The insulin tail

I often hear people talking about the the tail for insulin. Is this referring towards the end of the insulin activity (generally 4-5 hours) where you see a decrease potency effect on your BG?

I’m on novolog right now and my observation when giving correction boluses is that usually around the 20 minute mark, I should expect my BG to start coming down or turn the arrows from rising to steady. My feeling is and I can’t quite put my finger on it, is that by the end of the first hour, the correction bolus has done all it can and doesn’t have any effectiveness like its done it’s job and out of my system already. I can’t seeing it last for 4-5 hours. If my correction bolus doesn’t do a thing within the first 30 minutes, I basically write it off and do another stronger correction. While it seems I’m stacking insulin because my IOB increases, that insulin dose is not effective.

I’m not sure whether its insulin brand related, but this is what I’ve noticed on novolog.

So this brings me to think when people talk about the insulin tail, they seem to be referring the IOB at the end that still has some impact on the BG. For large boluses, I see this, but for smaller correction boluses under 2 units, I think the tail doesn’t exist. Do you have the same observation?

I think tail exists, but very small and not noticed.

If there are fats, proteins in meal, they usually take longer to digest, and do use up some insulin from bolus in the later hours. Consider that those on low carb/high fat-protein, often bolus additional insulin to cover, using extended bolus. Buts gets more confusing with IQ pumps also messing with basal at same time.


I wish the insulin I took was at full activity at 20 min, and out of my system in an hour. But that is not my experience at all. For example, below is a graph of a one unit correction I took yesterday. I’ve had T1 for 46 years, and use Tresiba for basal.

You’ll see that I get almost nothing for the first 30 minutes, but activity getting into gear at one hour. But my BG continues to go down for a full five hours, though obviously the slope of the curve is highest for the first three hours.

I use Humalog. But I tried samples of Novolog, and found no discernable difference in speed or longevity, so I stayed with Humalog simply because it was easier not to switch. But if my insurer ever wants me to switch, I would have no issues with switching, because as far as I could tell, these insulins are virtually indistinguishable to me.


Have you tried Afrezza ?
It is an inhaled insulin, very fast acting.

No, I’ve read about it and asked my doctor for a sample at one point, but he didn’t have any and I was making other changes at that meeting (started Tresiba and CGM), so didn’t press it. I don’t see doctors often, only once since then, and forgot to bring it up again. Thanks for the suggestion, hopefully I’ll remember to ask next time.

Anyway, I think Afrezza would be useful mainly for corrections, and the four unit dose limits its value for that (I know I can split, etc, but that starts to get more complicated).

Insulin tails are a lifelong challenge for me. I think in my case though it’s a matter of not absorbing the insulin right away and then it sits there and 3 to 6 hours later BG goes screaming down. I didn’t have this issue with Afrezza but because of some other issues I started having, it stopped working as well for me. Since switching to a pump it has become a daily issue for me but all I can really do is keep fast acting sugar around to proactively treat the delayed drop.

1 Like

The insulin tail does indeed exist but can be obscured by many other factors such as:

  • An insufficient basal dose that correction insulin must backfill before any effect on BG is noticed.

  • Other hormones like cortisol may be in play causing a temporary increase in insulin resistance and masking the effect of a correction dose.

  • Nutritional absorption of high carb plus high fat meals which can act many hours after the meal. High protein consumption can exhibit similar effects.

Diabetes glucose management is not simply about an insulin and glucose interaction. It is the dynamic interplay of insulin, food, other hormones, variable insulin sensitivity, variable and/or poor insulin absorption at infusion/injection sites, sleep deficiency caused insulin resistance, stress, exercise, hydration status and more.

More info can be found at this NIH article by John Walsh, author of Pumping Insulin.

I speak from painful experience that discounting the action of the seemingly insignificant action of the insulin tail can surprise and threaten your well-being at unexpected and random times. People can be tempted to write these incidents off as the chaos that is diabetes but, if truth be told, can be explained by the scientific fact that insulin duration and action of rapid acting analog insulin is in the 5-6 hour range.


That’s why I chose the graph I did to illustrate the lengthy action of insulin. That unit was a correction dose injected at 1:30am, long after my last food and meal insulin, during a quiescent time without activity. So it shows the insulin action unobscured by any of these other factors.

The duration and intensity of the insulin tail depends also on the dosage. If a bolus is only 1-2 units the tail might be a fraction of a unit “on board” a few hours later which can be easily cancelled out by other factors such as a basal rate that is slightly too low. But if the bolus is 10 units, there could be several units on board hours later which can overwhelm the other factors.

There are some plots of typical insulin duration that illustrate this but unfortunately I’m not able to search them out right now.

1 Like

I have my IOB set at 3 1/2 hours as for my purposes it’s mostly done by 2 1/2 hours.

But it does have a tail, insignificant to me while I am awake. I probably make any adjustments as needed when I snack, exercise etc. If I take insulin in the evening, I do go to bed at a slightly higher level (105ish) to allow for a mild further drop.


Personally, I find the opposite. Intensity of course varies in direct proportion to dosage, but duration is about five hours, with most insulin activity within the first 3.5 hours, regardless of whether I inject one or twelve units.

1 Like

When I was younger, Humalog worked as promised: peak at 1 hour, pretty much gone after 2 hours.
Then I got old, and Humalog works much more slowly.
“Getting old is no fun, but it beats the alternative.”

I think of the tail as being that little extra insulin effect beyond the expected effect. We might not call that tail, but ‘overdosing.’ The excess bolus kinda turns into correction. I see tail reliably from correction insulin.

So, I tend to see lows that start after correct around the 4 hour point. Thats because see see the effect go about 20 minutes beyond 4 hours. That 20 min is what I call the tail and if I deliver too much correction, I will very predictably see lows that start at the 4 hour point.

I don’t usually see a tail on meal bolus, unless I deliver too much.

BTW, I finally got these basals kinda re-established and stable. It took forever!
I had to bump up over 2 units/hour, which I was afraid to do. Its gonna cause mayhem…again when it drops back down to 1 unit/hour.

But, now I can do legitimate bolus testing. Still can’t eat in the morning because I have too much DP/insulin resistance there. I am taking a bunch of short term correction, waiting a couple hours, and then delivering the bolus and eating.