Insulin load index / most ketogenic foods

Heathenlass

Well-Known Member
Messages
1,631
Type of diabetes
Type 1
Treatment type
Insulin
Hi Signy

My review of the data from the University of Sydney didn't show that fat has an impact on insulin. TAG also doesn't account for fibre. The original TAG book seems to be out of print. Do you know what the basis of these factors is at all? I've searched, a lot, with no luck.

Marty Kendall
https://optimisingnutrition.wordpress.com/

Hi Martin :)

I have the original TAG book, somewhere ! I'm not the tidiest of people and haven't as yet fully unpacked from my move last year :oops: Perhaps this issue will motivate me to get round to it :D When I excavate the info, I'll let you know ;)

You are right, original TAG didn't account for fibre, and therefore I never have as I count carb minus fibre as this is the way the value for most foodstuffs are given in the UK. It's true that the Sydney study showed no impact from fat, but I find it does affect me. Why, I don't know, but I'm guessing we are as individual in response to fat as anything else . Fat is not really a large part of my diet really, as I go for high/er fat rather than high fat in conjunction with low carb. Having said that, I do have a serious cheese addiction, and when on a cheese fest then I really do have to factor in the fat :D

Great site, BTW ! Very , very interesting :)

Signy
 
  • Like
Reactions: 7 people

martykendall

Well-Known Member
Messages
56
Type of diabetes
Family member
Treatment type
Diet only
Hi Martin :)

I have the original TAG book, somewhere ! I'm not the tidiest of people and haven't as yet fully unpacked from my move last year :oops: Perhaps this issue will motivate me to get round to it :D When I excavate the info, I'll let you know ;)

You are right, original TAG didn't account for fibre, and therefore I never have as I count carb minus fibre as this is the way the value for most foodstuffs are given in the UK. It's true that the Sydney study showed no impact from fat, but I find it does affect me. Why, I don't know, but I'm guessing we are as individual in response to fat as anything else . Fat is not really a large part of my diet really, as I go for high/er fat rather than high fat in conjunction with low carb. Having said that, I do have a serious cheese addiction, and when on a cheese fest then I really do have to factor in the fat :D

Great site, BTW ! Very , very interesting :)

Signy

Thanks. Thanks. And thanks!!!

RE cheese, I would be interested to see how your glocse goes on high fat versus low fat cheese. See the list in https://optimisingnutrition.wordpress.com/2015/03/23/most-ketogenic-diet-foods/.
 
  • Like
Reactions: 2 people

Heathenlass

Well-Known Member
Messages
1,631
Type of diabetes
Type 1
Treatment type
Insulin
  • Like
Reactions: 11 people

Spiker

Well-Known Member
Messages
4,685
Type of diabetes
Type 1
Treatment type
Pump
Low fat cheese, low fat dairy product of any type, is an abomination against nature. :)
 
  • Like
Reactions: 8 people

martykendall

Well-Known Member
Messages
56
Type of diabetes
Family member
Treatment type
Diet only
Sorry I offended you with mention of low fat cheese. :)

Maybe I should have suggested comparing the blood sugar reaction to high fat cheese versus higher fat cheese?

I appreciate the religious fervour in respect to fatty cheese!
 
  • Like
Reactions: 9 people

zand

Master
Messages
10,784
Type of diabetes
Type 2
Treatment type
Diet only
@martykendall One of the many mistakes I made in the past was drinking far too much diet coke when I was trying to lose weight. I have read several articles that suggest because the body thinks diet drinks are sweet, the pancreas releases insulin which has nothing to work on because there is no sugar in diet coke. This therefore eventually leads to insulin resistance. It is what appears to have happened in my case.

So my question is do you have an insulin index figure for diet coke to prove this? Or are you going to tell me the theory is wrong and I would have become insulin resistant anyway?
 
  • Like
Reactions: 4 people

martykendall

Well-Known Member
Messages
56
Type of diabetes
Family member
Treatment type
Diet only
@martykendallSo my question is do you have an insulin index figure for diet coke to prove this? Or are you going to tell me the theory is wrong and I would have become insulin resistant anyway?

Good point Zand. There are lots of things that influence blood sugars and insulin. Food is just one of them.

My wife has seen big improvements since dropping Pepsi Max which she was addicted to.

