Can hba1c be too low?

[Andrew S]

This may be wrong, but a blog I read after diagnosis explained BG vs HbA1c something like1 this. BG is a measure of (increased) glucose in the blood which can cause damage. HbA1c is a measure of the damage that has been caused.

Even if this is not entirely true, I think it is a useful way to think about this.

By the way, my GP told me October that I am "too well controlled" with an HbA1c of 36. My Metformin dose has been reduced from 1000mg to 500 mg per day.

Without dragging the thread back to dietary matters too much, I am following roughly her guidelines on cutting down on carbs and "sweet treats" but I am not LCHF. I do a whole heap of exercise, and I think like Douglas, sometimes I do just need to eat for the energy! I am still not testing BG. I can't imagine I am having hypos but before my meds were reduced, I was getting quite ratty at times!

 

[Scardoc]

But it isn't an average. That's the point. It's an amount that increases incrementally as additional hemoglobin particles are glycosylated. Hypos don't cut HbA1c, or even show up in HbA1c numbers at all.

This is exactly my understanding and I don't unfortunately have the time to dig out the websites. This is why they moved away from giving your HbA1c as a % - it was misleading. HbA1c is a measurement of glycated haemoglobin, once haemoglobin is glycated it stays that way so a high HbA1c is indicative of prolonged periods of high blood sugar levels (obviously within a certain period of time, 6-8 weeks).

 

[Andrew S]

HbA1c is not an average, but as the scatter graph shows, for most people it provides results that act like an average!

 

[Bluetit1802]

My last two HbA1c tests have been quite a bit higher than expected from my own meter averages ..... and I test at 1hour, 2hrs, 2.5hrs and 3hrs after a carb meal. I have been studying my full blood count results over the last 18 months. It seems that as my MCV and MCH levels are always on the high end of normal, bordering on the limit, that this indicates my red blood cells are rather large (although not abnormal and not anaemic). I am slowly reaching the conclusion that this is the reason my HbA1c's are higher than expected. Could it be that the larger the red blood cell, the more glucose gets attached?

 

[LucySW]

The whole subject is probably much more complex than we know. And we have to get on with life too .. In my case, @bluetit, the explanation for why my HbA1c wasn't nearly as pretty as I'd expected was that I'd talen lots of lovely measurements of fasting BS. They were beautiful. But I hadn't taken account of the post-meal levels. These weren't huge (I low carb), but that was where the damage was.

This was completely clear once I got a Libre. End of mystery.

So for me, it meant either go and get some bolus insulin, or get my baseline level down by increasing basal. Which I did.

It's fascinating, isn't it. If I was starting out again, I'd love to work in this area.

 

[pavlosn]

My last two HbA1c tests have been quite a bit higher than expected from my own meter averages ..... and I test at 1hour, 2hrs, 2.5hrs and 3hrs after a carb meal. I have been studying my full blood count results over the last 18 months. It seems that as my MCV and MCH levels are always on the high end of normal, bordering on the limit, that this indicates my red blood cells are rather large (although not abnormal and not anaemic). I am slowly reaching the conclusion that this is the reason my HbA1c's are higher than expected. Could it be that the larger the red blood cell, the more glucose gets attached?

Sounds like you have something to discuss with your doctor next time you see him.

Just read that iron deficiency may cause elevated hba1c scores, so I thought I would repeat it in case it helps shed some light.

Pavlos

 

[phoenix]

Imagine your circulation with lots of red blood cells that have receptors onto which the glucose can bind. It only does this if it encounters a glucose molecule. Everybody has some glycated haemoglobin but If you have a high concentration of glucose ie have lots of glucose molecules in the circulation then this is more likely to happen. Once bound then it is irreversible, low glucose levels don't remove the glucose from the haemoglobin. The first step is actually reversible and glucose can un bind and return to the circulation( the full reaction is described as slow, I've never been able tofind that expressed as time )
However cells die after around 120 days, younger cells haven't been around so long so will have less glucose attached than older ones. HbA1c measures the percentage of the haemoglobin that has become glycated. (or now how many mmols out of a mol are glycated)

