Insulin dose for protein/fat?

Messages
1
Type of diabetes
Type 1
Treatment type
Insulin
Does anyone on a low carb high fat/protein diet calculate an insulin dose for protein/fat consumption? I am aware of the Polish algorithm, which provides a calculation for this, but when I have tried this it did not work for me, too much delayed insulin.

Thanks
David
 

catapillar

Well-Known Member
Messages
3,390
Type of diabetes
Type 1
Treatment type
Insulin
Does anyone on a low carb high fat/protein diet calculate an insulin dose for protein/fat consumption? I am aware of the Polish algorithm, which provides a calculation for this, but when I have tried this it did not work for me, too much delayed insulin.

Thanks
David

What's the calculation for the polish method @DavidXBlaxkman ?

I bolus for protein - I half the grams of protein and tell my pump that half is the grams of carbs I'm having, I extend the protein bolus over a couple of hours.

If you think the calculation you are using is resulting in too much delayed insulin, isn't a simple answe to just reduce the amount of insulin you are taking? Might take a bit of trial and error Until you get to a calculation that actually works for you.
 

tim2000s

Expert
Retired Moderator
Messages
8,934
Type of diabetes
Type 1
Treatment type
Other
I tend to use a far more simple approach when low carb. For protein I use an I : P ratio which is half my I:C ratio for an upfront bolus, and a quarter for an extended bolus, so when I'd do 1u per 10g of carbs, I do a dual wave with 1u per 20g protein up front and 1u per 40g protein extended over approximately two hours. I also don't bolus for fat as there is no point. The metabolic processes that convert fats to glucose convert only a tiny amount and take way too long for a standard bolus.

The tricky bit with the polish algorithm is that you need to go back and check your ratios precisely - typically we make a rough assumption on I:C ratio that includes a bit of protein or fat. You need to go away and test it with pure carbohydrate and plug that in.
 

AmandaD

Well-Known Member
Messages
109
Type of diabetes
Type 1
Treatment type
Pump
I give whatever half my carb ratio is (so I'm 1:12g for carbs and use 0.5:12 for protein and fat) and always as an extended bolus otherwise my blood sugar goes low, dunno if thats right but it works for me.
 

noblehead

Guru
Retired Moderator
Messages
23,618
Type of diabetes
Type 1
Treatment type
Pump
Dislikes
Disrespectful people
I am aware of the Polish algorithm, which provides a calculation for this, but when I have tried this it did not work for me, too much delayed insulin.

There's a lady on the forum called @ewelina who is very knowledgeable about the Polish algorithm, hopefully by tagging her she'll respond to your thread and help out.
 
  • Like
Reactions: Struma

Kristin251

Expert
Messages
5,334
Type of diabetes
LADA
Treatment type
Insulin
I'm not on a pump but MDI. I too bolus for half my protein as if it were carbs. I only eat a moderate amount of protein at each meal or I would have to take a small bolus at around 2 or 3 hours. I also need a good amount of fat in my meal to keep things low and slow or I will gluconeogenisis rapidly. Avocado is key for me to keep all slow and low.
 

ewelina

Well-Known Member
Messages
1,354
Type of diabetes
Type 1
Treatment type
Pump
Does anyone on a low carb high fat/protein diet calculate an insulin dose for protein/fat consumption? I am aware of the Polish algorithm, which provides a calculation for this, but when I have tried this it did not work for me, too much delayed insulin.

Thanks
David
Hi David,
I have the same problem with polish method. I joined Polish facebook group for T1 and asked them how they use it. Answers varied a lot. Some people set delay according to the method (number of fat/protein portions plus 2h) whereas many other just do what suits them. I don't need such a big delay either (unless there is LOTS of fat) and usually do 2h. I have somewhere an article on Polish method in English which I will post here when I manage to find it
 
  • Like
Reactions: azure

Blackers183

Well-Known Member
Messages
164
Type of diabetes
Type 1
Treatment type
Insulin
Dislikes
Commercial television
I tend to use a far more simple approach when low carb. For protein I use an I : P ratio which is half my I:C ratio for an upfront bolus, and a quarter for an extended bolus, so when I'd do 1u per 10g of carbs, I do a dual wave with 1u per 20g protein up front and 1u per 40g protein extended over approximately two hours. I also don't bolus for fat as there is no point. The metabolic processes that convert fats to glucose convert only a tiny amount and take way too long for a standard bolus.

The tricky bit with the polish algorithm is that you need to go back and check your ratios precisely - typically we make a rough assumption on I:C ratio that includes a bit of protein or fat. You need to go away and test it with pure carbohydrate and plug that in.

Thanks tim2000s. I'll have a look at the calculations I set up ina spreadsheet obtained from a document published in 2010 by Journal of Diabetes Science and Technology, lots of 'IF', 'AND' and 'OR' formulae! It seems however that in practice others are doing similar to you. Your comments are appreciated.
David
T1 MDI 25 years

Ps I used another logon for the post as I had trouble getting a new password!

Ps
 
Last edited:

Blackers183

Well-Known Member
Messages
164
Type of diabetes
Type 1
Treatment type
Insulin
Dislikes
Commercial television
Hi David,
I have the same problem with polish method. I joined Polish facebook group for T1 and asked them how they use it. Answers varied a lot. Some people set delay according to the method (number of fat/protein portions plus 2h) whereas many other just do what suits them. I don't need such a big delay either (unless there is LOTS of fat) and usually do 2h. I have somewhere an article on Polish method in English which I will post here when I manage to find it

Thanks ewelina. I'll have a look at the calculations I set up in a spreadsheet obtained from a document published in 2010 by Journal of Diabetes Science and Technology (lots of 'IF', 'AND' and 'OR' formulae) before deciding on next steps, I might have them wrong.

