HbA1c compared to Libre estimation

It can be dodgy! I had an hba1c last week it was 45. My libre estimated 31. This is because it is notoriously inaccurate under 4. It thinks i am hypoing through the night. This morning it thought I was 2.7 but fingerprick was 4.2.

My Libre HbA1c estimate, even calibrated using prick tests and the miaomiao is usually 4-5 mmol/mol lower than the lab test.

I tend to get a few compression lows during the night, which may well be affecting the overall total.

But then, to be honest, I am much more interested in the standard deviation than in the average number, whether high or low or middling.

I have been using the libre whilst I am on steroids and insulin. Before that, I just used diet and fingerprick. I recorded my numbers in the mysugr app and the difference between the estimated hba1c and the lab test was much closer.

I've seen alot of anecdotes on this site, which seem to point towards inaccurate HbA1c's for Libre's. I agree with @Brunneria, that variance is key with any of the devices.

Here's what my last use of the Libre showed (Actual was 37)


Again I would stress variance over HbA1c to an extent, due to specifics of your regime and bodily system. By extent I mean if you had a HbA1c of say 8.2, then sure this should be watched; but if you were maybe 39, had an accurate meter (maybe calibrated with your surgery), fbg of 4.6 and post prandial upper readings of 5.8, I wouldn't worry too much. Please note I have used arbitary numbers in the region of mime, this is not a recognised medical specific set of numbers. I am just comfortable a mid to upper remission number for HbA1c but dialed in meter numbers that I know are accurate (calibrated with surgery).

The below could be useful:

Matt_Harman said:

I was diagnosed nearly a year ago and for one reason or another (moving location, covid etc) I haven't had an official HbA1c test.

I've been using the Libre for about 6 months and was interested to know how accurate the 'Estimated A1c' is compared to a proper one?

Mine currently says 6.7% / 50mmol but I have no idea how accurate it is or if that's even a reasonable place to be.

Any help/input would be great

Click to expand...

I was wondering the same thing, I’m self funding and this is the second libre I’m wearing so I probably don’t have enough data yet. Be great if it was quite accurate though! As not sure when we will get a blood test for hba1c any time soon...
They base so much on that result so it would be good to know the current level.
My last few where 69, 56...so anything around that I’d be happy with or lower. 50 seems like a really good level to me well done! ...think 48-56 is a range to be in for good control...but I could be wrong there! Someone else may know better.

Matt_Harman said:

It seems as though it's good enough to give you a 'rough idea' of where you're at but comes in a bit lower than the actual reading.

Thanks Rose! You'll probably need to get through a few more sensors to get a more accurate estimation but heading in the right direction by the looks of it.

Click to expand...

I hope so! I haven’t had a hba1c since I’ve been on insulin...be nice to know it’s helping!

Hi, just seen this thread, sorry if it has been mentioned but I don't think it has yet... I asked my consultant this a few weeks ago as I had my annual clinic over the phone but hadn't had an HbA1c done. He said quite honestly that while, yes, the libre can be inaccurate, there is also a tolerance of allowed inaccuracy with HbA1c blood tests. It is well documented to be false if you have anaemia for example. I think someone on here referenced the fact that HbA1c, libre sensors and testing strips can be within a tolerance of +/-15% of the true value. So no health care professional would ever expect them to all be exactly the same anyway.

It is also essentially a snapshot, so while the HbA1c is a way bigger snapshot and so better than just capillary glucose tests, the pattern information you get from the libre or CGMs is invaluable. He urged me to keep looking at my graphs to try and avoid spikes, as it will ultimately also aid lowering your HbA1c, even if your true values are constantly 6 and the libre is telling you 5. He said he looks for graphs as flat as possible (within good range of course) over low HbA1c.

Of course as with all tech it is only as good as the thing taking the measurement. So with capillary test you want the best strips, and with the libre I assume they are working to improve the filament function. I would assume it will improve over the years, so try to do at least a weekly capillary test to see how they compare, but love the ease of the libre and graphs it gives. I can't wait to see how my projected HbA1c compares eventually!