I know I can go get this from my book, but I was told a couple of months back that I shouldn’t get from a book that which I can get here. (That was before I did a lot of talking… so maybe I’ll be sent back to my book).
When I received my first pump, my AIT was set for 4 hours. I had no idea what that meant, and the trainer just waved her hand at it and told me to leave it where it was. And that is where it remained for 10 years. (I’m a very good listener). About a year and a half ago-ish, my endo had me switch to 3.5 hours. She sat with calculator, crunching the numbers from my 2 BG readings a week I was taking at the time (I’ve come a long way), and she said she could see I needed it. The only other exposure I’ve had to any discussion about AIT is in the 670G world where people, all OVER the place, are lowering theirs to 2 or 2.5 hours. I’d like to bore you on my theories as to why, but I have some self-control. Oh, and one conversation with @Eric where he said, “what’s your AIT?” (Probably not like that), and I said, “3.5 hours”, and he said, “why?” I didn’t know. End of conversation. I still don’t know.
In my group today, a friend mentioned having just learned of the importance of her own AIT, which she has adjusted to 2 hours for auto mode, and we kind of exchanged a few, very few, words about it. I think we’re at about the same level in understanding. I told her I would come over here and enquire and drag back anything I find.
I don’t even know enough to know what to ask. I understand the setting is just the setting and that your body’s time never actually changes, but is it permanent? Can it fluctuate like insulin sensitivity? With blood sugar that runs high after food, would AIT be something to zero in on? How would one know if theirs needs to be adjusted? And what do people believe theirs to be??
I did know what to ask. But I’m open to hearing anything anyone has to offer. To go read my book included.
Different insulins can affect it a little bit (like Fiasp versus Novolog), or things like site location, exercise, temperature, etc.
Generally - faster in would also mean faster out. So if you notice the insulin is working faster, you might also see it leave quicker (this is not 100% guaranteed, just a general guideline).
So the active insulin time (DIA) is basically the time, from when the injection of insulin is made, till when the last bit working in the body to lower blood sugar gets used up.
When you inject the insulin, it has to travel through your fat, your interstitial tissue, into blood vessels, and then make its loop-de-loop through the body before it reaches your tissue and starts working. Once it’s in your bloodstream I believe it’s only active for about 4 minutes, but there’s a pool of insulin that’s sitting in your body but NOT YET in your blood stream that has yet to be put to use for several hours after injection.
Studies have shown that it’s typically in the 5 to 6 hours range for the vast majority of people. In other words, there’s usually some insulin sitting near your pump infusion site inside your body, but not yet in your bloodstream, for about 6 hours.
The reason it’s important is that it can prevent insulin stacking. I.e. let’s say you’re blood sugar is high and you think you need 2 units to bring it down. If your active insulin time is 2 hours, you may underestimate how much is still there later, and may have unanticipated lows.
Practically speaking, you can super roughly estimate that about 20 to 25 percent of the insulin you inject is going to be active each hour. (The amount that gets used up doesn’t go linearly and actually there’s a time of “peak insulin action” which is roughly an hour or 1.5 hours after injection, but as a very rough ballpark it’s useful to think of it like this for me).
Also practically speaking, with newer systems this DIA that you program is JUST A KNOB that you can dial up or down to make your system more aggressive or more conservative. It has a rough physical analog but I don’t think the settings you program in are necessarily going to work best if they correspond to the underlying physiological reality. Does that make sense?
And also if you exercise, maybe that increases circulation and so your body is maybe pulling that insulin into use faster, for instance. So there are definitely “absorption” factors that can change the DIA. But it’s never going to go from 6 hours to 2 hours physiologically, unless you’re injecting it in a different way.
That was interesting. Thank you for the link. In it one comment talked about how it was irresponsible of the doctor to recommend a change from 3 hours to 2 hours. I’m just curious… was it?
If I remember correctly, Kerri was post-pregnancy with kiddo #2 and had been dealing with a lot of highs. Her doctor suggested changing the DIA/AIT as a last resort to deal with those highs.
For me, I would have to try a lot of other things before I try that. I always try to get to the “why” of the problem when deciding on a solution. I cannot imagine a scenario for myself wherein artificially changing the DIA/AIT setting would be the answer. It certainly leads to more aggressive correction dose calculations, but it’s not getting at the real meat of whatever the real problem is because unless I try a new insulin, I don’t see Humalog’s duration going down to two hours on a consistent basis.
In my body, Humalog seems to last three hours when I’m pretty darn active…pushing to 3.5 hours. When I’m super sedentary, I feel like it lasts closer to 4.5 or 5 hours. Fortunately (or not), my control still isn’t quite tight enough for that calculation difference to make huge differences in my results just yet (i.e. if I was running closer to 90 all the time, I’d see more advantage to the calculation being closer to reality. But I’m still averaging high 140’s, so I have a lot more wiggle room).
