HbA1c and the Insulin Fairies

[Grazer]

Guess it depends what you eat pneu. by definition, a non-diabetic can go to 11 on an OGTT, so if the food was a plate of mars bars ( :clap: :clap: ) then I guess they could get there. I take your point that they normally wouldn't though. Main point I was making is the point you make

the peak being at around +1 hour

This is the same for a diabetic, so our +2 readings aren't a peak anyway. So if we're concerned about 7.8, then 6.5 at +2 may easily not cut it. I know from testing that my +1 can be anything from 15% to 40% up on my +2. But the 7.8 isn't a bogey number anyway. It's just a measure against non-diabetics. As Phoenix says, no real evidence that above that does any real harm. Having said that, I'm with xyzzy and, I'm sure, Borofergie, in that if my +2 was 7.8 or more I'd sulk for a day!

 

[borofergie]

Grazer.. from the GTT studies I have seen on non-diabetics the majority of individuals do not go above 5.5 mmol/l with the peak being at around +1 hour. A small percentage of these individuals may reach 7.8 mmol/l at +1 hour. 7.8 mmol/l is quoted as a figure as its the maximum a non-diabetic reaches after a significant volume of carbohydrate intake...

Here is a lovely picture from a reference that Phoenix dug up, which bears that out (although remember it's only a textbook illustration):

http://www.zuniv.net/physiology/book/chapter27.html

The point is that a diabetic "spike" isn't really a spike at all. The diabetic curve is also more representative of uncontrolled diabetics, most of us here (who manage their carbs) would probably have a profile that would look more like the "impared glucose tolerance" curve - returning to a baseline value in about 2 hours.

My point is that HbA1c is a measure of "spikes/peaks/bumps".

 

[Grazer]

The diabetic curve is also more representative of uncontrolled diabetics, most of us here (who manage their carbs) would probably have a profile that would look more like the "impared glucose tolerance" curve - returning to a baseline value in about 2 hours.

Agree. The diabetic curve, and the figures behind the illustration, assume the diabetic starts from a fasting BG of over 7.8; don't think we'd be happy with that.

 

[Pneu]

Gents.. I am off out now! but I have some graphs that I will post reference post meal glucose from my own testing with different insulin... I realise these are different to the above but perhaps it would be interesting to do a little experiment where we all eat the same and see how we react.. more details later!

 

[borofergie]

I realise these are different to the above but perhaps it would be interesting to do a little experiment where we all eat the same and see how we react.. more details later!

Wow, that's pretty geektastic - you've got me all excited now.

 

[xyzzy]

My point is that HbA1c is a measure of "spikes/peaks/bumps".

Well that depends...

Phoenix said yesterday that the effect of large spikes is unproven, Pneu said yesterday that in "healthy" terms those people who have tight control have better life chances than those who are spikey or peaky, apparently even to the extent that spikey people who end up with lower hBA1c's may do worse than well controlled people with slightly higher hBA1c's.

I like to think of it this way. Bear with me borofergie as I think this is where our disagreement (if any) is.

I agree the data I present below is staged but it's to simplify the explanation.

Take a one day period and split that into six four hour periods.

Take two people.

The first person averages a BG of 6.5 in each of the six periods.
The second person has average readings of 5.0 in five of the periods and 14.0 in the sixth.

I hope we agree that the average BG for both of these people is the same.

Person 1 is 6.5 x 6 / 6 = 6.5
Person 2 is [5.0 x 5 + 14] / 6 = 6.5

Both people repeat that pattern for 120 days and then do an HBA1c test both will have the same HBA1c result as on the day the blood is tested their AVERAGE glucose concentration across the sample taken will be the same at 6.5

Now the difference in our opinions is as far as I can tell is if having Person 1's steady daily pattern makes any PHYSICAL difference to having Person 2's big spike pattern in survivability or health terms.

Pneu's comments on stable BG's being better than spikey or peaky ones seem to suggest that it does as you can imagine a third person who within each of the six four hour periods averages say 6.7. This would give Person 3 an average of 6.7 (6.7 x 6 / 6) and a slightly higher HbA1c than either Person 1 and importantly Person 2 who although having a lower hBA1c has the 14.0 daily spike in his readings and according to Pneu does worse than either Person 1 or Person 3.

This as far as I see it is the crux of the issue. In my opinion it is caused by something along the lines of the following. By "along the lines" I mean what I've written down is just a general kind of idea and it's all a lot more complex in real life!

Imagine an eyeball blood vessel under attack by glucose laden blood cells. Lets use the real value of 7.8 to determine if it lives or dies.

Now you can treat the "liquid" the blood cells are contained in using TWO distinct but importantly different ways.

Way one is to believe that the attacked blood vessel will either live or die based on the AVERAGE concentration of the blood surrounding it. So in both Person 1's & Person 2's case with the daily patterns shown above the blood vessel survives because the AVERAGE concentration attacking it is 6.5 in both cases.

