How reliable are HbA1c tests?

[ziggy_w]

Hi ziggy_w

thanks a lot. This is very useful.
Do I understand that CGMs are the metres, e.g. Dexcom G6, which measure just beneath your skin?
For illustration Hb1Ac measurement of 53 mmol/l corresponds to a CGM of 160 mg/dl on average.
The plot contains 95% intervals, but even if I eyeball a 67% interval (1 standard deviation)
you can get a variation of Hb1Ac values about 6 values up to 59 and down to 47 which also will
give CGM of 160. This is quite scary for two reasons.
1) When people fret about being below the diabetes threshold of 42, they could be off by a huge margin.
2) I would be surprised if all GPs know that there can be such large discrepancies between Hb1Ac and CGM.

I also read that for individuals the Hb1Ac variation is much more stable.
Thus I have to accept that for me the rise from 42 to 55 is real. Maybe Christmas where I did not resist sweets (which I usually do) and also not fast had a bigger impact on my Hb1Ac.

Hi @Lupf,

Yes, I agree. I doubt many people realize there is quite a large margin of error for individuals.

The reason I looked into this that about one and half years ago, my HbA1c was about 6 mmol higher than anticipated based on finger pricks (and I did quite a lot of finger pricks, so doubt I missed much). My GP suggested that maybe blood sugars were elevated during sleep. I calculated that my blood sugars must have been 8.9 mmol on average at night compared to 5.1 average during the day to arrive at this average, which did not sound likely to me. This was also confirmed when setting my alarm clock at night to check and blood sugars and they were pretty much the same as fasting values during the day -- so no rise during the night.

Anyway when I next met with my GP, I brought some printout of the research. You are right, he had no idea that HbA1c could be off by so much.

Just eyebaling the graph, I would say the one standard deviation is about 15 mg/dl. So, if about 65% of the people are within plus/minus one standard deviation of the mean -- with an HbA1c of 5.0% (or 31 mmol), the "real" (i.e. observed) HbA1c could range from 4.5% (26 mmol) and 5.5% (37mmol). For the remaining 35% of the people, this variation would even be larger (hope my math is correct).

So, with a reading of 42 mmol, you could in reality be either non-diabetic, prediabetic or fully diabetic.

So based on this, I agree with @Mbaker and trust my meter readings more than the HbA1c as I also tested it against lab values and it is quite accurate. Wished I could afford a CGM though.

 

[Resurgam]

Thanks for the info.
Maybe Christmas where I did not resist sweets (which I usually do) and also not fast had a bigger impact on my Hb1Ac.
Breakfast is typically home-made muesli based on oats, some fruit and full-fat yoghurt. I now can't eat whole grains (more like whole sugar) anymore.

You are eating oats and fruit and not testing your response to them?
I'd be seeing high numbers after eating that for breakfast. My usual first meal is sausage or steak, or even a pork chop with mushrooms or a salad. I have to avoid high carb foods, particularly first thing in the morning.

 

[DCUKMod]

My fingerprick average and Freestyle Libre estimates for HbA1c are never close to the lab tests I have done.

I have found the A1cNow home A1c tests to be really very accurate for me, but others have founf the polar opposites to my experiences.

If I had to hazard a guess as to why your A1c has drifted up a bit, I would likely put my bets in the folowing order:

- easing off a bit over the festive period
- your exercise regime changing due to your shoulder situation
- stopping your Metformin

Were I in your shoes, I would look to recreate the conditions that led to your November A1c. If you don't need or want to drop any more weight, add a little more protein or fat to your diet, to take up any slack.

The way Metformin works is to help with insulin resistance and curb the amount of glucose your liver releases, but in reality, it gives almost a smoothing effect. If I recall, the average impact on A1c from metformin, without other changes is something like 0.1%.

A1cs will very over time. Life gets in the way of "perfect living". The culprits can be other health conditions, stress, fatigue, eating, drinking, the weather (literally!), a mixture of the foregoing, and of course,........... just because.

All any of us can do is or best.

Edited to add a little clarity

 

[DavidGrahamJones]

Thanks to losing 10 kg my Hb1Ac reduced to 42mmol/mol on 28 Nov 2019.
I agreed with my doctor to come off metformin (2x 1g /day) in mid-Dec 2019.

Weight loss is no bad thing, but you might be giving it far too much credit for lowering your HbA1c. I lost 40 kgs by lowering my carbs to <50gms per day, but it was the reduced carbs that lowered my HbA1c rather than the loss of weight, which I have proved by eating more carbs. Also, testing just before breakfast isn't going to reflect the effects of eating breakfast, and only testing a few times every week will not allow you to make a proper comparison with your HbA1c which as has already been mentioned only gives an indication of your average BG.

A CGM would be useful but expensive. Even a Freestyle Libre (technically not a CGM) is going to be £50 per 14 days sensor. What is also quite useful is a c-peptide and an insulin resistance test. Most GPs don't hand them out just because you ask, but you might be lucky. When my GP at the time was telling that my pancreas was making less and less insulin, I was able to show that my pancreas was making enough insulin but my insulin resistance was a bigger problem and the value kept changing.

