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

Which is worse? - 1hr at 8mmol or 10min at 15

Thank you all, this is a really interesting discussion. I hadn't given a lot of thought to damage from glycemic variability before, and will be paying closer attention to it from now on.
Now the question remains of how to put this into practise. I think many of us find regular numerical measures useful for knowing how well our efforts are working, and motivating further improvement. We can look at a day's or a week's readings and make a reasonable guess at how good it was, but having a number to see and improve might give an extra push.
With CGM or Libre readings we have the data available. What can we use to analyse these to get a value which is better than just time in range, or average glucose at revealing how much risk that level of control is exposing us to?
A quick search on glycemic variability did turn up this, which seems like a fairly intuitive measure:
http://www.healthline.com/diabetesmine/a-new-view-of-glycemic-variability-how-long-is-your-line#1
 
If you use the Libre, you get ambulatory glucose profile and that gives you a good indication of how "in range" you are. The narrower all of the lines, the better - Pale blue is 10-90%, Dark is 25-75%, orange is the median and dotted line is the Max and Min. :

d678bc6a8a71d3f110097ef447e6f133.jpg


If you use Diasend you can also see standard deviation information:

59b021708ca89bac755884c78baf1201.jpg


A good SD is considered to be one where the SD is around a third of your average glucose level. As you can see, these results of mine from over Christmas are rather less than spectacular, both in AGP and Standard Deviation!
 
tim2000s said:
No we shouldn't. Prolonged very high BG levels cause Micro- and Macro-vascular complications. Prolonged moderately high causes some damage. Regular spiking between high and low levels causes oxidative stress and endothelial damage and this predominantly affects the nerves, kidneys and eyes. If you google glycaemic variability there are a lot of papers on this.

The difficult bit is that the oscillation comes second only to prolonged very high levels. In this context, 8 is not high in comparison to 15, and it would really depend over an extended period where you were sitting. To put the eight into context, an Hba1C of 7.5% is roughly the equivalent of an average of 7.7 mmol/l.
Click to expand...
Endothelial damage, now we're talking. See Malcolm Kendrick's seven-part blog series on ED as one if the core causes of heart disease. Part I here: http://drmalcolmkendrick.org/2016/01/18/what-causes-heart-disease/
 
Last edited by a moderator:
It is much worse to be at 8mmol/L all day, and then shoot up to 15mmol/L for one hour, and then back to 8mmol/L for the remaining 23 hours. Its even worse to count the time getting to and from 15 as it will likely be 1-4 hours of extra time above 8.

8 itself for 24 hours is borderline a problem, however it is a CONSTANT problem which makes it a much bigger one.

I don't know enough to make a legit comment - but i can say - if you can sit at 8 you can sit at 6 or at least < 8, id fix that first and just not try to convince yourself that constant minor damage is OK.
 
Diamattic said:
I don't know enough to make a legit comment - but i can say - if you can sit at 8 you can sit at 6 or at least < 8, id fix that first and just not try to convince yourself that constant minor damage is OK.
Click to expand...
Oh don't worry, I don't, thanks! The numbers I picked were purely hypothetical, not my actual readings. I'm fully aware that even moderate highs for sustained periods are damaging. I'm simply trying to get a clearer picture of just how damaging different levels are relative to each other.

Of course we'd like to be able keep perfect control every hour of every day, but most of us don't manage that. We put a lot of time and effort into managing our diabetes, but ultimately the amount of effort we have available for this is finite (as we have lives to live). Therefore it helps to know where to direct this effort most effectively, and to have clear targets. Hba1c alone tends to get focused on because it can be relatively easily measured from a single blood sample, but it misses a lot of the picture. Now that many of us have continuous records of our glucose levels, we could be targeting improvements in different more sophisticated measures which are actually a better measure of reduced risk of complications.
 
Bluetit1802 said:
In other words, a rise of 1 to 1.2 is good, non-diabetic
a rise of 1.2 to 1.5 is modest variability
a rise of above 1.5 is high glycemic variability

Or am I misunderstanding?
Click to expand...

I think you are misunderstanding (sorry!)

It's not "a rise of" x mmol/L to y mmol/L they are talking about, but a glycemic variability index of 1 to 1.2 etc.

While, obviously, to get a lower glycaemic variability index number you would have to try to minimise your rises (and falls), keeping the rise between before and after isn't necessarily enough. For example, you could be going up and down by only 1-2mmol/L, but if you were going up and down like that all day, you would still end up with a longer line on your cgm than if you were steady.

I'm going to reread the bit in sugar surfing where he says the only person with a flat line is a dead person & try to feel a bit less like a diabetic failure!

P.s really interesting thread, even if it is a little depressing. Thanks for the links & info.
 
catapillar said:
I'm going to reread the bit in sugar surfing where he says the only person with a flat line is a dead person & try to feel a bit less like a diabetic failure!
Click to expand...

What I'd really like to do is get a cohort of ten to fifteen people together with Libre or CGM to undertake some testing.

