hi @anon76147412, IOB stands for “insulin on board.” Basically, when you give insulin, it takes anywhere from 4 to 6 hours for the body to use it up. First it is sitting in the fatty tissue, then it has to migrate into the blood stream and be used by the body. So when you give, say, 1 unit of insulin by injection, after about an hour about 0.8 . units is still left in the fatty tissue to be delivered, after 2 hours maybe 0.6 units remains, after 3 hours 0.4 units remains, and so on.
So if you gave her insulin from before (maybe breakfast) and some of it was still not used up by her body, it could have still been lowering her blood sugar hours later. IOB is a really helpful concept wen trying to control blood sugar.
I’ve studied the pharmacokinetics of insulin a lot. It is extremely valuable to point out, it is not linear. The half-life for a subcutaneous insulin injection is about 1 hour. Meaning, you generally get half of your dosage in the first hour, the remainder trickles out over the rest of your duration. Every once-in-a-while you may hit a quirky site that acts differently, but it is very useful to know this - no matter what formulas you may have on your pump or on any calculators, IOB is non-linear. Of all the pumps, Tandem does the best with their IOB formula. I’ve always just done it in my head, because it’s more accurate than anything I’ve seen so far.
Take a look at section 12.3, Pharmacokinetics:
(this is for Humalog, but they are all basically the same)
http://uspl.lilly.com/humalog/humalog.html#ppi Elimination — After subcutaneous administration of HUMALOG, the t1/2 is shorter than that of regular human insulin (1 versus 1.5 hours, respectively).
Let me know if you have questions about this. Because of this, I do my stuff a lot different than most.
Eric, I am very surprised by this. I don’t think this is true.
I think that what may be true is that t1/2 is 1 hour when it becomes active. There is a long period of inactivity when you inject. For my kid, it is 25-45 minutes depending upon what threshold you use. Once the insulin turns on, which is probably when it makes its way into the bloodstream, (say 40 minutes after injection), then you may get 50% of your dosage within an hour there.
What I am saying is that the 0 of the time scale does not correspond to injection time. For is, I can tell you for sure that we get no more than 15% in the first hour after injection, probably less.
For you it is probably longer because you are using those short needles. So your start time is delayed. If you try them, you’ll find the 5/16" needles work better.
But regardless, I’m not making this up. Look at the Lilly site. They do their test with a glycemic clamp. (And they probably did not use 1/6" needles !!!)
yes, of course IOB is nonlinear, BUT I think it’s very helpful to have these kinds of rough rules-of-thumb in your head for when you don’t want to do a calculation – the fact is few of us are going to take a calculator out or draw lines on a graph to calculate IOB (even those of us not using a pump). In general I find this type of ballpark estimation (what I think might be called Fermian estimation) extremely useful. Precision is really great if you are able to do it, but for us, in our daily lives, we often rely on this type of estimation and it works out okay. Not perfect, but okay. The reality is that there are a lot of variables taht, even with precise dosing and calculation, are not accounted for, and so I see a lot of the slop as canceling out. But perhaps that’s also a justification for a hack that has essentially made our lives a lot easier.
Another type of estimation we do – eyeball carb counts in restaurants.
@Michel and @Eric, I think there is also some subtle distinction between how much insulin is left and how long it will continue to lower blood sugar.
For instance, what we’ve found based on our pump’s estimate of IOB – and tracking it with actual blood sugar lowering effect – is that insulin activity peaks in 60 to 90 minutes. To me this insulin activity value, the time derivative of IOB, is way more useful.
TiaG, you are absolutely right. I, also, charted standard drops. For my son, the most active part of insulin action starts around minute 40 and stops around minute 110.
It does not have anything to do with the size of the needle:-)
I remember reading about the measurement process. There is a part of the experiment linked to how they measure the effect which adds time to the beginning and the end of this curve you see there, which must be added in order to compare it to DIA. You cannot take this curve and make it start at injection time.
If you remind me of it in a few days I will look it up and document it. But possibly someone can provide the information faster:-)
No doubt (although every kid in the US uses the same needles, and probably most recently diagnosed adult patients). But this is not the reason, though. Every formula I have read of the amount of insulin available in the first hour shows a small number – the mainstream, nonlinear formula I know best, for instance, gives 11% as the first hour number for a 5-hour DIA. I think this is the reason for the discrepancy:
Courtesy of Pumping Insulin, John Walsh & Ruth Roberts, p.99
I believe this applies to the curve you showed – I think the curve you showed does not include the 0-C duration on the left side of the DIA curve. I could be wrong, though.
Insulin may not absorb as fast when they’re sleeping, but that seems to be pretty variable between people. In general, insulin action seems to depend on the person, the site and a bunch of other things we can’t predict all the time.
OmniPod and other pumps all track the insulin on board. They each use different formulas for this, though.
I had a g6 but had to go down to G5 because of medicare. I no longer can tell qhen my sugar is low. My G6 was always spot on. G5 is crap. I was feeling icky but my G5 said I was at 145. I did a fingerstick and it was 38 and I was starting to feel it. Why was my G5 so far off? Had I been sleeping I would not have woken up as i cannot get any of the Dexcom apps on oir new phones. Not supported I guess but the G5 receiver would never wake me because its so quiet. I have it set to alert me at 80 because I am very sensitive to insulin. I have been a diabetic for over 30 years T1. I also have gastroparesis so I don’t dugest as fast as I should. I also have bronchiectasis with adult CF so managing my diabetes is hard. In Jan the EMTs were at my home 16 times in 14 days because i was sick and ended up having blood transfusion and Iron infusions. After that happened I fought hars to get my G6 with my insurance but was switched to medicare in June and was put back to G5. My question is what can we do to make this CGM work properly? I calibrate when needed and I do calibrate when there is an extreme differene. I have to do fingersticks at min 6 tumes a day because the CGM is so unreliable. And don’t get me started on sensor failed(s) any suggestions will be greatly appreciated. I am overtired because i am alone a lot because my other half bless his heart works abd isn’t always around and I am too scared to sleep.