Seems that these diet drinks have a major influence on the gut microbiome which in turn has a big influence on insulin sensitivity.

Hard to go wrong if you go for nutrient dense, high fibre, whole foods without bar codes. Interestingly enough its these foods that also have a low insulin index. See https://optimisingnutrition.wordpre...ood-sugar-regulation-and-nutritoinal-ketosis/

Cheers

Marty Kendalll
 
  • Like
Reactions: 8 people

zand

Master
Messages
10,784
Type of diabetes
Type 2
Treatment type
Diet only
Good point Zand. There are lots of things that influence blood sugars and insulin. Food is just one of them.

My wife has seen big improvements since dropping Pepsi Max which she was addicted to.

Seems that these diet drinks have a major influence on the gut microbiome which in turn has a big influence on insulin sensitivity.

Hard to go wrong if you go for nutrient dense, high fibre, whole foods without bar codes. Interestingly enough its these foods that also have a low insulin index. See https://optimisingnutrition.wordpre...ood-sugar-regulation-and-nutritoinal-ketosis/

Cheers

Marty Kendalll

Thank you @martykendall Your index is very interesting. I see it to be most useful for type 1s and also for type 2s who are stuck with weight loss. For some T2's weight loss is plain sailing once their BG's are under control. My BG's are great with 80g carbs daily, but to lose weight I need to drop to <30g daily. For the moment this is working fine, but should my weight loss come to a standstill again then I will have a look at your index to see how I can improve things.

Thank you for your work on this.
 
Last edited by a moderator:
  • Like
Reactions: 6 people

martykendall

Well-Known Member
Messages
56
Type of diabetes
Family member
Treatment type
Diet only
Thank you @martykendall Your index is very interesting. I see it to be most useful for type 1s and also for type 2s who are stuck with weight loss. For some T2's weight loss is plain sailing once their BG's are under control. My BG's are great with 80g carbs daily, but to lose weight I need to drop to <30g daily. For the moment this is working fine, but should my weight loss come to a standstill again then I will have a look at your index to see how I can improve things.

Thank you for your work on this.

Thanks Zand.

I have some other thoughts on weight loss, rather than just reducing the insulin index at

https://optimisingnutrition.wordpress.com/2015/03/22/weight-loss/

https://optimisingnutrition.wordpress.com/2015/03/28/optimal-foods-for-weight-loss/

https://optimisingnutrition.wordpress.com/?p=1756

This approach emphasises reducing calorie density and increasing nutrient density rather than just reducing carbs / insulin load.

I would be interested in your thoughts.

Cheers

Marty Kendall
 

Spiker

Well-Known Member
Messages
4,685
Type of diabetes
Type 1
Treatment type
Pump
How did you do the analysis to backfit the U Sydney data to your equation i = c + 0.54p?

Do you have a scatter plot or something for all the data points, and then you fitted a curve to it? I have to say I am very sceptical that such a simple mathematical relationship exists. Though I would love to be wrong! :)

You didn't do something like take an average did you?

Really what you need to do with any predictive formula is back test it against all the empirical data in the U of Sydney study and demonstrate you are replicating the empirical data within some reasonably narrow confidence level.

Otherwise, worst case, if it's just an average across the whole data set, it will only work to predict the insulin load of an actual meal if that meal is an blend of all 211 foods in the study, or a statically valid sub sample of them.
 

Spiker

Well-Known Member
Messages
4,685
Type of diabetes
Type 1
Treatment type
Pump
Ok I see in your manifesto where you do fit to a scatter diagram and get an improving R^2 numbers as you add in net carbs and protein.

For me I need to see some more heavyweight stats tests applied to your data because to be honest, by eye I don't see the scatter plot tightening that much around the trend line. As a T1, importantly, I could not afford to dose insulin with that much uncertainty.

Having said that, if FII is correct, I already am! :)

You say it's "conventional wisdom" to dose about 50% for protein. I would say that is wrong, there is no conventional wisdom. There are very few people dosing for protein and almost no agreement as to how to do it. I will tell you for a fact that your 0.54 ratio will not work for me and will make me go hypo. I know this from repeated experience.

