Importance of Matching Meal-GI to Insulin

[Bananas 2]

This is for illustration purposes so folks can conceptualize better what is happening in the body regarding different insulin types vs meal plans. The following graphs are by-the-numbers calculations, not drawn-in curves.

First, a look at the spreadsheet layout to establish these are indeed numerical analyses.

_______________________________
Next, let's zoom in on the R+N example: 2 combi injection of "R" plus "N" insulin. The first graph shows the timing of the injections (with inverse insulin activity, as it behaves in the body, reducing blood glucose) vs the sum of meals in the meal plan. The next graph shows a breakdown of the appropriate meal plan and how the low-GI meal overlap. There are also metabolism and mild exercise factors included.

(Obviously this is an ideal, and there are always hickups with cold/flu, metabolism changes, exercise changes, etc.)

______________________________
Next, an analysis of "A" (Aspart, Lispro, etc) plus "D" (Detemir) with similar exercise and metabolism factors as above.
You can see the meal schedules are quite different. Caveat, the meal-GI at the bottom of this graph are estimates based on the best info I have, as I am not using this schedule. I have asked for input on this forum and elsewhere...
https://www.diabetes.co.uk/forum/threads/meal-gi.132372/

____________________________
OK, so... What happens if we mismatch our meal-GI to our insulin schedule? The next graphs take the "same carb" plans and switches them -- mismatching the GI to insulin.

This illustrates the "roller coaster" so many seem to talk about.
I hope this helps illustrate the importance of matching meal-GI to your insulin schedule.

 

[Bananas 2]

Here is the insulin activity comparison chart.

 

[tim2000s]

Hi @Bananas 2 - what you're describing are the two forms of MDI. The one that many of us started on when it first came out, using NPH and Regular (in Novo terms, Insulatard and ActRapid), which then got moved up to Aspart, Lispro, etc and Glargine or Detemir.

It's worth noting in your graphs that Glargine isn't flat like you've presented, and rarely lasts 24 hours. It also suffers an absorption variance that is similar to NPH. Detemir is best used as a twice daily dose due to its peakiness and duration at average doses, rather than a single dose.

The other thing I'd note is that your schedule misses a couple of important points.

1. By adjusting the timing of the insulin in relation to the meal, you can account for different meal GIs, such that meal GI plays much less importance, and many people do this.
2. Your schedule assumes a uniform absorption in both food and insulin, and this generally isn't the case.

 

[Bananas 2]

Your point 1 -- somewhat, yes, but not all THAT much, unless taking multiple/split boluses... then we're talking pincushion... which is kinda outside the point of the demonstration. Yes, the demo assumes most people might want to minimize the number of injections. It also assumes people might want to eat a complete and satisfying meal, and not eat, say, half a burger (with a high-GI bun), then the other half an hour later.

2 -- Uniform? in the case of low GI food, yes... since that is mostly the case with lowGI foods. The graph of mid-GI foods/meals (2nd set) is not regular, but skewed as is the case with higher GI foods, as with rapid insulin (which the graph also includes a knee - the tapering off end of the curve). Even the "R" curve is skewed (to a lesser extent) and has a knee. So I dont get your reference here. If you are talking about any other personalized irregularities, then it is irrelevant to the point of the demo, which must assume a standard and cannot cover all scenarios.

As for glargine, the flatness is what is presented on all the numerous graphs I found on the web. Since I am not using glargine, I can only go with what data is available. Aside from that, it is not even used in the two examples. However, if you have a graph (and data) that shows a more accurate curve, I'd be happy to use it in the future!!

Thanks for the feedback, but seems you're splitting hairs a bit here.