HbA1C much higher than finger prick and Freestyle Libre 2 sensor

Taz786

Newbie
Messages
4
Hi All,

I was diagnosed with type 2 diabetes about 6 months ago. My HbA1C was 54 mmol/mol at that time and my weight was 73kg with BMI of 23.

Since then I have made siginicant life changes including low carb diet, smaller portions, daily excersie, regular checking of glucose readings (3 to 4 times a day). I have lost ~7kg weight, My glucose readings after 2 hours of a meal always been below 6, fasting readings always around low 4 and before meals generally between 4 to 5.5 or so. The glucose readings were tracked with mySugr app, and this was predicting an A1C of 5.0% or 31 mmol/mol.

However, I went for 6 month test and was surprised to receive the result of 49 mmol/mol. I was very surprised with this as I had eaten less than 50g of carbs a day for 6 months , no sweet treats, regular exercise, weight loss and so on and literally no difference in my HbA1C reading!!!

Then I used a Freestyle Libre 2 sensor to monitor the glucose continuously rather than spot checks with finger readings. With no surprise my continuously measurements are measuring an average of 5.1 over 14 days, with hardly any spikes over 7.8, and estimated A1C of 29 mmol/mol......

Has anyone else had similar situation where the HbA1c test is coming much higher than the finger prick readings or continuous glucose monitoring device such as Freestyle Libre?

I am questioning if I have another underlying health condition that is creating the high HbA1C reading?
 

Goonergal

Master
Retired Moderator
Messages
13,465
Type of diabetes
Type 2
Treatment type
Diet only
Hi @Taz786

It’s quite common for HbA1c results to vary - either higher or lower - from finger prick tests. The finger pricks are snapshots in time and it’s fairly easy to miss highs or lows.

With the Libre, did you cross check at all with finger pricks to see whether is was similar? The Libre takes its measurements from interstitial fluid, rather than blood and it is common for them to run lower (sometimes significantly) or higher than blood glucose meters, so again there is room for difference, even though the monitoring is continuous.

As you only used the Libre for 14 days (if I understand correctly), then again, it won’t be a direct comparison as the continuous monitoring was only for a short period.

Editing to say that my estimated HbA1c from the Libre was much closer for my recent test, now that they use the Glucose Management Indictor rather than just the average scores:

‘GMI tells you what your approximate A1C level is likely to be, based on the average glucose level from your CGM readings for 14 or more days. GMI gives you the A1C level that would usually be expected from a large number of individuals with diabetes who have the same average CGM glucose level as you.’

Having said all of that, there are people for whom the HbA1c is consistently above (or below) expected results due to differences in the length of time red blood cells last in their bodies.

I’ll try and find some links and add them later. Here’s a link (not to the full article, unfortunately), but I found a PDF of an older version of the same study - attached - which should give you the general idea: https://academic.oup.com/jcem/article-abstract/104/3/707/5183246

Also editing to link to some relevant threads on here:
https://www.diabetes.co.uk/forum/threads/can-your-a1c-be-too-low-article.147926/page-2

https://www.diabetes.co.uk/forum/threads/hba1c-test-biased-towards-the-most-recent-weeks.151557/
https://www.diabetes.co.uk/forum/th...-can-be-misleading.127353/page-2#post-1590384

It’s disappointing to get a result that’s not as low as you hoped, but a 5 point drop is not to be sniffed at, so try not to be too discouraged!


Edited for typo.
 