You say that the basis of the conventional wisdom is unclear. I can tell you where it comes from: Dr Bernstein. Start by reading his book. Actually don't. I read 3 editions of his book cover to cover trying to accurately summarise what he says about protein dosing. It was inconsistent. And he is the authoritative source for low carb T1s. He cites a number of 58%, similar to your 0.54. He does not cite any paper. You cite a number of 0.54 from a paper of 1920. Before insulin was isolated, incidentally.

I contacted Bernstein directly and he did not resolve the question directly. Except he did. Because he gave the best and correct answer which is "it depends". Never good when a lawyer says that and much worse when your doctor tells you that. But it depends.

Apart from anything else it is completely unknown how much protein will enter gluconeogenesis and be converted to glucose. (I take it you know that's why protein causes an insulin effect.) It totally varies from individual to individual and from day to day in the same individual. This is because protein is also used structurally and that value can be estimated but not known. And a variety of different inconsistent formulas exist even for estimating it. So an unknown variable part of the protein will not undergo GNG.
. GNG is also a variable process that is down regulated by insulin. Another unknown.

So to summarise I will say this is what I believe about protein dosing:

- varies by individual
- varies within individual over time
- absolute maximum GNG of 58%
- carb diet raises insulin and suppresses GNG
- protein diet to a lesser extent probably suppresses GNG (self - suppresses)
- base protein demand for structural protein needs to be discounted from GNG
- base protein demand varies widely and is hard to estimate or measure with any confidence
- in practice, ignore protein unless low carbing
- test and titrate to find your personal protein ratio just as you would test and titrate to find your other insulin ratios
- be observant for ratio changes, as you would be with other insulin ratios
 
  • Like
Reactions: 4 people

martykendall

Well-Known Member
Messages
56
Type of diabetes
Family member
Treatment type
Diet only
How did you do the analysis to backfit the U Sydney data to your equation i = c + 0.54p?

Do you have a scatter plot or something for all the data points, and then you fitted a curve to it? I have to say I am very sceptical that such a simple mathematical relationship exists. Though I would love to be wrong! :)

You didn't do something like take an average did you?

Really what you need to do with any predictive formula is back test it against all the empirical data in the U of Sydney study and demonstrate you are replicating the empirical data within some reasonably narrow confidence level.

Otherwise, worst case, if it's just an average across the whole data set, it will only work to predict the insulin load of an actual meal if that meal is an blend of all 211 foods in the study, or a statically valid sub sample of them.

Hey Spiker

Thanks for taking a look in more detail. The formula is Insulin = carbs + 0.54P - fibre.

The fibre seems pretty clear cut. It's simply not digestible and hence doesn't impact insulin or glucose.

The protein allowance is subject to a bunch of variables. I could argue it's 54% based on the precendent of Wilders' formula, 60% based on the actual data, or 80% based on the glugenic protential of amino acids versus ketogenic (see http://en.wikipedia.org/wiki/Glucogenic_amino_acid).

In reality the glucogenic potential of protein is going to depend on a while range of factors including how much protein and carbs you're eating, whether you're eating more or less calories than you need an whether you're excercising a lot or a little.

I would prefer to be on the conservative side - to allow something rather than nothing but not too much so you're going to overdose.

Using this approach to help with refining insulin dose is one thing. But I think it becomes more useful when you use it to help you select foods as well. It helps you minimise your errors more - a la Berstein's law of small numbers.

If you're going to live on a high carb, low protein dietary approach then the insulin index probably isn't worth the hassle. However if you're going for a low carb, moderate protein, high fat approach then it will help refine your insulin dose for protein which is a significant proportion of the insulin demand in that paradigm.

Cheers

Marty
 
  • Like
Reactions: 4 people

martykendall

Well-Known Member
Messages
56
Type of diabetes
Family member
Treatment type
Diet only
Ok I see in your manifesto where you do fit to a scatter diagram and get an improving R^2 numbers as you add in net carbs and protein.

For me I need to see some more heavyweight stats tests applied to your data because to be honest, by eye I don't see the scatter plot tightening that much around the trend line. As a T1, importantly, I could not afford to dose insulin with that much uncertainty.

Having said that, if FII is correct, I already am! :)

You say it's "conventional wisdom" to dose about 50% for protein. I would say that is wrong, there is no conventional wisdom. There are very few people dosing for protein and almost no agreement as to how to do it. I will tell you for a fact that your 0.54 ratio will not work for me and will make me go hypo. I know this from repeated experience.

