

For reference only. Verify against your clinical protocol before acting on this number.

A1C against average glucose

A1C %	IFCC mmol/mol	eAG mg/dL	eAG mmol/L
5	31	97	5.4
5.7	39	117	6.5
6	42	126	7
6.5	48	140	7.8
7	53	154	8.6
7.5	58	169	9.4
8	64	183	10.2
9	75	212	11.7
10	86	240	13.3

Method and background

This is what the calculation gives for the numbers you entered. It is an estimate, not advice, and it knows nothing about your situation beyond those numbers. Rules for United States change on a published schedule; the effective date is shown on every rule-based tool.

How this is calculated

The conversion comes from the ADAG study, which paired A1C with continuous glucose monitoring in around five hundred participants and fitted a straight line. The line is a good average and a poor individual predictor: the published intervals around each A1C span roughly forty mg/dL. A1C also measures haemoglobin glycation over the lifespan of a red cell, so anything that changes that lifespan changes the reading without any change in glucose, which is why anaemia, haemoglobin variants, recent transfusion and pregnancy all break it.

the ADAG regression relating A1C to mean glucose, fitted on around five hundred participants

eAG: Estimated average glucose, mg/dL
A1C: Glycated haemoglobin, NGSP percent

Worked examples

Each of these is asserted on every build. If a change to the engine ever moved one of these answers, the build would fail before the page could print it.

an A1C of 6.5

What you have: An A1C percentage
A1C: 6.5 %
Average glucose: 0
Glucose unit: mg/dL

Estimated average glucose140

28.7 x 6.5 - 46.7; the IFCC figure is the familiar 48 mmol/mol

Open this example

converting back from a glucose average

What you have: An average glucose
A1C: 0 %
Average glucose: 140
Glucose unit: mg/dL

Estimated average glucose140

boundary: the inverse of the same regression must round-trip

Open this example

Method and limits

What it assumes

Normal red cell turnover, which is what makes A1C a glucose measure at all.

What it deliberately does not model

The A1C to glucose relationship has wide individual scatter and the estimate is not a substitute for monitoring.
Anaemia, haemoglobin variants, recent blood loss, transfusion, pregnancy and chronic kidney disease all distort A1C.
Reference bands are diagnostic thresholds used with other findings, not a diagnosis on their own.

Formula version 1.0.0 · definition 1.0.0 · United States · Report a problem with this calculator

Frequently asked questions

Why does my meter average not match my A1C?: Two reasons. The conversion is a population regression with roughly forty mg/dL of scatter around it, and your meter average only covers the moments you tested, which are rarely the moments that matter most.
What is the mmol/mol figure?: The IFCC unit, used across most of the world. It measures the same thing on a different scale, and both are reported here because international results and guidelines mix them.