Attachments

  • 2421.full.pdf
    70.4 KB · Views: 147
Last edited:
  • Like
Reactions: Robbity

barrym

Well-Known Member
Messages
803
Type of diabetes
LADA
Treatment type
Insulin
If I were you I'd feel pretty happy with all those figures. Don't stress over it. And current view from my HCPs is Time in Range is the key now. HbA1c is IMHO an outdated measure based on old tech.
 

searley

Well-Known Member
Retired Moderator
Messages
1,889
Type of diabetes
Type 1
Treatment type
Pump
Dislikes
Diabetes, not having Jaffa Cake
In my opinion The hba1c estimation on mySugr sucks to start with it reads atleast 20point lower than any other method I use

Also mySugr seems to be very heavily biased to the last 7 days…. In my case if I’ve had 3 really poor months but the last 7 days I’ve been really good the hba1c estimate is good

There is also a problem with using finger blood tests to estimate hba1c in that finger tests are only a snapshot at that time. And the higher results in between are not seen and therefore not calculated

So you eat.. and for 2 hours after you are higher.. At 2 hours you take a test and you are back low again…. But your tests won’t know about the higher results

Also what happening over night you could spend 10 hours in bed and the levels could creep up and come back down

The only way to get a reasonable hba1c estimates is cgm as this gets a reading every 5 minutes and sees all the highs and lows


In the attached picture you will see mySugr sees the 6.1 and the 6.9. But if you look at my cgm line in between I was between 10 and 15 most of the time




9fbeb996be07a878a037a95af425b42a.jpg
 
Last edited:

Taz786

Newbie
Messages
4
Hi @Goonergal,

Many thanks for your comments and information shared. I will have a read of it. I have cross checked the Freestyle Libre readings with finger prick blood samples and glucometer several times and these are not too dissimilar, with +/- 0.1-0.4 variations observed.

My Glucose Management Indicator (GMI) is 5.5% or 37 mmol/mol on Freestyle Libre.

@barrym thanks for the encouragement. Greatly appreciated.

@searley I appreciate finger pricks are just a moment in time, so you do not get the full picture in terms of what is happening inbetween those times. Hence, the use of Freestyle Libre sensor for continuous monitoring.
 

searley

Well-Known Member
Retired Moderator
Messages
1,889
Type of diabetes
Type 1
Treatment type
Pump
Dislikes
Diabetes, not having Jaffa Cake
Hi @Goonergal,

Many thanks for your comments and information shared. I will have a read of it. I have cross checked the Freestyle Libre readings with finger prick blood samples and glucometer several times and these are not too dissimilar, with +/- 0.1-0.4 variations observed.

My Glucose Management Indicator (GMI) is 5.5% or 37 mmol/mol on Freestyle Libre.

@barrym thanks for the encouragement. Greatly appreciated.

@searley I appreciate finger pricks are just a moment in time, so you do not get the full picture in terms of what is happening inbetween those times. Hence, the use of Freestyle Libre sensor for continuous monitoring.

Libre will give better estimate but you’ll need the hba1c estimate from the libre app not mySugr

Libre will still be slightly out as they always tend to read slightly high or low for most people
 

Daphne917

Well-Known Member
Messages
3,320
Type of diabetes
Treatment type
Diet only
@Taz786 it’s going in the right direction and you’re getting there! Did they prescribe you statins when first diagnosed as this can affect BS?
 

Taz786

Newbie
Messages
4
Libre will give better estimate but you’ll need the hba1c estimate from the libre app not mySugr

Libre will still be slightly out as they always tend to read slightly high or low for most people

@searley yes I agree looking at the comments from other users. It is just my estimate is much lower than the actual HbA1c result.
 

searley

Well-Known Member
Retired Moderator
Messages
1,889
Type of diabetes
Type 1
Treatment type
Pump
Dislikes
Diabetes, not having Jaffa Cake
@searley yes I agree looking at the comments from other users. It is just my estimate is much lower than the actual HbA1c result.

