• Guest - w'd love to know what you think about the forum! Take the 2025 Survey »

Is GMI more useful than HBA1C

Art Of Flowers

Well-Known Member
Messages
1,299
Location
Kent
Type of diabetes
I reversed my Type 2
Treatment type
Diet only
Dislikes
Statins
HbA1c is short for glycated haemoglobin (glycated means joined with glucose). The test is also sometimes called haemoglobin A1c. Haemoglobin (Hb) is the protein in red blood cells that carries oxygen through your body. Red blood cells live for about three month, so the HBA1C is a measure of your blood sugar for the previous 3 months, weighted towards the recent time. HbA1C ranges are described here ... https://www.diabetes.co.uk/what-is-hba1c.htmlz

HBA1C Ranges

* Normal - Below 42 mmol/mol or 6.0%
* Prediabetes - 42 to 47 mmol/mol or 6.0% to 6.4%
* Diabetes - Above 48 mmol/mol or 6.5%

CGM Data

GMI - Glucose Management Indicator approximates the laboratory A1C level expected based on average glucose measured using continuous glucose monitoring (CGM) values. It is usually displayed as a percentage e.g. 6.5% corresponds to a HBA1C or 48.

You get a calculated GMI with 14 or more days data from a CGM such a Freestyle Libre or Dexcom.

In addition to a GMI value on a CGM you also get an Average Glucose value, a Time In Range (TIR) which is the percentage of time between a lower and upper limit (defaults to 3.9 and 10.0) which has a target of 70% or better. A Coefficient of Variation which indicates the variability of the glucose readings (target under 36%)

Is GMI more useful than the HBA1C

A HBA1C value of 7% could result from a person who has a roller-coaster of high and low blood glucose values or a relatively flat range of blood glucose numbers. This was discussed in the following article - From A1c to Time in Range — New Glycemic Management Metrics
 
Last edited:
To answer your question - I would say no. The HbA1c is the agreed international standard diagnostic test since around 2010: the test is somewhat more accurate than a CGM; and the GMI approximates to the HbA1c rather than the other way round.

I've found CGM use invaluable in understanding my BG fluctuations and I'd certainly recommend using one even for a short period. But while it was internally consistent, the CGM (Libre) consistently underestimated my BG as determined by fingerprick.

As with many things a CGM-derived GMI might be quite useful in keeping track between HbA1c tests, but don't expect your medical records to be using it.
 
Have to agree with @KennyA the discrepancies between actual and CGM readings have to have a effect on HB1AC , I am a fan of the Libre system, I know there are discrepancies as I mentioned but as a system to track blood sugars it is invaluable, plus as an aid to stop hypos its really good for me, it has helped me get my HB1AC down from high 90s + to 56-58 which is totally acceptable for me ,
 
Just having a good HBA1C figure does not really mean you are unlikely to get diabetic complications. Time In Range (TIR) is also a good indicator of diabetic complications risk ... see https://diatribe.org/diabetes-management/time-range

A study of 141 patients with Pre-Diabetes found that 8 had retinopathy ... see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10089014/
A lot of people get glucose spikes when they eat high carb food. This can cause diabetic complications even though the person has normal fasting blood glucose and a good HBA1C.

There have been studies regarding Time In Range and the likelihood of retinopathy ... see https://www.healio.com/news/endocri...-diabetic-retinopathy-risk-in-type-1-diabetes which suggest a 5% drop in TIR is associated with a 27% increase in risk of retinopathy.
 
I also agree with @KennyA.
My reason is that HbA1C is cheaper.
GMI requires a significant amount of CGM data.
Otherwise you could get someone “cheating the system” (and themselves) by eating very low carb for only the 2 weeks of their sensor which would give a false sense of security until the next test in 12 (or, if you are lucky 6) months.
 
Points;
1. & 2. Actually, @KennyA is an example of someone who developed diabetic neuropathy while still having a 'pre-diabetic' HbA1C, so his doctor said it could not be due to diabetes. However, I'm not sure if it could have been due to brief highs (TIR), people get neuropathy (sometimes) even when not diabetic - so even normal HbA1C does not guarantee no neuropathy will develop.

3. This study was of Type 1's so the standard range is 4 to 10. For us T2's (excluding those on Insulin) we tend to aim for the highs (at 2hrs post prandial) to be less than 8, so there is a difference there. Also to note that so far as I'm aware, time below the low of the range is not a contributor to neuropathy.
 
Hi
Statistics can prove anything as the saying goes but you do have to have a recognized benchmark.
Before the libre my levels were good but since the libre my control is better and my last hba1c was 40.
My current TIR is 86% but i class that as 93% because below 3.9 isn't dangerous to me and i have a 7% below 3.9.

And there you have it i have manipulated my own figures to suit my own viewpoint.lol

Interestingly enough i am due a blood test tomorrow and my libres estimated hba1c is 5.5% or 37mmol/mol over the 90 day period.

I am almost excited to see the outcome and my own prediction is another 40.

I get a bit twitchy as there are people out there struggling so i don't want anyone to think i'm being smug.
Ive had it for 50 years and have that experience to draw upon and remember the dark days of clinitest tablets which makes me embrace the new tech available.

i wish you all well

Tony
 
They're both useful in different ways.

You can have a 100% TIR with the officially adviced range by runninh between 8 and 10 all the time. This would give you an hba1c of about 56 mmol/mol, well into the diabetes range. TIR only wouldn't flag this.
On the other hand, you could have a non diabetic/prediabetic hba1c caused by many lows and highs. The hba1c wouldn't flag this.

Both measurements are complementary.
 
Last edited:
You must log in or register to reply here.
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn More.…