You say that the basis of the conventional wisdom is unclear. I can tell you where it comes from: Dr Bernstein. Start by reading his book. Actually don't. I read 3 editions of his book cover to cover trying to accurately summarise what he says about protein dosing. It was inconsistent. And he is the authoritative source for low carb T1s. He cites a number of 58%, similar to your 0.54. He does not cite any paper. You cite a number of 0.54 from a paper of 1920. Before insulin was isolated, incidentally.

I contacted Bernstein directly and he did not resolve the question directly. Except he did. Because he gave the best and correct answer which is "it depends". Never good when a lawyer says that and much worse when your doctor tells you that. But it depends.

Apart from anything else it is completely unknown how much protein will enter gluconeogenesis and be converted to glucose. (I take it you know that's why protein causes an insulin effect.) It totally varies from individual to individual and from day to day in the same individual. This is because protein is also used structurally and that value can be estimated but not known. And a variety of different inconsistent formulas exist even for estimating it. So an unknown variable part of the protein will not undergo GNG.
. GNG is also a variable process that is down regulated by insulin. Another unknown.

So to summarise I will say this is what I believe about protein dosing:

- varies by individual
- varies within individual over time
- absolute maximum GNG of 58%
- carb diet raises insulin and suppresses GNG
- protein diet to a lesser extent probably suppresses GNG (self - suppresses)
- base protein demand for structural protein needs to be discounted from GNG
- base protein demand varies widely and is hard to estimate or measure with any confidence
- in practice, ignore protein unless low carbing
- test and titrate to find your personal protein ratio just as you would test and titrate to find your other insulin ratios
- be observant for ratio changes, as you would be with other insulin ratios

Awesome, thanks for engaging Spiker!

I agree with most of what you're saying.

I think this is Bernstein's latest on dosing for protein -
. If you asked me to write down what this means next time you go to calculate a protein dose I'm not sure I could. What I get from this is that it's something and not nothing.

Recently I've been looking to understand the proportion of glucogenic amino acids versus the ketogenic amino acids. I just uploaded an updated version of the manifesto doc - https://www.dropbox.com/s/if9cs6u0achx4lj/Optimising nutrition, managing insulin.pdf?dl=0. It seems that up to about 80% of proteins could be converted to glucose if someone was very inactive and they were eating a lot of protein etc... So 54% is not a maximium it is more a typical average as shown by the FII data from The University of Sydney testing.

I have heard that insulin / carbs suppress GNG but I can't find a reference. By logical extension this means that excess protein is excreted and doesn't count as proteins. I'm not a biochemist but I've been trying to understand it. I would be interested if you could direct me to something that elucidates the issue. I have a separate blog post on the GNG issue - https://optimisingnutrition.wordpre...e-insulin-reaction-to-protein-dose-dependent/. Seems to be an area worth more investigation.

> Base protein demand for structural protein needs to be discounted from GNG.

Yes, but I think you've got to start at 80% glucose potential and work back from there. FII data suggest you get to about 60%. Some Type 1s dose for 100% of their protein as carbs!

> Base protein demand varies widely and is hard to estimate or measure with any confidence.

Yes. But are you 100% confident in protein. I suppose it's a little but more predictable, but the body's reaction to carbs depends on a range of factors too, but everyone has accepted carb counting as standard fare. In the end a type 1 is going to need to use correcting insulin to some extent.

> In practice, ignore protein unless low carbing.

Agreed. The insulin from the carbs will be more than enough to deal with. Also the protein is typically lower and slower so less of an influence. So not really worth the effort. Considering protein is a consideration for type 1s on low carb and also a growing number of non type 1s who are trying for ketosis rather than Atkins or LCHF. If insulin is the master hormone of the metabolism, then the insulin index is potentially a more refined tool to manage insulin than carb counting alone.

> Test and titrate to find your personal protein ratio just as you would test and titrate to find your other insulin ratios.

Definitely, but the insulin index gives a starting point for that and a basis for that starting point. I have been searing for that for a while but haven't come across it, until now.

Some people will go to the extra effort to refine and test to find their own factors, most people won't though.

