I keep seeing the idea that higher fat, lower protein for people on a ketogenic diet will yield higher ketones. (This just came up in the context of someone needing higher ketones for therapeutic reasons. I know that it’s often not useful to chase ketones, but in the therapeutic context, it might be better to have higher ketones.)
Are there any studies on this? Even a single person, with a cross-over, would be good. I can find no studies whatsoever.
PS – maybe I should do one? High protein, lower fat one week; then high fat, low protein one week. Not sure I’d need a washout period. My only problem is that the pin-prick glucose and BHB meters have so much error that I’m concerned there wouldn’t be much to find. That’s why I haven’t done this study in the past.
Edit: another reason I haven’t done this: I’d have to weigh everything and take food data. I don’t like doing that.
I think the concept is probably that protein can derail ketosis by being converted to glucose, so less protein = better results. But IIR, we debunked the idea of that, beyond what glucose the body actually needs. More protein does not equal more glucose, ergo lower protein / more fat wouldn’t be any more beneficial.
@kib1: But the old therapeutic keto had high fat, low protein and it yielded higher ketones and those helped with epilepsy if I remember correctly, what about that? And carnivore is considered not ketogenic due to the protein or something…?
I am unsure as it doesn’t matter to me, I am physically unable to eat moderate protein, just super low and high and that works me well enough. But I still like to know things, I just don’t work for them if they aren’t relevant or extreme interesting to me.
I recall reading research at some point debunking the myth that protein above some level interrupts ketosis (in a carb-restricted context). The idea was interesting, but not based on the realities of actual human metabolism.
I don’t have any such studies stashed away in my files so I can’t be of more help to @ctviggen in his request.
Not eating carbs forces a body to produce ketones (and GNG for glucose) to stay alive. The idea of trying to elevate ketones further than what a healthy body is producing (putting aside therapeutic needs in the context of disease) is a fool’s errand anyhow.
Dr. Benjamin Bikman - ‘Insulin vs. Glucagon: The relevance of dietary protein’
There’s about 8.5 minutes of good stuff about protein and ketosis, starting just before 24:00.
The entire 36 minute video is really good - one of the best I’ve ever seen. Bikman’s very solid point is that the insulin to glucagon ratio makes an enormous difference with respect to several things, including ketone production, and that a person well-established on a keto diet will react much differently to protein than one coming from a “standard American diet” or the like.
I think the bottom line is that if one has relatively low insulin and relatively high glucagon (as with eating very low-carbohydrate), then protein is not to be feared at all, as far as lowering ketone production.
@OldDoug That is one of the first I saw for the idea of protein not causing higher glucose. Protein causes insulin to go up, which one would think would cause blood sugar to crash. However, then glucagon rises, meaning that blood sugar is raised.
I’ve even gotten both insulin and glucagon tested twice. The problem is that both are pulsatile, so looking at them at only a few times isn’t meaningful. You get a picture, but of what, I’m not sure. For instance, here are my two glucagon results:
The same insulin (and c-peptide) results:
I have no idea why my insulin was so high, as it always used to run around 10 for a while. I hypothesize it’s due to a drug I had two switch to, but there’s no way to really know (I can’t go off the drug to test).
I started taking the drug sometime before March 2023, and I didn’t change what I was eating. So, gaining 0.4 in HbA1c in about 6 months with no dietary changes seems like a drug issue.
But let’s ignore the possible effects of this drug. I’ve had people on Reddit tell me they’ve been keto for 8 years (I’m on year 11), and yet they can get ketones of 5-6 mmol/l eating higher fat. Meanwhile, I get 0.1 or 0.2 mmol/l in the mornings and I’ll be lucky to get >0.5 at night. While I eat a higher protein diet, it seems unreasonable for me to be able to go from 0.1 mmol/l or even 0.5 to 5 mmol/l. The last time I fasted 4.5 days, I only got to 2mmol/l.
If there are two people who have been keto for about the same time (assume 8 years = 11 years), why would one get 5mmol/l ketones and another get 0.1 mmol/l? I can’t see protein making that difference.
Maybe I’ll have to buy some ketone strips and do a two week test? High protein one week (my normal diet) and high fat the next week. My guess is that there could be a slight increase in ketones, but nowhere near what people are suggesting.
Edit: I’ve ordered ketone and glucose strips. I’ll test this.
The test idea sounds good, at least you will be the same person! There must be huge personal differences or else we hardly would get these wildly different results for different people…
I’d say to be really sure, the individual would need a continuous glucose monitor, and see what happens when they eat a substantial amount of protein.
To me, the 800 lb gorilla (do metric system users have a set number of kilograms in a similar statement?) in the room is whether or not one has been eating very low-carb.