The test is very simple. Over a week, each person will consume a normal diet. Each and everyone of us will eat exactly the same foods and keep the records of all the CGM data over that period, which will be sent through to the researcher (me).

The following week, we'll do the same thing eating a purely low carb diet.

The week after that it will be a purely high carb diet.

The diet sheets would be prepared in advance so that we'd all eat the same thing and each participant would be expected to specify whether they use a pump or MDI. We'd have a record of your I:C ratios and correction factors.

For the week of testing, you'd be expected to go about normal life but there'd be no use of the gym or "normal exercise".

We'd then overlay the data for each of the phases and see which one resulted in the highest food based variability.

While this wouldn't be an RCT it would be an interesting experiment to do...
 
tim2000s said:
We'd then overlay the data for each of the phases and see which one resulted in the highest food based variability.
Click to expand...

It's an interesting idea, but wouldn't you have to exclude a lot more than exercise to be sure you were only seeing food based variability: stress, illness, sleep, menstrual cycle...? It all impacts on insulin sensitivity and therefore what variance the food might have. Also, timing of the insulin v the food... I'm sure we could all come up with more to add to the list!
 
catapillar said:
It's an interesting idea, but wouldn't you have to exclude a lot more than exercise to be sure you were only seeing food based variability: stress, illness, sleep, menstrual cycle...? It all impacts on insulin sensitivity and therefore what variance the food might have. Also, timing of the insulin v the food... I'm sure we could all come up with more to add to the list!
Click to expand...
That's why I stuck the idea in the forum! Happy for people to work out what would need to be normalised.
 
tim2000s said:
Get some funding and put everyone in a bubble....
Click to expand...
Or... Do the experiment on non-diabetics? I don't know if that would work to show variability, think it probably would. But you could be sure their body was already compensating for other factors so you would only see the food variability? Hmm.. Maybe?
 
catapillar said:
Or... Do the experiment on non-diabetics? I don't know if that would work to show variability, think it probably would. But you could be sure their body was already compensating for other factors so you would only see the food variability? Hmm.. Maybe?
Click to expand...
You tend not to see the glycaemic variability in non-Ds because the pancreas/liver pair is so darned efficient!
 
tim2000s said:
You tend not to see the glycaemic variability in non-Ds because the pancreas/liver pair is so darned efficient!
Click to expand...

Ah yes, excellent point - we are special. Saw this comparison on another group, the non-D still goes up & down a bit tho:

image.jpeg
 
Hey @catapillar - Weird question here - You husband on the left seems to have some spikes at around midnight of >7.8mmol/L... Is that cause for concern ?

Also - @tim2000s your experiment seems like a great idea, and honestly i think that is exactly what is needed. Doctors and universities will rarely be able to pull off such experiments due to funding and participation but on a community like this it really seems like some the perfect place to be able to shine some light on things. I do think it will need to be tweaked as you cannot have people go about their 'normal exercise' but restrict diet - there is no way I could do that without eating before hand lol
 
Diamattic said:
Hey @catapillar - Weird question here - You husband on the left seems to have some spikes at around midnight of >7.8mmol/L... Is that cause for concern ?
Click to expand...

It's not my husband - it's a posting taken from another group. But the same question was raised in that group & the answer was that he had had a midnight feast of a tonne of cookies. It's not that unusual for a non diabetic to peak to 8+ with something really carby.
 
catapillar said:
It's not my husband - it's a posting taken from another group. But the same question was raised in that group & the answer was that he had had a midnight feast of a tonne of cookies. It's not that unusual for a non diabetic to peak to 8+ with something really carby.
Click to expand...
Ahhh bickies. Yesterday I fell off the wagon big time. I had a minor hypo, so had some cheese thin bics to boost me, and it took me from 4.1 to 13.8 an hour later. Bye Bye hypo, hello bogeyland. my 4hr PP that followed was 4.7 so my pancreas still works.
 
@tim2000s I've previously done the experiment on myself - minimal variability on low carb! My hypothesis is that despite significant individual differences in weight / hormones / stress / whatever will result is MUCH less variability for everyone. More hypos might occur as minimal physical activity on a nice straight line of 5s will result in lows that wouldn't occur on long post prandial highs...but if you know how exactly how much glucose you need and catch the hypo before it becomes an evil rebound hypo, the difference between 5 and 3 isn't much.

I did the experiment with CGM to show dietitian who kept telling me to eat carbs when pregnant / trying to become pregnant.

I eat normally now and whilst my HbA1cs are 'good/ok', I have HUGE daily variability. I'm normally up to 14 and down again, probably after every time I eat more than 25carbs and have been for years...starting to think I need to get motivated to do low carb permanently.

I'm up for doing your experiment, not bought a libre yet though.
 
Thanks @London36_. Your experiences reflect mine. Managing variability is a great deal more difficult for me with a heavier carb diet.

I've been giving it some more thought and it would have to be presented as a case study rather than a "trial", just a case study of ten participants.

Let's see if there are any more volunteers.
 
Back
Top