But you can look at it in another way. Imagine each blood cell that passes the blood vessel INDIVIDUALLY. Now in Person 1's case each individual blood cell no matter what time of day only has a 6.5 load of glucose because the person kept to 6.5 during each of the six four hour periods.

It's not the same for person two though. For five of the six periods each individual blood cell as a glucose loading of 5.0 so causes no damage but blood in the sixth period is loaded at 14.0 so BOOM the vessel dies.

It's obviously not quite a straightforward as that because of things like the CHANCE of the vessel meeting any individual blood cell and the TIME a blood cell hangs around before an insulin fairy hits it. What I've described is a very simple model but I hope it gets the point over.

To me Pneu's survivability factoids tend to suggest that you may need to pay more attention to INDIVIDUAL blood cells BG glucose values rather than treating blood as an AVERAGE concentration of BG glucose values.

So that leads me to conclude that yes while HbA1c may be a very good predictor of a persons AVERAGE goodness or badness it can mask or hide the effects of dangerous spikes as I suggest. If I was wrong it would imply you don't need to bother having a "don't exceed 7.8 on any meal" rule and it could just be changed to "don't exceed an average of 7.8"

As Phoenix says its all unknown and still under research but I will err on the side of caution and go with Pneu's stable is best approach.

 

[borofergie]

I hope we agree that the average BG for both of these people is the same.

Person 1 is 6.5 x 6 / 6 = 6.5
Person 2 is [5.0 x 5 + 14] / 6 = 6.5

OK xyzzy <trumpet fanfare> I finally get what you're saying! :idea: </trumpet fanfare>

Suppose that someone, Grazer say, can get by on 150g of carb a day (3 shreadded wheat and 9 chocolate hobnobs).

He could either:
Eat it all in one meal (and get one huge spike and be low all day)
or
Eat three smaller meals (and get three smaller spikes spread throughout the day).

In both cases his average BG and (probably) his HbA1c would be about the same, but in the first case he'd probably be doing himself more damage.

I agree.

 

[xyzzy]

I agree.

Well maybe you shouldn't

According to Phoenix its not known (although I think, never can tell with him) he had some sympathy towards my viewpoint!

(Ducking now)

 

[Grazer]

I hope we agree that the average BG for both of these people is the same.

Person 1 is 6.5 x 6 / 6 = 6.5
Person 2 is [5.0 x 5 + 14] / 6 = 6.5

OK xyzzy <trumpet fanfare> I finally get what you're saying! :idea: </trumpet fanfare>

Suppose that someone, Grazer say, can get by on 150g of carb a day (3 shreadded wheat and 9 chocolate hobnobs).

He could either:
Eat it all in one meal (and get one huge spike and be low all day)
or
Eat three smaller meals (and get three smaller spikes spread throughout the day).

In both cases his average BG and (probably) his HbA1c would be about the same, but in the first case he'd probably be doing himself more damage.

I agree.

Could I have no shredded wheat and 15 hobnobs?

 

[borofergie]

According to Phoenix its not known (although I think, never can tell with him) he had some sympathy towards my viewpoint!

(Ducking now)

Ha ha. Well if "HE" had any sympathy towards you then "HE" probably won't have now because "HE" is actually a "SHE".
:lol: :lol: :lol:

I'd stay ducked for a while if I were you!

 

[xyzzy]

OH S**T I'm really, really,really, really,really, really,really, really,really, really,really, really, really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really,really, really sorry.

Honest with cherries on top and everything.

 

[mistee71]

xyzzy, I loved your story, i have a great imagination and stayed with you all the way. You could make a PC game out of that.. Will you tell me a bed time story, your really good. :lol: :clap: Seriously though, you did explain yourself really well..

 

[smidge]

Hey everyone!

I have to say, I kind of agree with the 'spikes are most harmful' philosophy and I do my best to control them, but that's actually really hard with LADA - easier I guess with Type2. However, if it is actually the existence and density of glucose attached to red blood cells that is the problem in itself and it doesn't matter how they get there, would I be better inducing mild hypos (down to 3.5ish) to counter my spikes? BTW, before anyone shouts at me, that is a theoretical question - not something I'm going to do or suggest anyone else does - but you see my point?

Smidge

 

[xyzzy]

Smidge

I wouldn't worry to much so long as you really aren't spiking on the grand scale I set out in the example (i.e one massive great mother of a spike once / day) If you are just being "peaky" like Sid mentioned earlier then I personally think you're fine.

I set the example numbers up just to show borofergie what I was getting at regarding HbA1c being just an average so I made the numbers extreme just to make the point. Even with the stuff about a steady BG's being better that spikey ones won't doesn't mean ROCK steady BG's as I would suggest that's nigh on impossible even for the most controlled Type 2!

The stuff about average versus individual cell bg levels is just again my own thoughts. In reality I was just pointing out that you could look at what Phoenix said was "unknown" in those two very simplified scenarios. Again I tried to simplify just so show there could be two different views of how glucose laden blood cells COULD operate on other cells not that it did! Undoubtedly its far more complicated than that. If it was as easy as my examples suggested then the answer still wouldn't be "Unknown".