If one is low-carb like that, then I’ve never even seen an apparently rational proposal for why higher protein would necessarily mean meaningfully lower ketones.
I’m testing this for myself right now. On high fat low protein, I’m getting very high breath acetone and low blood glucose. Wondering what it will do for BUN, creatinine, and other markers which were very high last time, when I was (over?)eating leaner meat. Do you have a baseline for those numbers on your high protein?
Measuring fat content of meat is hard, though. For untrimmed beef brisket, no way Chronometer’s “visible fat eaten” accounts for a 3 inch fat cap! How do you track this?
Bikman says protein does spike insulin on SAD but doesn’t on low carb diet. “The need for the liver to create glucose determines the I:G response to protein”
I watched the video, I think I watched a similar one before… I don’t worry about insulin due to high protein since many years but I still remember reading many times that therapeutic keto with high ketones requires higher fat… Oh well, it doesn’t matter to me personally, I am just curious.
Very true. I am aware I simply can’t track fatty meat. I do try and give some educated guess but I know very well that I may be very off, the fattier the meat, the more. I vaguely know my preferred fat content for fried/roasted pork and it helps a bit but if I eat something way more fatty, it’s a mystery and anyway, certain cuts aren’t comparable to others. Visible fat (not like I could guess that) doesn’t help as the pure meat may be very lean or fatty! And the fatty one doesn’t feel like some lean one with a fat cap even if the fat content is the same. So I am lost.
And if my meat has bones (and maybe skin), I should just give up. I cooked some duck wings and meaty beef bones today (some parts are for the cats, some for me and some extra fat is produced), I don’t even try to track that, it’s totally impossible.
I usually guess something for chicken thigh (it is very fatty in the beginning - but then it loses a lot of fat. I measured and now I just track half with skin, half without while eating it fully with skin. good enough, I can’t possibly do any better).
I often saw how extremely different fattiness two slabs of meat, same cut can have. And then I go and fry only the fattiest pieces of the already fattier than usual green ham… Or the pork shoulder loses a lot of lard while cooking… The air fryer is especially good at that even if the meat isn’t super fatty. I can weigh the lard but what for if I don’t know the starting fattiness…?
So one just can’t track anywhere near accurately these sometimes. Or often, it depends on the food choices.
I don’t know of any studies but I can share a couple of real life examples.
My friend has bipolar 1, her mental health is better when she gets her ketones to around 4-6. I would never call her diet healthy and neither would she, but uncontrolled bipolar 1 symptoms are so awful and potentially dangerous she eats in a way that results in highish ketones: a pathetic amount of protein, a bit of raw cabbage, and a massive amount of mayonaise. If she ups her protein to a bit less pathetic (but still a fair bit under what is recommended here in terms of grams per kg) her ketones drop to 0.3-0.8. 0.8 is an absolutely fine number for Joe Average but not for someone needing a medical keto protocol.
She recently decided to eat higher protein because she wants to build some muscle (some cardio but mostly weights and resistance work). And (thankfully!!!) she got her protein from my freezer (80/20 ground beef) and not from the crap sources she usually does. She removed the source of the very few carbs she was eating (the cabbage) and just ate beef and a ton of mayo. Her protein intake was just under the low end of what we recommend here. Yes her ketones plummeted but she felt great in other ways. Way more energy and sleeping way better.
So what she’s decided to do is protein cycle. Medical keto interspersed with protein feeding periods.
I did carni for 14 months (may 2022-july 2023) and ate 1.5-2kg of meat a day (6 foot 3 sedentary 62 year old woman). I only ate the fat that came with the meat but also ate full fat cheddar and unflavoured unsweetened greek yoghurt and cream in my coffee. I lost 19kg despite eating a ton of calories. I never had a blood ketone reading higher than 0.3, and it was usually 0.2.
I stopped for 5 months, and ate very cheap crap. Put on 8 kg and got full of inflammation again -> a lot of physical pain in muscles and joints. I restarted carni in feb this year. I didn’t eat anywhere near as much meat (300-500 grams a day) when I restarted because I just can’t afford it, which was a big reason I stopped carni last year. Out of interest I took a ketone reading a few weeks back and it was 0.8. I was astonished. I am definitely experiencing the reduced appetite phenomenon caused by being in (a decent amount of) ketosis but I was not eating enough to fully nourish myself. A recent 6 day hospital stay where I was nil by mouth and lost zero weight has shown me this. My body thinks there is a famine so I am now increasing my meat and cheese intake, but that’s an aside.
Sooooo, for me and for my friend, higher protein = much much lower ketone production. I tried the high fat, much lower protein protocol during my 1st carni stint but I hated how the food tasted. Apart from pork crackling fresh from the oven when doing a shoulder roast, I’m really not keen on the taste of animal fat, butter (ewwwww), ghee (double ewwww) etc. Cold pork fat when I eat the rest of the roast cold is ok too.