For some reason some people's haemoglobin seems to become more readily glycated than others . In the DCCT they found ' patients with an MBG of, say, 10 mmol/l can have mean HbA1c values between 6 and10%' As an earlier poster says, this same variation has been found in the recent research used to create world wide standards of HbA1c testing.That is why most countries have not introduced the concept of estimated average glucose level (ie what you get with the convertor on this site) onto lab reports.
People who have lower HbA1cs than 'expected' by their average levels are termed 'low glycators' One piece of research found evidence that low glycators had more 'labile' glycated haemoglobin. ie higher levels of glucose bound in the first reversible stage where glucose can go back into the blood circulation The researchers hypothesise that 'labile' haemoglobin acts as a 'safe' glucose store, to be released into the blood stream when needed .If it remains there though, then it becomes permanently bound,http://care.diabetesjournals.org/content/33/2/273.full


According to Kilpatrick who has analysed the DCCT(T1) data, the best risk predictor for micro vascular complications was HbA1c (and not average glucose), however, for cardiovascular disease average glucose levels (based on 7 readings in a day, every 3months) was a better predictor. http://crossfitmarina.typepad.com/files/diabetologia-2008-kilpatrick.pdf
HbA1c molecule (blue blobs, glucose in circulation and attatched (from you tube video by Casey Steffen)

 

[sanguine]

Once bound then it is irreversible, low glucose levels don't remove the glucose from the haemoglobin. The first step is actually reversible and glucose can un bind and return to the circulation

???

 

[phoenix]

@sanguine
Does this help?

The actual chemistry of the reaction between glucose and protein is complex. In the short term, an intermediate is formed which may dissociate into glucose and protein again. However, over a period of many hours, the intermediate undergoes molecular ‘rearrangement’ and becomes a stable entity. So the initial linking of glucose to protein (haemoglobin in this case) is ‘reversible’ but the overall process is ‘irreversible’.

http://www.diabetesinfo.org.nz/hba1c.html
This one does give a time factor. . It places an interesting perspective on the debate about the significance and contribution of short term glucose variations (ie the daily ups and downs) as opposed to the overall glucose levels to complication risk. (but I've not seen it discussed outside the paper above; there are though a lot of papers!)
There is no consensus amongst specialists as to the significance of short term glucose variability.
first the case for and then the case against http://care.diabetesjournals.org/content/36/Supplement_2/S272.full

 

[douglas99]

Pavlos,,

Have a look at Jenny Ruhl's summary of the latest findings from the ACCORD study. That's the one that seemed to associate lower HbA1cs with increased mortality.

http://diabetesupdate.blogspot.dk/2014/12/accord-final-analysis-lowering-blood.html?m=1

The blog has some interesting information.
'that as A1c rises out of the 5% range the risk for cardiovascular disease starts to rise steeply,

Which is in line with the median of a normal population being 5.6%, and above that being an out of normal range (just)

The blog then goes on to make a great leap though, which seems to have no substance.

'Ideally an A1c near 5% is best,'

Which I suspect is where we have come into this thread,
why aim for a target lower than the average 'healthy' population, and indeed is it actually beneficial, or does it have it's own risks?

 

[pavlosn]

Pavlos,,

Have a look at Jenny Ruhl's summary of the latest findings from the ACCORD study. That's the one that seemed to associate lower HbA1cs with increased mortality.

http://diabetesupdate.blogspot.dk/2014/12/accord-final-analysis-lowering-blood.html?m=1

Thank you for the link.

The blog itself, not being a scientific study, does not seem to justify everything it claim( eg an hba1c of near 5% 31,1 mmol is best).

It does have a link to the actual paper published in Lancet

http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(14)60611-5/fulltext

and looking at that it seems to have findings that again group all non diabetic hba1c (< 6% or 42 mmol,) individuals as one group so does not tell us where within the non diabetic range we should be aiming for.

Thankfully, it did find that people with hba1c <6% were facing a lower risk of ischemic heart disease than those under less tight control, so it does seem to contradict ACCORD.