Thanks for your time! I do really appreciate your feedback. I will feed some more of these thoughts back into refinement of the blog and the manifesto!

Cheers

Marty
 
  • Like
Reactions: 3 people

Spiker

Well-Known Member
Messages
4,685
Type of diabetes
Type 1
Treatment type
Pump
The formula is Insulin = carbs + 0.54P - fibre.

The fibre seems pretty clear cut. It's simply not digestible and hence doesn't impact insulin or glucose.
Outside of the US and some other countries this can be ignored as fibre is already excluded from the carb count in the nutrition information. Are you sure that the Sydney data hadn't already excluded the fibre from their data?
 
Last edited by a moderator:

Spiker

Well-Known Member
Messages
4,685
Type of diabetes
Type 1
Treatment type
Pump
In reality the glucogenic potential of protein is going to depend on a while range of factors including how much protein and carbs you're eating, whether you're eating more or less calories than you need an whether you're excercising a lot or a little.

I would prefer to be on the conservative side - to allow something rather than nothing but not too much so you're going to overdose.

Me too, and in my experience your suggested ratio of 54% is much too high to be conservative. I will hypo when low carbing if I use a ratio above 33%. I suggest a conservative number to start with is 25%, test and titrate from there.

54% is also implying a degree of precision that simply doesn't exist given the level of variation. Given all the uncertainties, both theoretical and practical, I think you should use a round number, say 50% or 60% - though as I say I think those are too high.

Whether the maximum value is 54% or 58%, there's no way you can say it's "conservative" to use a maximum value that would never be reached even if a person ate nothing but pure protein (since even then not all the protein would enter GNG).

I've never heard this 80% number quoted. I would dismiss it as an outlier, rather than using it to justify well referenced experimental reports and well analysed metabolic sequences of the theoretical maximum as being "middle" values.[/QUOTE]
 

Spiker

Well-Known Member
Messages
4,685
Type of diabetes
Type 1
Treatment type
Pump
I think this is Bernstein's latest on dosing for protein -
. If you asked me to write down what this means next time you go to calculate a protein dose I'm not sure I could. What I get from this is that it's something and not nothing.
Yes as I said, he is unfortunately pretty vague about it, and no one else is speaking with any authority about it. So like you have done, I had to go to the underlying research papers. I then (as Dr Bernstein has always done) had to use myself as the experiment. So I have empirical data from a real human. There are a u other humans in this small data set, enough to convince me that effective protein ratios vary in T1 LCHF people, from below 25% to as high as 60%.
 

Spiker

Well-Known Member
Messages
4,685
Type of diabetes
Type 1
Treatment type
Pump
> Base protein demand for structural protein needs to be discounted from GNG.

Yes, but I think you've got to start at 80% glucose potential and work back from there. FII data suggest you get to about 60%. Some Type 1s dose for 100% of their protein as carbs!
I really disagree with this suggestion there is anything like an 80% maximum protein conversion potential. I don't see any evidence for it and frankly I think it is a dangerous claim to make unless you are really convinced about there being solid evidence. Two points on this.

1. Yes there are variations in the GNG yield of different amino acids. But we are not talking about huge variations. More to the point we do not consume single amino acids, not even athletes using specialised supplements. We are always consuming a broad mix of amino acid types in our protein, particularly natural protein. So at best we are "tinkering at the margins" if we try to shift our amino acid profile.

2. You cannot just look at a theoretical chemical yield of glucose from amino acids / protein. Only a limited number of metabolic pathways are available. Very large numbers of theoretically possible reaction sequences are simply not available in our biology. It's not useful to look at what a chemist in a lab might do to extract glucose from amino acids in a lab (even restricting the options to exothermic reaction sequences not endothermic). GNG is the only game in town, and only conversion ratio that matters is the GNG ratio. Be that 54% or 58%.

Whoever is reporting stable blood sugar with a 100% protein to carb ratio, I would immediately question if their basal rate and carb ratios are correct. I would verify that by doing systematic variation tests in their bolus, for example doubling the protein in a meal while holding the carbs constant, and vice versa. I expect you would see a hypo in the first test and probably a hyper in the second test, or alternatively you might discover they are inadvertently using an excess basal rate / dose to cover their typical boluses.