I don’t know of any studies but I think it’s dependent on the individual. Their metabolism and where they are in their diet.
I just followed the advice given from the carnivore doctors that I followed and the numerous articles I read on eating carnivore. By eating a ratio of approximately 70-80% fat and 20-30% protein I had very little problems with healing my body and losing weight and inches. In ten months I lost 55 pounds and 36.5 inches. Now maybe I just fit into the group that that works on or if that’s the norm, I don’t know but what I do know is it darn sure worked for me. I’m still eating the same way now that it looks like I’ve hit my optimal weight and I’m having no trouble maintaining it.
Well if we talk about fat-loss too, that’s even more individual, just like what fat percentage we can have for it to happen. I simply can’t lose fat eating really high fat (like 70-80%), it’s too much for me. My protein and energy need and plans and bodyfat just results in that. (I am a normal person and can’t lose fat when overeating and my energy need is smallish. Could be worse though.)
Whenever people talk about what fat percentage is best for keto, I don’t understand why they don’t address the relationship between energy and protein need. Whenever it’s about protein, there is a g/kg range but if one eats 1500 kcal vs 4000 kcal, the percentage will be highly different… If one needs more energy, they probably need more protein but it’s quite possible that energy need shoot up way more. I already have very varied days calorie wise and my protein need is about the same all the time. (My intake isn’t due to my tastes but it’s a single day for me, it wouldn’t be sustainable, I believe even I have a maximum limit for protein, it seems inevitable due to having a human body though some carnivores make me wonder about this… )
I have good macros when I eat somewhere about 60% fat. It depends but 65% never worked (it’s theoretically possible if I do everything right like keeping my protein to the minimum but of course I can’t do that every day. my minimum is already over 2g/kg for LBM but I noticed my body tends to accepts it but I can’t go lower).
I am glad I don’t have a special need for a specific percentage because I would be in trouble. My desires/tastes and needs already point at different percentages and I had to give up eating similarly every day, it just can’t work. But this way I can explore new percentages, not just the 65% fat I had for years (no matter how I tried to eat, no matter if I ate barely or a ton, it was very close to 65% virtually every day. well that happens if your staple food has this percentage, you do your best to avoid added fat and if you eat some leaner or fattier bites, you immediately feel the irresistible pull towards the other direction. I had to focus real hard to get my single 80% fat day out of curiosity. I never could go below 58-60% until recently though, I mean, not even for one day, it’s super lean to me and I couldn’t handle it. now that I have 90% days, I can do 50% too).
Another anecdote here: I noticed several years ago when I was still measuring blood glucose and ketones that when I ate less protein my morning blood sugar was lower and my ketones were higher. I really do think there is something to it. I’ve mentioned this before, but if it didn’t matter, why does Paleomedicina limit the amount of protein their patients can eat?
I think this is a valid reason to want higher ketones. Does she take MCT oils? Those should help (assuming her digestive system can handle them – start slow.)
Supposedly, higher PUFAs (polyunsaturated fatty acids) also can lead to higher ketones, so more pork and chicken is probably better than beef.
For her, it has to be a tough balancing act. If you’re older, you want more protein, to have stronger bones and muscles, but you don’t want bipolar episodes (or at least from what I hear, they are really bad).
Has she tried metformin? That was shown to reverse bipolar, but only in those where metformin worked. 20 people took metformin, but only 10 reversed their insulin resistance using it.
She also had a plan for people for whom metformin did not work.
@VirginiaEdie There could be something to it, but it’s hard to test. For instance, I have a picture somewhere where I have 3 ketone meters showing the following from a blood draw at about the same time: 0.2, 0.4 and 0.8 mmol/l. Which one is true? This is particularly problematic for me, where there’s a world of difference between 0.4 (could be there) and 0.8 (never there).
And then you look at the new continuous ketone meters and everything changes. This is from someone on Reddit:
Makes me wish I could get a CGM and CKM. Unfortunately, the best I can do is pin-prick Keto Mojo.
And this is really only important for those folks who actually need higher ketones.
Anyway, I bought some ketone strips and blood glucose strips. I’m going to try to go higher protein/lower fat next week, then very high fat the next week and see what happens. Not sure that 1 week for each is good enough, but it’s really all the time I can devote.
@VirginiaEdie and @beannoise Assuming that you had higher morning blood sugar and lower ketones with higher protein, what were the perceived detriments of this? In other words, were there changes on a monitor but no other changes you could discern? Or did you actually experience a detriment you could perceive?
There must be huge personal differences or else we hardly would get these wildly different results for different people…
)
