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

HbA1c test

P

paulandbez

Active Member
Messages
34
I wondered if anyone could give me some advice? I recently had a private blood test for c peptide and fasting glucose. I am producing insulin (1046 pmol/L) but my glucose was 9.1mmol. I think this is correct because I have been getting morning fasting readings of about 8mmol and I have noticed my readings throughout the day being on the high side. A couple of days ago I went for our diabetic review and blood test. I have just had a phone call from our GP surgery to say the my HbA1c is 37mmol. I was told to stop taking my Metformin and to have a repeat blood test in September. Can a HbA1c blood test go wrong? As I say, my blood glucose readings have been higher over the last few months, especially in the mornings but even when I have eaten the smallest amount of carbs. I would really appreciate your thoughts on this. The receptionist at the surgery said she would pass my concerns on. The strange thing is that last May my HbA1c was 35 and they didn't take me off my Metformin. I had another HbA1c in October last year and it was 43. Many thanks for reading this.
 
paulandbez said:
my HbA1c is 37mmol.
Click to expand...

For what it's worth, that is supposed to correspond to an average bg of 6.2mmol/L.

Yes, hba1c tests can be inaccurate for a number of reasons
Falsely Lowered A1C and Falsely Elevated A1C | Time of Care
The article has a chart at the bottom with some possible reasons for high or low levels...

If your red blood cell turnover is relatively high or low then that can affect the result, but there are other reasons too.
 
Hi EllieM. Many thanks for your help. The chart is interesting. I didn't realise there could be so many things to influence the HbA1c. My wife has just had a call from the surgery to say they are taking her off her metformin as her HbA1c is 39mmol. I'm a bit worried about coming off the metformin but I guess I have no choice. I was hoping that the diabetic nurse would have phoned but instead I got a call from the receptionist. I'm the sort of person who likes to ask questions about my body but it seems that some GP surgeries don't want you to ask questions and just accept what has been said. Thanks again.
 
I reduced my HbA1cs fairly quickly to generaly lowish pre-diabetic levels. MY GP has always been more than happy with my results which I believe mainly due to my low carb diet, but it took him several years to decide to discontinue my metformin. This has definitely had no adverse effect on my glucsose levels/HbA1cs. Any slight blips I've seen both with and without the metformin have coincuded with and been mainly due to health/pain or stress issues.
 
I'm assuming you're T2.

Well, there is an allowable error in the HbA1c test of 6%. For example, a true BG value of 40 could (in theory) actually show results anywhere between 37.6 and 42.4 and the test would still be valid. In practice, you probably wouldn't expect to see such extreme swings but that's why they usually do two tests to confirm diagnosis - especially when the reading is close to the diagnostic threshold.

The link EllieM posted is excellent - BG is affected by more things than just what we eat.

The other thing to remember is that the A1c and the fingerprick tests are testing different things. The fingerprick test looks at your blood sugar at the point of testing (and fingerprick tests have an allowable error, as well) while the A1c counts the number of red blood cells that hvae had a glucose molecule attached to them at some point. It's not really an "average" but it does give a picture of how much glucose has been in your blood over the last three months. It's heavily skewed towards how things were in the last month, because red blood cells only live about three months.

The other thing about fingerprick tests is that you get potential sample error - if you are mainly testing when your BG is at its highest (first thing in the morning, after eating) you'll tend to inflate your idea of what your usual level is. Your level might drop a lot while you're sleeping, for example, and you wouldn't know (without using a CGM).


So if I read this right - over the last twelve months you've had A1cs of 35, then 43, then 37. From this, you might start to think that the 43 is the odd result that needs some explaining. You've had two low-normal A1cs, and one high. Can you think of anything that might have been going on in September/October last year that might have contributed to an elevated A1c?
 
Never be afraid to ask the ques. I see so many mistakes made wich then impacts on patient care and well-being. Too many people think that the Dr is right and therefore take whatever they tell them too without a thought.
Always question their judgment and reasoning for what they are recommending.

Xx

(mod edit)
 
Last edited by a moderator:
For what it is worth I had 2 HbA1c tests at different locations 8th June and 13th June.
First was 7.6%, second was 7.4%.
So it looks as though test results are not precise.
Very rough calculation is that if you assume 7.5% was the true value, the difference was about +/- 1.5%.
This is well lower than @KennyA quoted as 6%.
6% on a reading of 7.5 is 0.45, which could give a range of 7.05% to 7.95% which is quite an alarming range when you think about it.
 
The tolerance is i thought 5%, ie 95% accuracy. The reading could be spot on, or up to 5% out.
 
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.…