My advice would still be try your best to get as stable BG's as possible. As Type 2 I understand what I need to do to get stable BG's and I kind of understand how well controlled Type 1's achieve the same thing. Not at all saying you at all wrong but I'd like to understand why LADA makes it more difficult? I can then add that to my little black book.

 

[smidge]

Hi Xyzzy!

LADA is difficult because it's very unpredictable. You still produce some of your own insulin, but no first phase insulin and the insulin you do produce varies from day to day. So, you have to keep the injected basal low or you have lots of hypos when your own insulin kicks in. That means most of your control has to come from the injected bolus. So I can be on target after 2 hours, but high (8 to 9) after 4 hours or low depending on how my pancreas is doing that day even with exactly the same meal. If I increase the basal, it would help with the highs, but on a good day my pancreas produces a fair amount of its own insulin and I would struggle with hypos. I generally keep the carbs very low to avoid this situation and correct with extra bolus when necessary - but by then, the spike has happened. The after food spike would be similar to what a Type 1 would get rather than a Type 2 so rapid acting insulin is essential, but the lack of control over the amount and timing of basal makes control interesting!

Smidge

 

[xyzzy]

I'm sure it makes control very interesting indeed!

Have you ever analysed for what makes a good or bad day pancreas wise for example what you ate or drank prior to a good or bad day? If you could get even a rough predictor on that it might help you a bit. Really no expert so could be complete b****ks in which case feel free to point and laugh.

 

[Sid Bonkers]

Take a one day period and split that into six four hour periods. Take two people.

The first person averages a BG of 6.5 in each of the six periods.
The second person has average readings of 5.0 in five of the periods and 14.0 in the sixth.

I hope we agree that the average BG for both of these people is the same.

Both people repeat that pattern for 120 days and then do an HBA1c test both will have the same HBA1c result as on the day the blood is tested their AVERAGE glucose concentration across the sample taken will be the same at 6.5

I think you may be wrong in assuming that both would have the same A1c as I would have thought that the person who had a daily period at 14 mmol/L would almost certainly return a higher A1c than the person who never goes over 6.5. surely thats the whole point of an HbA1c test to show how much glucose has been attaching itself to the blood cells, or am I missing something?

 

[Grazer]

Both people repeat that pattern for 120 days and then do an HBA1c test both will have the same HBA1c result as on the day the blood is tested their AVERAGE glucose concentration across the sample taken will be the same at 6.5

Just one point Xyzzy.
I understand and agree the maths in principle, but they would be unlikely to have the same HbA1c as they would be likely to have different Glycation rates; the rate at which glucose in the blood attaches to platelets. Apparently glycation rates can vary quite widely. So a true average BG of "X" in me could give an HbA1c of 6 for me but 5.7 (say) in you.

 

[phoenix]

So far the models suggested all seem to be simplistic with very gentle movements in glucose levels. Maybe that's what happens in some T2s but I know my glucose levels are far more volatile at times, they can rise very quickly and drop very quickly.
I may have a a spike but it is a spike, not a plateau and probably lasts for minutes. I also have troughs, very low levels. I'm very lucky in that my glucose levels are a plateau overnight... a long period of 'normal' glucose levels.
I'm not at home or I would post a cgms trace so I went looking on the web for a good example. Unfortunately it's on another forum so google: children with diabetes forums cgms accuracy (with luck it's the first link, we're looking at Darryls posts)
The graphs show what I mean by variability. This child had an HbA1c of 5.6%. (about the same as mine) You will note the high levels can be in the 250mg/dl area ;so approaching 14mmol/ but they are not every day and they are transient, there are also lows below 3mmol/l . Days vary, some have lower levels than others.
Her averages are in the 5mmol/l ball park and the standard deviations are within the third advocated by Hirsch.

http://www.diabetesmine.com/2010/12/the ... -care.html
Tells you about Hirsch but also Amy also discusses the evidence on variability

Note what she says about variability between HbA1cs, this seems to be a very bad thing. No use playing hot and cold with control, ie low HbA1c followed by high, followed by low over the years.

One other thing that I haven't mentioned is that the DCCT post hoc analysis showed a difference between micro-vascular complications and macro-vascular complications. It was the HbA1c that was predictive of micro-vascular comps but the average glucose level that was predictive of macro-vascular comps (and they are not the same thing, 2 people, 1 a low glycator, the other a high glycator would with the same averages have widely different HbA1cs)


PS,xyzzy As long as they're fresh cherries!

 

[Grazer]

they are not the same thing, 2 people, 1 a low glycator, the other a high glycator would with the same averages have widely different HbA1cs)

Point I was trying to make above xyzzy; glycation rates can't be overlooked when you're trying to compare average BGs with HbA1Cs. Helps to explain my fairly disapointing HbA1C of 5.9 (indicating average BG of c. 7.6; not checked exactly) when my range of BGs before, after meals, fasting, before bed ALL averaged below 6, some well below! Even allowing for spikes my A1C should have been 5.5 tops, probably lower.