Why bolusing for fat is beneficial

Those aiming for tight control may have intuitively figured out this technique, but now researchers are finally getting around to showing it: Giving an extra bolus for fat reduces post-prandial blood sugar and reduces inflammatory markers in the blood. They are bolusing an additional 30 percent for the fat at the 180 minute mark; I suspect the right dose may be more personalized.
Of course, the downside is that it requires remembering to bolus 3-hours after eating.


This may be a bit of semantics, but the way they are describing it, it sounds like they are bolusing for the carbs later, in addition to the bolus that comes with the meal. Not that they are bolusing specifically for the fat.

Maybe it’s just the way they phrased it.
From this text:
“An additional bolus of rapid-acting insulin to normalise postprandial cardiovascular risk factors following a high-carbohydrate high-fat meal in patients with type 1 diabetes: A randomised controlled trial”

Since it is a high carb meal, it sounds less like they are bolusing for the fat, and more that they are adding a 2nd bolus later for the carbs that are delayed from the fat.

When you eat a meal with a lot of fat, the fat delays the carbs, so you extend your bolus. People have been doing that for a long time. Is this just validation of that idea?

Am I understanding correctly, or maybe I am missing it?

I read your post a second time, and I see I may have inferred from the words “for fat”, and assumed they were talking about specifically bolusing for the fat.

Maybe another way of saying it is - “beneficial to consider fat content when calculating a bolus”

I apologize if I am being unclear. That’s quite possible.

1 Like

I am kinda laughing about this article – they are very proudly expounding upon how they are the first to show that the resulting BG is lower etc., when many have already used this very technique in the past. When I was researching our pizza strategy last summer I reviewed a bunch of strategies doing just that, and published a post in another forum about such a sequence as well, which we still use regularly, in particular for Costco pizza:-)

But it does not matter – the point is that the technique appears to work. However, I am also puzzled by the same thing as Eric:

  • When I read the article, they are clearly not bolusing for the fat but for the delayed carbs, since the rule is 30% of insulin originally injected

  • It is doubly clear that this is what they are doing, since they are making the amount of carbs proportional to the body mass (1g/kg body mass) but keeping the fat content fixed (50 grams).

So, their implication is that it does not matter how much fat you are eating, and that either you are eating too much fat (and should take the extra insulin injection), or not – and you don’t take the extra insulin injection.

This troubles me. Like Eric, I think, I believe that the extra injection is not for the delayed carbs but for the fat itself, some of which will eventually turn into carbs. So the injection should be proportional to the fat, not to the carbs.

Another reason why would be – if the carbs are just delayed, then why aren’t we injecting less in the first injection?

On the other hand, something that may support them is that I have actually experimented with my son, and found that, for him, 30 grams of fat per meal is not quite enough to warrant a separate injection, but that 50 grams are.

I’m not sure why the researchers chose the 30% strategy when I believe another study showed that the variation between people (again, for pizza) was something like using a delayed bolus of 10% up to 80% of the original bolus for the additional fat.

But I do I think there’s a threshold effect … and then a stair-step effect above that threshold. So if you eat a low-fat meal you mostly don’t bolus for the fat. If you eat high-fat you need to. But I think the primary driver of the stair-step effect is the number of carbs initially eaten. So for instance, when our son eats salmon and cauliflower and cheese for dinner, we don’t have to bolus at all… even if he’s eating a fair number of calories of fat. But if he eats the same amount of fat with a bowl of pasta, we need to consider the fat. The more carbs he eats with the fat, the more we need to bolus for the secondary fat spike. But we haven’t yet figured out the right formula. It’s possible it is proportional to the fat grams, and we just haven’t been carefully tracking those because we were taught we didn’t need to.

1 Like

It seems like they’re having you bolus for the carbs upfront, and then an additional 30% (relative to the carbs) for one group of people 180 minutes after the first bolus… so they get the same initial carb bolus, and then 3 hours later they get 30% of what they initially got, on top of it – but only if they ate the high-fat, high-carb meal.

I am guessing they stuck with this arbitrary ratio just for experimental purposes to make it easier to analyze the results…

1 Like

So that’s what I was saying. You bolus for the carbs, not the fat. All the fat does is delay it.

But I agree, there comes a point where you need to take some insulin even if there are no carbs.

I think the idea of bolusing only for carbs is a bit flawed. I have a formula you can try which is a bit different than standard but I can try to find it. It boluses for carbs, protein, and fat at different percentages and with extensions.

When it comes to bolusing for meals, I personally don’t use formulas or calculations, I just do it. :slight_smile:

But the total amount we bolus for just carbs is LESS than for the same number of carbs PLUS fat… and we bolus for the fat later, on top of the initial carb bolus. i don’t actually think the fat simply delays the carbs. I think it increases insulin resistance and increases the total amount of insulin needed for a meal by a significant amount.

1 Like

I agree that you should add insulin for the fat. I would take some insulin for a meal of nothing but fat and protein. I just know a lot of people don’t do that, they only count carbs. They use an IC ratio only, they don’t use IF (insulin to grams of fat) or IP (insulin to grams of protein) ratios.

When you said this I thought you were saying you don’t bolus for it:

I will try to find the formula I have. You can see what you think of it. I think in a big part, people are not instructed to bolus for fat and protein because it adds complexity to the process.

well, that’s the weird thing. If it’s dinnertime and it’s just fat and protein we _don’t_have to bolus for it. If it’s breakfast, a meal of just fat and protein will require twice as much insulin as an equivalent number of calories made up mostly of carbs. No idea what’s going on there, but it’s a very repeatable pattern.