 

[Bluetit1802]

After some questions on this site @Bluetit1802 said she tested on another assay, which produced a lower result. I did 2 tests at Medichecks, one assay was the same type my hospital uses and the other was another type @Bluetit1802 mentioned. The first test was the same result as my Doctors test, the other was 4.5 mmol/L lower.

Yes indeed. After a couple of years of HbA1cs considerably higher than my own comprehensive data from frequent finger pricks, A1cNow and the Libre suggested, my GP went to great lengths speaking to the Consultants at 2 different labs that use different types of machinery for HbA1c testing, and arranged for me to have it tested at both these labs from the same blood draw. The new lab returned a result 4mmol/mol lower than the usual lab. This has now been done 3 times and produced the same difference.

 

[Lupf]

If I had to hazard a guess as to why your A1c has drifted up a bit, I would likely put my bets in the folowing order:

- easing off a bit over the festive period
- your exercise regime changing due to your shoulder situation
- stopping your Metformin

Thanks a lot for this info. This confirms what I thought.
At this time I can see no reason to go back to Metformin,
I will need to find a way to exercise despite the shoulder pain, if it persists.
Regarding diet, I hope to make the weight loss permanent at the end of November level,
i.e. lose another kg, but I seem to have reached a limit in that I need the fast days
just to stay put.

 

[DCUKMod]

Thanks a lot for this info. This confirms what I thought.
At this time I can see no reason to go back to Metformin,
I will need to find a way to exercise despite the shoulder pain, if it persists.
Regarding diet, I hope to make the weight loss permanent at the end of November level,
i.e. lose another kg, but I seem to have reached a limit in that I need the fast days
just to stay put.

If you are unable to do your usual excercise, without pain or doing more damage, then consider something even like walking.

At the moment I am somewhere warm, so I appreciate that helps, but I have a very short walk early morning, the. A longer one later in the day. Of course, if it's blowing a holier or if the rain were horizontal, I'd find it trickier.

I think what I'm saying is exercise doesn't have to be structured.

 

[Lupf]

Weight loss is no bad thing, but you might be giving it far too much credit for lowering your HbA1c.

Thanks a lot. I "disagree" with your literal statement, my Hb1Ac was lowered by losing weight. However we very likely "agree" on the principle. Of course it is the carbs. I lost my weight by eating no carbs for twice a week (5+2 see books and videos by Michael Mosely) and I am eating smaller portions of carbs on normal days. Reading the book "The Diabetes code" by Jason Fung and watching lectures by him and others, confirms that Diabetes and being overweight or obese go hand in hand. Sugar/carbs is converted into fat in your liver. If you eat too much of it, this makes you fat and causes insulin resistance and Diabetes type 2. This needs to (can) be overcome by changing your diet. Sugar/carbs must be reduced and/or temporarily removed.

 

[JohnEGreen]

What about those who are not obese or fat but are diabetic sometimes we take our conjectures and turn them into incontrovertible fact.

 

[HSSS]

Thanks a lot. I "disagree" with your literal statement, my Hb1Ac was lowered by losing weight. However we very likely "agree" on the principle. Of course it is the carbs. I lost my weight by eating no carbs for twice a week (5+2 see books and videos by Michael Mosely) and I am eating smaller portions of carbs on normal days. Reading the book "The Diabetes code" by Jason Fung and watching lectures by him and others, confirms that Diabetes and being overweight or obese go hand in hand. Sugar/carbs is converted into fat in your liver. If you eat too much of it, this makes you fat and causes insulin resistance and Diabetes type 2. This needs to (can) be overcome by changing your diet. Sugar/carbs must be reduced and/or temporarily removed.

So you cut carbs to lose weight, how can you differentiate which part of that equation resulted in lower bgl? There are studies that show if you cut carbs without losing weight then bgl still improves.

 

[DavidGrahamJones]

If you eat too much of it, this makes you fat and causes insulin resistance and Diabetes type 2.

Or does insulin resistance cause excessive amounts of insulin to be produced to overcome the insulin resistance and as we all know insulin controls fat storage. All I'm suggesting is that losing as little as 10kgs which is obviously a good thing, may not be what actually caused the drop in HbA1c. In fact, don't you say that although not gaining weight your HbA1c has gone up and your assumption is that the HbA1c is inaccurate.

I can also say with great confidence that I lost 40 kgs and that weight loss did not lower my HbA1c, in fact my BG dropped so rapidly that I was able to stop taking Gliclazide and Januvia within three weeks of starting the low carb diet and when very little weight had been lost. Also my total cholesterol dropped so rapidly that I stopped taking atorvastatin as well.

Also, as others have suggested, measuring your BG when you do and as infrequently as you do, makes it difficult to compare with an HbA1c which reflects BG for between the last 8 and 12 weeks.

All the best.

 

[Lupf]

I don't know why we are arguing. It seems that what worked for you is different from what I experienced. I log the data and observe a linear correlation between my blood sugar and weight measurements, which I consider meaningful.