Just to clarify: I’m not a looper. I don’t think Kerri was a looper in the blog post I linked to. So as far as IAT and looping, @TiaG is the source of great info. I’m only commenting on DIA in general since it sounded like you had questions about that, too.
I think it makes a huge difference and is one of the main levers to pull when you’re using hybrid closed-loop systems. We found with Samson that there was no way to avoid unanticipated lows with a DIA of 4-hours… we upgraded it to six and eliminated those bizarre phantom lows. The issue with these closed loops is you’re stacking upon stacking insulin at an enormously frequent rate. So a difference in 30 minutes in your DIA could mean that you will invariably tick up in the last hour or two of DIA, or that you will inexplicably plummet when there’s supposed to be zero or even negative IOB.
But, in my opinion the DIA isn’t that useful for manually administered correction doses at all. That’s because sometimes you just need so much more insulin for the same number, and to me the main factor determining how much you need is how sticky that number is and past experience. So a high of 200 from overtreating a low with juice will come down with much less than 1 unit of insulin. A high of 200 from eating pizza will take 2+ units to bring down. If we’ve already issued one correction and things are not dropping like they should, we just give between 20 and 50 percent more than we might otherwise. And in the end, you want to come down quickly and that’s easier to do with just overbolusing and feeding the tail anyways. The little bit extra in the tail won’t be a huge hassle to compensate for, usually just another glucose tab.
To me, the rate of change from the last number is the best indication of how much insulin is needed. WE give 0.4 units if Samson is rising +8 or +9, 1 unit if he’s rising +15, and so forth.
I use a DIA of 6 hours in AndriodAPS, which is a realistic number. I do not correct highs using a bolus calculator so the IOB is just used as a gauge of how much insulin I have on board.
DIA also depends on what pump you are using and how IOB is calculated. So my AndriodAPS DIA is 6 hours, my Omnipod DIA is 3.5 hours. Omnipod uses a linear curve for IOB (i.e. a 4 hour DIA will mean 25% of the insulin is used ear hour) while Andriod APS uses a curve that is not a line (that takes into account the time it takes for the insulin to peak and then the time it takes to absorb). The non-linear curve means more insulin is used up front and for the last 3 hours of my 6 hour dia, there is just a small fraction of the original dose left.
The problem with a traditional bolus calculator is it tells you how much insulin you need based on IOB and the difference between your current BG and your target BG… BUT in many cases you may need more insulin.
So… I think people who use IOBs that are less than 6 hours are either using an Omnipod, or are making a choice to me more aggressive with corrections when they use a pump bolus calculator.
I usually do not use a bolus calculator to calculate how much insulin I need to correct high BG. I just guess how much insulin I need based on the CGM trend. This often looks like giving 2 units or so every 1/2 hour until I see the CGM trend change in direction. If it is a steep change I give more, if it is a slow change I give less. If I am really high, I just give lots. This usually results in getting enough insulin to correct the high. (and sometime too much so then I have a cookie to cover the IOB).
So the only time I use IOB is to eat cookies or take glucose to prevent a low when I overcorrect
@Aaron, @TiaG, and @T1Allison, I just wanted to let you know I’ve read all of your responses and really appreciate your taking the time. I want to respond but have been really busy. I’ll shower you all with unnecessarily long responses as soon as is humanly possible… just don’t want you think it was an “ask and run”.
My endo has been adjusting my Active Insulin Time. (670g with Guardian sensors.) It’s now up to 5.5 hours. I understand this is a long time. However, once I get used to the new time, it does work better than what it was before.
Hi @Jack! That’s interesting… sounds like your endo is going the opposite way with your AIT on your 670G. Usually they they head in the other direction and take it all the way to 2.5 hours. Are you in Auto?
3.5. Never budged. Now if you want to hear all about it, you’ve been around long enough to hear it. Seriously, my endo wouldn’t adjust mine because she didn’t like the premise, and then a call to Medtronic confirmed the decision.
That’s the short version. The long version is a LOT more exciting.
Most likely… But only cuz we (me?) kinda like you like a younger sister. As an aside, I’m reading your post and CCR is on the radio playing “I put a spell on you”, so maybe there’s some hidden truth?
Now Chicago is playing on the radio - “Does anyone really know what time it is?”. Weirdness… FWIW, mine’s 3.25. I use this number in the face of all logic because the bulk of my bolus is done by then. An insulin tail is just extra goodness. OK… Enough
I am proud to say, and I hope @Eric is listening, I do 3.5 hours strictly because my endo told me to. It’s one of the very, very few things left that she gets to hold onto. Well, that, and all the prescription power and all my copays and now a global fee I will be handing over to her at every appointment… but that’s it.
Just a 670G enthusiast doing 670G chitchat stuff. 
Seriously, my endo wouldn’t adjust mine because she didn’t like the premise, and then a call to Medtronic confirmed the decision.