As I say mySugr is far out for me
 

Andydragon

Well-Known Member
Retired Moderator
Messages
3,324
Type of diabetes
Treatment type
Diet only
@searley yes I agree looking at the comments from other users. It is just my estimate is much lower than the actual HbA1c result.
Finger prick most definitely are not a good way to try and match to hba1c due to individual point in time testing and the 15% margin of Error

For CGM, others have posted inaccuracies quite often when the sensor is laid on overnight, new sensors etc. It's good for a view and trends and for type 1s i can see for risk of hypo, but for hba1c tracking it doesn't seem too good

I know the 5 point drop in hba1c isn't what you hoped but it isn't as insignificant as you seem to think,it's a good drop and 49 is nearly pre-diabetic, that's good! it's in the right direction and is a good drop, time may well could continue a drop and you are doing the right things
 
Last edited:

UK T1

Well-Known Member
Messages
334
Type of diabetes
Type 1
Treatment type
Insulin
Hi, in terms of libre 1, when I first had it prescribed I found my HbA1c prediction in the librelink app wouldn't show until I had 90 days of data, perhaps this has changed now? The reason being the the HbA1c usually gives an indication of blood glucose over the previous 90 days, so using data which covers a shorter time period is likely to give inaccurate predictions.

Also, I have always found the AGP prediction of HbA1c which you get on the LibreView website to be more accurate than the prediction formed within the LibreLink app. Hope that helps a little.
 

LittleGreyCat

Well-Known Member
Messages
4,247
Type of diabetes
Type 2
Treatment type
Tablets (oral)
Dislikes
Diet drinks - the artificial sweeteners taste vile.
Having to forswear foods I have loved all my life.
Trying to find low carb meals when eating out.
Hi, in terms of libre 1, when I first had it prescribed I found my HbA1c prediction in the librelink app wouldn't show until I had 90 days of data, perhaps this has changed now? The reason being the the HbA1c usually gives an indication of blood glucose over the previous 90 days, so using data which covers a shorter time period is likely to give inaccurate predictions.

Also, I have always found the AGP prediction of HbA1c which you get on the LibreView website to be more accurate than the prediction formed within the LibreLink app. Hope that helps a little.

Thanks.
I was just about to ask where I would find the AGP as it wasn't showing on my LibreLink.
 
  • Like
Reactions: UK T1

Jaylee

Oracle
Retired Moderator
Messages
18,232
Type of diabetes
Type 1
Treatment type
Insulin
Hi All,

I was diagnosed with type 2 diabetes about 6 months ago. My HbA1C was 54 mmol/mol at that time and my weight was 73kg with BMI of 23.

Since then I have made siginicant life changes including low carb diet, smaller portions, daily excersie, regular checking of glucose readings (3 to 4 times a day). I have lost ~7kg weight, My glucose readings after 2 hours of a meal always been below 6, fasting readings always around low 4 and before meals generally between 4 to 5.5 or so. The glucose readings were tracked with mySugr app, and this was predicting an A1C of 5.0% or 31 mmol/mol.

However, I went for 6 month test and was surprised to receive the result of 49 mmol/mol. I was very surprised with this as I had eaten less than 50g of carbs a day for 6 months , no sweet treats, regular exercise, weight loss and so on and literally no difference in my HbA1C reading!!!

Then I used a Freestyle Libre 2 sensor to monitor the glucose continuously rather than spot checks with finger readings. With no surprise my continuously measurements are measuring an average of 5.1 over 14 days, with hardly any spikes over 7.8, and estimated A1C of 29 mmol/mol......

Has anyone else had similar situation where the HbA1c test is coming much higher than the finger prick readings or continuous glucose monitoring device such as Freestyle Libre?

I am questioning if I have another underlying health condition that is creating the high HbA1C reading?

My estimated A1c for Diaconnect logging my meter is always 15% higher than what comes back from the lab.
Same with the 3rd party app (0,1 lower.) i use in my CGM set up.

I tend not to read too much into it. Just manage day to day..?
 

EllieM

Moderator
Staff Member
Messages
9,323
Type of diabetes
Type 1
Treatment type
Pump
Dislikes
forum bugs
There is another possibility. For some people (not many) an hba1c can be inaccurate due to various conditions, notably anaemia or B12 deficiency or folate deficiency. The test assumes that your blood cells last 3 months, which isn't the case for everyone. It can over read or under read, depending on your personal pathology. You could ask for a fructosamine test, which will give you an estimate over the last 2-3 weeks, but it generally only gets done for pregnant women, where you need a recent average.