 

[VinnyJames]

Great post and deeply thought out @pavlosn - I'm always a little sceptical of NHS arbitrary measurement limits and parameters although I tend to accept they are close to the truth, they are continually revised and updated to reflect current evidence.
The good food plate is based on evidence albeit flawed!

I've had a virus for 8 weeks now - longest ever. Am I lacking energy to fight it?? Dunno, but I'm trying to up my carbs temporarily.

If you had asked me what was the most common range of hba1c I would have said 4.5 to 5,5 which is pretty close and represents a standard 'bell curve' approximation.

You would see the same with BP and cholesterol.

I suspect other variables will be at work in all cause mortality and with your morning walks and tennis games you will have more cardio protective benefits than many.

Your doing great and blood sugar levels although highly important for all of us are but one component in the intricacies of the human body.

 

[Squire Fulwood]

Gotta bump this since we are not done yet

 

[pavlosn]

This Harvard Medical school paper from 2011 entitled "how low should you go?" appears to paint a grim picture of the effectiveness of tight sugar control in reducing the risk of cardiovascular complications in t2s.

http://www.health.harvard.edu/newsl...blood-sugar-in-diabetes-how-low-should-you-go

However they do qualify by stating that:

"Not all patients with type 2 diabetes respond to tight blood sugar control the same way. When researchers analyzed subgroups of patients, they found that those with newly diagnosed diabetes enjoy cardiovascular benefit from tight control, while those with longstanding diabetes and established heart disease do not. It may be that tight control can protect healthy blood vessels from damaging glycosylation, but once damage has occurred, it's too late for tight control to help."

Pavlos

 

[pavlosn]

This is an interesting article which seeks to explain why lowering HbA1c below 6% or 42mmol is not dangerous by addressing the findings of the Accord and similar studies.

http://www.phlaunt.com/diabetes/35169265.php

I am generally a little hesitant about trusting blogs and Internet sites, as I often suspect that they tend to filter the evidence they present to support previously held personal beliefs on a subject, but this particular one seems well argued. I have not had the chance to study it well yet, but I intend to do so when time allows.

I posted the link here as it seems relevant, thought it would be of interest and because it presents the opposite point of view to the Harvard paper I posted in my previous post.

Pavlos

 

[LucySW]

This is an interesting article which seeks to explain why lowering HbA1c below 6% or 42mmol is not dangerous by addressing the findings of the Accord and similar studies.

http://www.phlaunt.com/diabetes/35169265.php

Pavlos, here is the postscript to that article, which she has just posted:

http://diabetesupdate.blogspot.dk/2014/12/accord-final-analysis-lowering-blood.html?m=1

 

[phoenix]

There is also this paper from the Accord trial researchers which compares the value of combinations of intensive glycaemic control/lipidcontrol/BP control.
Adding fenofibrates to statins didn't change outcomes, nor did increasing statins to high doses. BP control was important intensively lowering it worked as well as against CVD mortality as intensive glycaemic control.
'These secondary analyses newly suggest that in diabetes patients resembling those in the ACCORD BP trial, either intensive BP or glycemia control reduces major CVD compared with combined standard treatment, but the combination was no better than the individual intensive interventions.'
http://care.diabetesjournals.org/content/37/6/1721.long

The intensive BP control had an aim of reducing systolic BP below 120.

It's as well to remember that Accord did use a cocktail of hypoglycaemic drugs so is a mile away from just using metformin or diet and exercise alone. (and also intensive meant below 6% not low 5s or less; it doesn't really provide any evidence about that)

