How Accurate Is Your HbA1c?
Anyone else think their HbA1C doesn’t match their glucose averages?
Even if you test 10 times a day on a regular blood glucose meter, there are not enough measurements to make a case that your measured HbA1C might be wrong.
I now have 22,074 Dexcom G4 CGM readings over the last 3 months, and it’s been pretty accurate against my meter. That’s one measurement every 5 minutes!
My CGM average for that time, according to the upload at Diasend, is 6.7 mmol/L. That tells me I should have an A1c of around 5.8% or 5.9%, at worst, 6%, if I’m being conservative.
I just wanted one single A1C under 6% in my life to prove I could do it with only a couple of mild hypos a week, if that! But that’s another story.
So what comes back today from QML labs? HbA1C is 6.2%!
That’s quite a difference, don’t you think?
In the past I’ve had high haemoglobin and my endo would adjust my A1C down, but my haemoglobin was getting better and within normal range. I’m now kicking myself that I didn’t get haemoglobin tested at the same time. Still, what a difference!
So now I have to ask… how accurate is the A1C test? How different is the result between different labs?
There are lots of sites to convert average blood glucose to A1C, like here: http://www.diabetes.co.uk/hba1c-to-blood-sugar-level-converter.html … look further down that web page for the conversion from average blood glucose in mmol/L to HbA1C.
That site says my HbA1C would be 5.8%. So why was the lab result so much higher?
How interesting… look at the post by sugar2 here:http://www.diabetes.co.uk/diabetes-forum/viewtopic.php?f=15&t=20432. She says that she got an A1C at two different labs on the same day, two hours apart, and they came out different – 7.4 and 7.1. Maybe I’ve answered my own question. Labs really are different.
My standard deviation isn’t much over the three months, so it shouldn’t make much difference. Besides I have thousands of CGM readings. I seem to recall reading something about people being high or low glycators and that can make a difference to A1C.
There you go… found the research:
High and low hemoglobin glycation phenotypes in type 1 diabetes: a challenge for interpretation of glycemic control: http://www.ncbi.nlm.nih.gov/pubmed/12200073
“…29% of the patients had HbA1c levels that were statistically significantly higher or lower than predicted by the regression equation. The observed individual differences in the relationship between MBG [MBG = mean blood glucose] and HbA1c were not related to erythrocyte age and there was no evidence of analytical artefact. We interpret these results as possible evidence of high and low haemoglobin glycation phenotypes within the population. We conclude that MBG and HbA1c are not necessarily interchangeable estimates of glycemic control and that haemoglobin glycation phenotype may be important for the clinical assessment of diabetic patients.” (Spelling corrected for UK/Australia.)
From what I can tell, high haemoglobin wasn’t an issue in the research, where I assume people had normal haemoglobin. It was more a phenotype distinction, or rather conclusion.
I’ve heard the terms “high glycator” and “low glycator” bandied about, but I wasn’t aware that it was because of phenotype rather than higher than normal haemoglobin values, which if high, would add to the inaccuracy, according to my endo and others.
I’m going to assume that I’m a high glycator, considering the differences between the lab result and the calculated result.
That would mean, if the research paper above was correct, that 29% of us don’t get an accurate A1C! Freaky considering our lives are judged by it.
Think about it! If someone was tested and got an HbA1C of 7.4%, their diabetes team would tell them they have to improve to get under 7%. Yet if they wore a CGM like me, and had tens of thousands of measurements and they were really 6.8%, they’d be congratulated on a great HbA1C.
HbA1C results seem to be taken as fact, when indeed there are extenuating circumstances that can’t be seen. At the other end of the scale, Anaemia will make your HbA1C lower, and so will being a low glycator, which is genetic. Blood disorders that have an effect on haemoglobin do matter, and so does your glycation phenotype.
I’ll take the average between the lab measurement and the A1C/CGM average worked out by formula. Far as I’m concerned I got my A1C under 6… I’ll even take a 5.9999%. I reckon that’s fair, don’t you?
(I welcome any corrections to the conclusions I’ve come to in this post. I’m not a doctor and there are others far more qualified to more properly interpret.)