 

[HSSS]

I don't know why we are arguing. It seems that what worked for you is different from what I experienced. I log the data and observe a linear correlation between my blood sugar and weight measurements, which I consider meaningful.

Did you do the same for carbs and bgl though?

Most of us find that it’s the carb reduction - regardless of weight loss - that makes the difference. The carb reduction is what causes the weight loss as well as the bgl control in type 2, because you stop feeding yourself sugar that we can’t metabolise thus have been storing as weight gain. So the lack of blood glucose control is what made us fat not the other way around.

Though that’s not to say the weight loss is not beneficial.

 

[JohnEGreen]

@Lupf We are arguing because correlation does not equate to causation. Especially when you are considering only one variable out of many.

 

[Lupf]

@JohnEGreen @HSSS @DavidGrahamJones
Our arguing seems to be a case of "violent agreement" if you know the term. We all agree that eating too much carbs is what makes us fat and causes T2 diabetes and that LHCF works. Of course I understand that the relation between my weight loss and blood sugar is a correlation and not a causation. I never meant to imply otherwise.

Only recently I learned (on these fora) that LCHF can reduce Hb1Ac very rapidly as e.g. David reports. This is very interesting as I came from a slightly different angle where it was all about weight loss. Soon after I was diagnosed with T2, I read about the Newcastle diet (Dr Taylor) where you only eat 600 calories for 8 weeks, which possibly can reverse diabetes. At the time I decided that I would not be able to do this. Since long I had my suspicions about (low-fat) diets which almost always fail. Before a year ago I never dieted in my life. The books by Michael Moseley also seemed to focus on losing weight, especially belly fat. Thus when a year ago I was asked to take additional medicine apart from Metformin, I decided to give 5+2 a go and I wrote down that I want to lose 10 kg. It turned out that for me intermittent fasting worked and that I achieved my goal. I now weigh the same as I did when I was a 19 year-old athlete. I am very happy about this and that (despite the recent setback) my Hb1Ac has come down considerably and my GP agreed that I stop taking metformin.

Over this year I learned a lot. In particle the "The Diabetes Code" by Jason Fung puts all of it on a scientific basis. It all makes sense now. The medical establishment had it all wrong, in particular the cause of getting fat and T2 diabetes - (Hyperinsulinemia) and insulin resistance is - and the vilification of fat.

Regarding my future steps I am continuing with fast days and as someone suggested I try to reproduce the situation I was at the end of last November. I hope the weather improves soon so that I can go cycling again.

 

[lindisfel]

Yes indeed. After a couple of years of HbA1cs considerably higher than my own comprehensive data from frequent finger pricks, A1cNow and the Libre suggested, my GP went to great lengths speaking to the Consultants at 2 different labs that use different types of machinery for HbA1c testing, and arranged for me to have it tested at both these labs from the same blood draw. The new lab returned a result 4mmol/mol lower than the usual lab. This has now been done 3 times and produced the same difference.

I found a similar result where Newcastle was 2 pts lower than Carlisle, measured in same week.
It could be the difference between being diagnosed diabetic or prediabetic.

One cannot really compare hba1cs with differences in an absolute sense. It 's an indication, it all depends.....for example how quickly do we turn over red blood cells, do we have anemia etc etc.
It's problematical and not much good if one has Reactive Hypoglycemia.
D.

 

[Resurgam]

I really can't agree with the statement about too much carbohydrate making us fat and type two - it is the inability to deal with the carbohydrate we are advised to eat which is the problem for me - always has been ever since I was in my twenties - I am almost seventy now. I have always known that I needed to eat low carb to control my weight and feel well - but GPs and dieticians have always thought that they know better.

 

[DavidGrahamJones]

I really can't agree with the statement about too much carbohydrate making us fat and type two - it is the inability to deal with the carbohydrate we are advised to eat which is the problem for me

Isn't that the same thing? Also, I think it depends on the individual. I obviously can't deal with the recommended amount of carbohydrate (too much carb) whereas my wife can't cope with less than the recommended amount of carbohydrate. When I say can't cope I mean she will go weak and wobbly, almost like a hypo . . . almost.

 

[Resurgam]

Isn't that the same thing? Also, I think it depends on the individual. I obviously can't deal with the recommended amount of carbohydrate (too much carb) whereas my wife can't cope with less than the recommended amount of carbohydrate. When I say can't cope I mean she will go weak and wobbly, almost like a hypo . . . almost.

No it is the reverse - being pushed to eat more than the amount of carbs I could cope with made me fat - I should never have allowed anyone to dictate what I ate.
False hypos are quite common when cutting back on carbs - but it is only the brain having a tantrum. Carbs are not an essential food group and once adapted to using fat the metabolism can work perfectly well.

 

[DavidGrahamJones]

being pushed to eat more than the amount of carbs I could cope with made me fat

Pushed? I'll say that the advise is wrong it doesn't apply to everyone. My wife doesn't have to worry about the carbs she eats.