Also, when I say we bolus for the fat, I mean we are doing it on the fly or based on history – we don’t have a fixed ratio for fat, just intuition based on past meals. Perhaps that’s because most of our meals are low in fat unless they’re restaurant meals, in which case we’re just swagging on a lot of the values.

1 Like

If I eat a large meal of just fat and protein, I bolus for it, but I assume I’m covering the protein, not the fat. My understanding was that only a very small/negligible amount of fat can elevate the blood sugar, but in the absence of carbs, protein will be at about a 50% rate or so.

Maybe that’s because when fasting, his body metabolizes the protein for blood glucose because it’s desperate for some, but by dinner, his body doesn’t need it as much, so it’s not converting the protein over? It seems like people eating a keto diet are more prone to metabolizing protein into blood glucose, so maybe your son is essentially like that when he wakes up and hasn’t eaten any carbs for many hours.

1 Like

You know, I read about the “total available glucose” method of calculating dosing and elevations in blood sugar but it didn’t seem to work well for us. My son just doesn’t eat enough protein for it to have a dramatic effect on his BG (despite our nudges)!! He has progressively been rejecting more and more types of animal protein, he hates eggs, he enjoys yogurt but will only eat small portions, he’ll eat a small slice of cheese, but that’s mostly fat, he likes beans and lentils but those are actually quite high in carbs.

There are several components to BG: The first is how many calories of each macronutrient – fat, protein and carb – is metabolized into actual glucose molecules in the blood. The second is how those foods affect the liver and its decision to release glycogen stores or not. The third is how the food you have eaten affects the receptivity to insulin in other cells in the body. My totally unsubstantiated theory is that my son develops insulin resistance and that the absence of carbs triggers a liver glucose dump, when he is confronted with just protein and fat in the morning, because his body interprets the meal as a starvation signal and then takes extra efforts to raise blood sugar. It’s not just that he shifts to metabolizing protein or fat; it’s that the insulin works less well than if he had eaten a meal with more carbs.

It’s the rare day he will eat more than 10 grams of protein in a sitting at breakfast, despite our efforts. So on a good day: Even if every single gram of protein translated directly into blood sugar in the same way as a gram of carbohydrate did, that would raise his blood sugar at most about 100 points. But we’ve seen him stuck at 350 for hours from a BG 80 after eating a little bit of scrambled egg and cheese. So you really have to add in an element of insulin resistance or else it just doesn’t add up mathematically for us.

1 Like

Ah ok, that makes sense! Interesting. Yeah, liver glucose release in particular has been a major factor I’ve had to learn how to deal with, but it definitely doesn’t do that for me in the mornings. In addition to all in the individual differences with that, I’m guessing age is a big factor in that too—will be interesting to see how that may shift over time.

1 Like

In Pumping Insulin, John Walsh suggests starting with an additional bolus for fat and protein in pizza equal to 25 percent of the carb bolus spread out over a 2.5 hour period…and then keeping track of the results and adjusting from there. I can see that my activity level would probably make a big difference, too, in the amount of additional insulin needed. Now that I am pumping, I am going to give this a try!


Morning insulin resistance?

yes, but the insulin resistance is enhanced by the absence of carbs and the presence of fat.

Does some activity before eating breakfast help?

well, that’s a tough question to answer, because he gets up and often wants to eat at 6 or 7 and then he goes to daycare. I guess I could take him out for a vigorous run beforehand, but to be honest I don’t have the stomach for that so early :slight_smile:

I bolus for fats/ protein when I have a really heavy meal.

But the rest of the time, if I did I would take too much insulin. So it’s hard to figure out an exact rule.

1 Like

i am glad that i am on the pump and able to experiment with the dual bolus feature. i have found that ANY food with protein and fat must be split up with a small amount of insulin at start, and then a lagging, square bolus sustained over a selective amount of hours. and there is no set-in-stone formula until i have experimented with the particular food. for example, if i am eating dark-meat chicken, i need to extend my bolus accounting for 10gms of carbs/fat over a 2 hour period with a 40/60% split. if i have a burger on a bun with fries (anything greasy/fried) i need to cover 75gms carbs w/ a 25/75% split over 3 hours. if i am having a grilled cheese sandwich on rye with tomato, i cover for 65gms with a 50/50% split over 2 hours. how i have figured this out? years of experimenting at the SAME restaurant, so that i can keep the portion size right. then i know that if i should eat something similar at a different restaurant, i can kind of guess and modify my formula. i hope that this makes some sense to anyone.

all i know is that the more fat, the slower the carbs digest. if i take too much insulin up front, i crash and then spike later. i could go down to a 30bg in 1/2 hour after eating, and then shoot up to 300 4 hours later. and i hate having to chase those annoying high bgs back down into target range.


i agree with this 100% . this is what happens to me when i eat a high fat meal with carbs. if i were to simply count my carbs and bolus for them, say at a 1:10 ratio, then when i were having a grilled cheese sandwich, i would bolus for the bread, the cheese (a little) and that would be that. it might come out to say 36 grams of carbs. but the sandwich is made on a grill with a pound of butter, and when the cheese is added it slows down the digestion. if i were simply to bolus for the carbs and the cheese and do a dual bolus, it would be rather simple. but instead, when i eat a diner grilled cheese sandwich,i have to bolus for 65 gms at 50/50% over 2 hours extended. that means that i have, approximately, IOB for 6 hours. and that is floes to double the carb count if i weren’t grilling it. its all that darned grease (but isn’t that one of the reasons it tastes so good :wink: )