There is also an interesting point made about long term possible outcomes in this Medscape article. pointing out that there was no benefit from lower glycemia on CVD demonstrated during the UKPDS trial but it was demonstrated with long term follow up The 'legacy' effect of a short term period of intensive glycemic control may be important (that was clearly demonstrated in the DCCT)
Also ,cardiac problems take a long time to develop so implementing tight control in people with longstanding diabetes may not be effective.
They conclude
F

or now, it is reasonable to pursue the lowest possible A1c level that can be reached with a reasonable level of medication burden, at least for those patients who can safely attempt it. Even if there is no CVD benefit, if intensive therapy for relatively short periods can slow or reverse beta-cell deterioration, then the long-term medication burden for people with diabetes might be reduced. Such a strategy could even prove to be cost-saving while improving quality of life

http://www.medscape.com/viewarticle/833202_3
(still adds nothing to the question of lowering levels to that below many non diabetics)
Personally, I take my age into account and note that I have an HbA1c within 'normal realms' for someone of my age. I take insulin so I have another factor to take into account but below 6.5% is at the moment 'good enough' for me (and if I had severe hypos I would reevaluate)

 

[pavlosn]

I am nearly 50 years old, diagnosed t2 5 years ago and have been on diet and exercise control plus metformin ever since.

I also take statins for familial cholesterolemia and bp lowering medication.

Personally I have never, as far as I know, been out of the non diabetic hba1c range since diagnosis, 42 having been my highest hba1c score ( although I may have been higher during periods for which, I have not been tested).

Personally I think of a number in the low to mid thirties as a good score to aim for.

On balance, I remain unconvinced by the argument that I am exposing myself to increased risk by maintaining my glucose at non-diabetic levels , when I am able to achieve a non diabetic score without resorting to high doses of glucose lowering medication.

Equally, I am not aiming for Hba1c of less than 5% or 31 mmol, as on the one hand I am not convinced that there are significant reductions in risk to be gained. Once the risk involved in the intervention necessary to take me a diabetic to such low levels are taken into consideration, any marginal reduction in risk from the reduction in HbA1c may well be eliminated.

Equally importantly, I am happy and feel comfortable with my current diet, exercise routine and medication regime. The effort required to take me below 5% would probably start affecting my quality of life.

So unless some highly convincing new evidence is put to me my own personal intention is to stay as I am.

But of course this discussion was not about me, or not just about me, but aimed to address an issue that is of wider interest and relevance.

I thank everyone that has participated, especially all those that have provided links to scientific papers and other documents. My intention is to revisit these and study them again to try and get a better appreciation of the subject.

Although this post, feels like wrap, being my own personal conclusion from the discussion, it does not have to be the end of the thread. If people still have points they wish to bring to the discussion, I am sure we would all still like to read them.

Pavlos

 

[dannyw]

Douglas, my guess is you are probably not diabetic. Possibly pre-diabetic. With your limited meds and the foods you eat your BG numbers don't suggest diabetes. This is a good thing though eh

 

[pmtbrew]

I've read this thread with great interest, since being diagnosed in August with a A1C of 112mmol/mol (~12%), I managed to get it down to 46mmol/mol in December and now its down to 37mmol/mol (the fact that I restrict carb intake and exercise for around 10hrs a week does not enter into the equations)

Quick bit about me, I'm 38 and have a BMI of 23 so I don't fall into the 'normal' range for diabetics (the first question I ask when I see someone new is how many patients they have seen like me).

The nurse has advised me to remove gliclazide (40mg/day), which I've tried but the A1C jumps up to 50mmol/mol by the third day (with an average of 2hr cardio exercise) so I have went back on the gliclazide very quickly and its taken almost 2 weeks for the levels to return back to normal (I use a Freestyle Libre so have enough data to see short term changes in blood sugars).

Having looked into the NICE guidance and the primary literature (some of them), the recommended levels for a diabetic (48mmol/mol or 5.5%) has a risk of developing complications greater than 1 (more likely) so I do not understand why the levels are not lower (I would say 37mmol/mol should be the target as the risk approaches close to 1).

One point that has not been made yet is the age of the diabetic, I have another 50 years (hopefully) to deal with my diabetes, so any increase risk would affect me more than a diabetic in their 60's.

I'm going to try to see if I can argue the case because of my age, I need to be allowed to have a tighter control of my blood sugars in order to prevent any complications later in life, I have my appointment on the 19th March so will come back with the outcome.