Insulin pre-bolus: a necessary technique for flatlining

A well-timed pre-bolus can do a great job with many meals! We find that there is no way we can stop meal spikes unless we do a good job pre-bolusing.

First, we established how long it takes us for our insulin to become active: for us, 40-45 minutes after injection, Novolog (or Humalog) becomes strongly active, and strongly inflects the BG track downwards. We call this “turning the corner.”

If we are already within range when pre-bolusing (we like to be between 90 and 100), we time the pre-bolus exactly according to the activation time, i.e. for us we pre-bolus 40-45 minutes prior to the meal. If we are a bit high, we add enough time to come down by that much. If we are lower, we typically eat very few carbs to bring us back up - maybe start with a lunch appetizer that has 3-4 carbs, just enough to give us a boost (and, of course, take off the appetizer from the carb count).

Then we wait until we turn the corner: once the corner is turned, we eat RIGHT AWAY!

Below is my boy’s lunch curve today:

Pre-bolus at 11:05, lunch (59 carbs) at 11:45. The red line is at 70, the yellow line is at 120.

It is not always as smooth :slight_smile:

Notes and comments:

  • Time it takes for insulin to activate varies widely per person: YDMV!

  • The glycemic index of the meal makes a difference to how early you would want to pre-bolus. A meal with a low glycemic index needs a shorter pre-bolus interval. Again, YDMV!

  • Time of day makes a difference to the efficiency of pre-bolusing. Again, YDMV!

  • Because of the significant individual variations, this is a process that needs personal experimentation.

End of wiki ---------- comments start here


I agree that pre-bolus is a powerful tool. It is harder to know when you are “turning the corner” (I assume this is an Insulin Surfing term?) when you don’t have a CGM (I don’t have a CGM), but I know my insulin action curve well enough that I don’t need to see it visually to predict when it will be occurring with decent accuracy. So I would add that pre-bolus is a powerful tool even if you don’t have a CGM.

Another note, is that I have found I need to consider not only where my BG is heading, but also the food that I’m going to be eating. I find quite a difference in timing when I eat lunch, for which I almost always eat bread of some type. If I am eating 100% whole wheat bread (my typical lunch), then I need to give myself a very short pre-bolus or I will go low. But if I am eating white bread (hot-dog, hamburger, …) then I use a prebolus of 30-40 minutes.


What an interesting demonstration of the Glycemic Index!

if my BGs are in range before eating, i wait 10-15 minutes for pre-bolus. if i wait longer than 15 minutes i go low very quickly. but if my BG is high before a meal, i will wait around 20 minutes, depending on how high and out of target i am in order to bolus. if i am eating something like oatmeal, i need a full 30 minutes of pre-bolusing before delving into the bowl.

i am astounded when i hear about these long pre-boluses. it would be a complete disaster if i tried to wait that long. also, it was my understanding that the insulin ( i use Novolog) starts working at around 15 minutes, so why would i wait longer than that if i am in range?

can someone please explain it to me, b/c perhaps i am doing something wrong.

Daisy Mae

I think this is one of those cases where you need to test yourself to see what your results are. Starting from normal range (BG 80-120) I usually don’t see much of a downward trend until about 45 minutes, but it generally moves quickly at that point. I really don’t see any BG movement at all until 30 minutes have elapsed (with Humalog, though when I tried Novolog I saw the same result).

1 Like

Indeed, which is why I am always surprised when I see someone say that the glycemic index has no value for them. I’m sure I see differences in carb response time depending on the composition of the food I’m eating.


i know YDMV, but there are so many who do what you do, waiting for this downward trend. am i the exception to the “rule?” also, what do you do when you are in a restaurant or at a dinner party?

i remember the days when i was taking “Regular” (R") insulin and i was told to wait 30 - 45 minutes before i ate after injection. when i began on the Humalog/Novolog, i was told to eat immediately after i bloused. there was no pre-bolusing back then…although that could have just been a lack of education on my part. i thought that Novolog was the miracle insulin of the century. it made everything so easy. i can’t begin to tell you how confused i am. all of you guys pre-bolusing for so long without crashing. its deff beyond me.


I pre-bolus by 15 mins if in range, 20 mins for “moderately” carby things, and 30 mins for very carby things (like ice cream!). If I regularly did a 30-45 minute prebolus, I’d be low.


yay ! someone who understands me!!! yahoo.



We find that Time of Day is relevant for pre-bolus timing as well as the food. Morning works best with a 45~50 minute pre-bolus.


I find so many different factors effect how quickly the hunalog acts - some that make sense others that don’t

  • time of day definitely - I don’t eat carbs in the am anyway but when I inject my 7am hunalog it doesn’t start working for at least 1.5hrs - in the early pm it’s 30-40mins - evening it’s 20mins before I see effect on the cgm
  • amount I take - no idea why but if I take 3 units for a low carb meal (egg and avocado) vs 8 or even 10 for high carb (wholemeal sandwich) then the larger dose starts working faster as well as obviously creating a steeper curve
  • if I drink a black coffee around the time I inject then it starts working faster also as if the caffeine is accelerating the effect !
  • if I have a drink with dinner (vodka / soda) this will often completely negate the rise afterwards - my doctor tells me that my liver stops producing sugar so it can tackle the alcohol (poisoning) and my basal insulin then immediately switches to the food quicker than the bolus injection (I also reduce the bolus amount) - I actually find it’s quite a good way to smooth the rise - not sure why this hasn’t been researched more as a means to help manage bolus spikes
  • lastly per my other posts - when on trulicity the reverse happens and the release into my blood of the sugar is slowed down so much (up to 2 hrs delay when eating fatty meal ) that I can eat the food first - assess how much carbs and what type I ate and then inject to avoid an unwanted rise up to 20 mins after the meal !! - the trulicity doesn’t effect the timing of the bolus working but slowing the food release down it allows the bolus to always be ahead - I think this is one of the best effects of trulicity for type 1s that should be looked at closely

I don’t need to wait that long either. Only if very high before.

I think our pharmacokinetics and pharmacodynamics are very similar. :wink:


That probably means that your insulin needs 30 minutes to activate :slight_smile: For my son, it is 40-45 minutes. That is why the first step in the process is to find out how long it takes for you.

The point is not the duration – it varies with everyone – but the fact that you need to measure how long it will take for you! The nurse I saw today told me her pre-bolus is 1 hour and 25 minutes, for instance. Each of us has a different duration. We all need to know what it is. We can’t use each other’s timing. But the same principle applies to most of us — although even for that there may be exceptions :slight_smile:


must be, b/c i was starting to feel all alone out there. :wink:

I’ve been reading a bit on #DIYPS about this. Dana Lewis calls it “Eating Soon Mode”

What’s good is that it can be accomplished with MDI fairly easily. This prebolus technique also ties in with carb absorption rate.


This carbohydrate absorption rate was a great idea!

Right! Definitely a YDMV situation.

1 Like

I rarely do more than 15 minutes. Today was about 10.

It just depends for everyone.


I updated the wiki to reflect input so far.