Hba1c estimate

S

Sharron1

Well-Known Member
Relationship to Diabetes
Type 2
When it gets close to my annual Hba1c I use different testing methods to give me some idea what the result might be. I am aware Hba1c is the gold standard and my chosen guesstimators won't be the exact result. I use accu check and check the 90.day average and I also use mysugr app.
Accucheck.has 6.1 average and mysugr gave me 5.8 (34.7).

Probably not the best use of my time but... which is the more reliable one?
 
Sharron1 said:
Probably not the best use of my time but... which is the more reliable one?
Click to expand...
The best one is the blood test.
I always think of these estimates as another good way to track our progress but close to zero value for finding out the real number.
One of the problems with using finger pricks for HBA1c estimation sis that you have no idea what is happening between finger pricks. But even with CGM like Libre which gives you a reading every (I think) 5 minutes, it can get the estimated HBA1c wrong (it was over 20% lower than the blood test for me).

By all means continue to track your progress with Accucheck and/or mySugr - it is a great way to calculate the affect of your work.

Good luck with your upcoming test.
 
There are formulae that can be used to estimate HbA1c from average glucose levels.

It’s important to recognise that HbA1c % values are not the same as an average of fingerstick readings (even though the numbers might look like similar values). HbA1c and fingersticks measure different things, and no direct conversion is possible.

I think @Docb has been conducting some interesting experiments attempting to predict HbA1c values from his 90 day average fingerstick readings (with some success I think) but I’m not sure which conversion formula he uses. 🙂

Hope you are pleased with your HbA1c result when it comes in 🙂
 
How long* before blood HbA1c gets superseded by "Time within Limits" for interstitial fluid, averaged over 90 days by some appropriate statistical method, as a measure of long-term glucose?
Does this require more accurate CGM sensors?

* probably the usual "in ten years' time"!
 
John Gray said:
How long* before blood HbA1c gets superseded by "Time within Limits" for interstitial fluid, averaged over 90 days by some appropriate statistical method, as a measure of long-term glucose?
Does this require more accurate CGM sensors?

* probably the usual "in ten years' time"!
Click to expand...
Surely, it first requires funded CGM for everyone.
For people with type 2 who are not on insulin, many struggle to get test strips funded. One blood test every 6 months is a lot cheaper and, for many (albeit a smaller "many" that the ones who struggle to get test strip funding) it is enough.

I can see TIR replacing HBA1c for people with Type 1 who do, in genera, have funded CGM. But it is important to remember that CGMs do not suit everyone.
 
everydayupsanddowns said:
There are formulae that can be used to estimate HbA1c from average glucose levels.

It’s important to recognise that HbA1c % values are not the same as an average of fingerstick readings (even though the numbers might look like similar values). HbA1c and fingersticks measure different things, and no direct conversion is possible.

I think @Docb has been conducting some interesting experiments attempting to predict HbA1c values from his 90 day average fingerstick readings (with some success I think) but I’m not sure which conversion formula he uses. 🙂

Hope you are pleased with your HbA1c result when it comes in 🙂
Click to expand...
Yes I reckon I can get within 2 units of the result. Last one I predicted a HbA1c of 50 and the result came back at 49.

The way I do it is to look at my past data, which I have been collecting for several years now, and whenever I have had an Hba1c result I have calculated the average finger prick result and average waking result for the previous 90 days.

1718614235896.png
The graph shows that as the 90 day overall average increases, my Hba1c increases linearly. This allows me to fit a line to the spots and then use the equation for that line to predict my HbA1c for any 90 day average.

So for me, a 90 day average of 8 will give an HbA1c close to 50. An average of 6.5 will give an HbA1c of 42. In reality i have all the information in a database and have set up routines so that I can click on a button to display graphs like the one above and every time I enter data my current predicted HbA1c is recalculated and displayed on the screen. It is currently predicting an HbA1c of 47.

I have to stress that is for me and my diabetes and the way I test (lots of random tests through the day) but I suspect that many people would get a similar result if they were nerdy enough to record all their data and to do the analysis.
 
I check my BG fairly regularly before my evening meal, especially if I’m eating something I’ve never checked before. My readings are usually around 6, give or take, before food and around 7.5, give or take, 2 hours later. Like a lot of folk, if I’m naughty, I have the occasional spike, but I’m always slightly surprised and pleased that my annual hba1c is usually between 44 and 48 I can only assume that, at periods in the day, my BG must get quite low to average out in the 40’s.
 
Last edited:
helli said:
I know it is "capable" of that but I don't know if it uses the 60 second readings in the estimated hba1c calculation.
Is this documented anywhere?
Click to expand...
I was hoping you’d tell me. It’s logical to assume it calculates from every shred of data it receives from interstitial fluid interpretations. My lab results come back consistently lower than what the app says. My focus is on TIR. But I did find this. Which doesn’t say a lot.

www.freestyle.abbott

What is eA1c?

Estimated A1c level (also called HbA1c) is based on available Sensor glucose data from the last 90 days. The more data available, the better your estimation will be. However, the estimated level may not match your A1c measured in a laboratory.* A1c can be used as an indicator of how well your...
www.freestyle.abbott www.freestyle.abbott
 
Many thanks to everyone for their assistance. I return to my ever present thinking, absolutely nothing remotely connected to Diabetes is ever simple Sigh.
 
helli said:
I know it is "capable" of that but I don't know if it uses the 60 second readings in the estimated hba1c calculation.
Is this documented anywhere?
Click to expand...

I’d have to look deeper back into notes from previous meetings and events I have attended, but my strong suspicion (half memory) is that only the 5min data points go towards the app / reader’s statistical analysis.
 
I know this is an old thread - but I wanted to add my thoughts anyway. I’ve been using the Freestyle Libre 2 sensors for about six months now. The app estimated the hba1c at 49, which turned out to be exactly the reading I got with the actual blood test recently. I don’t know exactly how the app works it out, but it’s pretty impressive.
 
I have been using Libre for a few years and mine is never the same as Blood Test.
 
I thought that I'd try to reduce my HbA1c by reducing the maximum daily intake of carbs. My HbA1c was 42, eating no more than 50gm of carbs a day. Right - I reduced to no more than 40gm per day - surely that would reduce me into the 30s.
A year later I got the result. 42.
 
Personally I find TIR a better variable to consider - you could have a low HbA1c which would indicate you are doing really well, but you may have 2-3 hypos a day (which is not good!), meaning your glucose variability is too high (a rollercoaster which is not good for overall control) - saying that my Libre estimation is not too far off when I have an actual blood test (within 5 everytime)
 
mashedupmatt said:
Personally I find TIR a better variable to consider - you could have a low HbA1c which would indicate you are doing really well, but you may have 2-3 hypos a day (which is not good!), meaning your glucose variability is too high (a rollercoaster which is not good for overall control) - saying that my Libre estimation is not too far off when I have an actual blood test (within 5 everytime)
Click to expand...
Not many T2s have CGMs, most would have to self-fund, so TIR is not available to all.
 
You must log in or register to reply here.
Back
Top