The last comparison was 45gm CHO and 80gm protein (streak) for dinner (plus other green veggies of course). The polish calculation was 9 units insulin at meal time and 8 units 8 hours later. My normal calculation was 12 units of insulin and I only needed 1.5 units correction 8 hours later. Hence I'm seeing that the polish calculation would have been too much too late!

It seems however that in practice others are doing similar to you so this is very valuable information. I have been considering a LCHP diet for some time but want to have a much clearer perspective on insulin doses first.

(I used another logon for the question as I had trouble renewing my password)

Cheers
David
T1 MDI 25 years
 

tim2000s

Expert
Retired Moderator
Messages
8,934
Type of diabetes
Type 1
Treatment type
Other
Thanks tim2000s. I'll have a look at the calculations I set up ina spreadsheet obtained from a document published in 2010 by Journal of Diabetes Science and Technology, lots of 'IF', 'AND' and 'OR' formulae! It seems however that in practice others are doing similar to you. Your comments are appreciated.
David
T1 MDI 25 years

Yes. I built something similar. You can find the algorithm fairly easily on line, but it's a pia if you are out and about and don't have a spreadsheet. That's why I moved to something simpler that basically seemed to work.

Looking at your evaluation for ewelina, I wouldn't have bolused 8 or 9u a number of hours later. THat was never going to work. It's really supposed to be a square wave over a period of time to that level using a pump, which you can of course cancel if you see too much action.
 

azure

Expert
Messages
9,780
Type of diabetes
Type 1
Treatment type
Pump
@Blackers183 If you're still having password or log in issues do say and a member of the tech team can sort it for you :)
 

Blackers183

Well-Known Member
Messages
164
Type of diabetes
Type 1
Treatment type
Insulin
Dislikes
Commercial television
What's the calculation for the polish method @DavidXBlaxkman ?

I bolus for protein - I half the grams of protein and tell my pump that half is the grams of carbs I'm having, I extend the protein bolus over a couple of hours.

If you think the calculation you are using is resulting in too much delayed insulin, isn't a simple answe to just reduce the amount of insulin you are taking? Might take a bit of trial and error Until you get to a calculation that actually works for you.

Hi catapillar
I cannot put all the calculations in the post as there is a lot of (MS Excel) 'IF', 'AND' and 'OR' formulae all which I obtained from a May 2010 paper from Journal of Diabetes Science and Technology. I built a spreadsheet from this paper to see if it works, I'll go back and see if I got the formulas correct before doing anything else.

People in this blog, including yourself, seem to be following a similar methodology for dosing for Protein so this is very interesting information
 

ewelina

Well-Known Member
Messages
1,354
Type of diabetes
Type 1
Treatment type
Pump
Does anyone on a low carb high fat/protein diet calculate an insulin dose for protein/fat consumption? I am aware of the Polish algorithm, which provides a calculation for this, but when I have tried this it did not work for me, too much delayed insulin.

Thanks
David
I found the article I mentioned before. It relates to t1 children , but the method is similar to adults. How did you establish insulin doses for this method?
 

Attachments

  • Application of novel dual wave meal bolus.pdf
    128.9 KB · Views: 488

Blackers183

Well-Known Member
Messages
164
Type of diabetes
Type 1
Treatment type
Insulin
Dislikes
Commercial television
Hi all who have replied to my post on this matter.

I was advised the other day via Facebook that the American Diabetes Society have just published '2017 Standards of Medical Care in Diabetes', the equivalent of the NICE standards.

I had a look through it for any reference to dosing for fat/protein which was briefly mentioned in Section 4 Lifestyle Management. 3 references were referred to, one of them being the below which is easy to access on the internet. This paper published in September last year is "Optimized Mealtime Dosing for Fat and Protein in Type 1 Diabetes. Application of a Model Based Approach to Derive Doses for Open-Loop Diabetes Management", right on this topic!!!

http://care.diabetesjournals.org/content/diacare/39/9/1631.full.pdf

I'll let you find and read it but for me there are a couple off very interesting things to note.

1. They indicate that people using the fat-protein units (I assume the Polish algorithm) method was associated with high incidences of hypoglycaemia (~1-3).
2. The recommended strategy for dosing for fat protein is very similar to what some of you have indicated you follow, well done.

Cheers
David
 

alaska

Well-Known Member
Messages
475
Researchers have tested out insulin pump doses for meals with different levels of fat.

They used a dual-wave bolus.

Interesting to see how the doses changed as fat in the meal was increased.
As fat in the meal increased, higher dose was needed for the second wave of the dual-wave dose.

This seems to back what some of us have found ourselves and I suspect it sort of roughly fits in with the Polish WBT method.

https://www.diabetes.co.uk/news/201...sulin-dosing-in-type-1-diabetes-93144227.html

and
https://care.diabetesjournals.org/content/early/2019/08/22/dc19-0687
 

LionChild

Well-Known Member
Messages
225
Type of diabetes
LADA
Treatment type
Insulin
Dislikes
whisky
Does anyone on a low carb high fat/protein diet calculate an insulin dose for protein/fat consumption? I am aware of the Polish algorithm, which provides a calculation for this, but when I have tried this it did not work for me, too much delayed insulin.

Thanks
David
I don't take any insulin for animal protein....I was told that I didn't need any for protein or salads or vegetables. I am fairly new to this, but it seems to work okay...?
 
  • Like
Reactions: porl69