Learn more about Peter Attia’s digital program
You may have heard from your doctor that ketosis is a life-threatening condition. If so, your doctor is confusing diabetic ketoacidosis (DKA) with nutritional ketosis, or keto-adaptation.
First, some semantics. Our body can produce, from fat and some amino acids, three ketone bodies (a “ketone” refers to the chemical structure where oxygen is double-bonded to carbon sandwiched between at least 2 other carbons). These ketone bodies we produce are: acetone, acetoacetone, and beta-hydroxybutyrate (B-OHB). [For anyone who is interested, they are the 3 most right structures on the figure, below.]
Why do we make ketones? For starters, it’s a vital evolutionary advantage. Our brain can only function with glucose and ketones. Since we can’t store more than about 24 hours’ worth of glucose, we would all die of hypoglycemia if ever forced to fast for more than a day. Fortunately, our liver can take fat and select amino acids (the building blocks of proteins) and turn them into ketones, first and foremost to feed our brains. Hence, our body’s ability to produce ketones is required for basic survival.
What is diabetic ketoacidosis? When diabetics (usually Type I diabetics, but sometimes this occurs in very late-stage, insulin-dependent, Type II diabetics) fail to receive enough insulin, they go into an effective state of starvation. While they may have all the glucose in the world in their bloodstream, without insulin, they can’t get any into their cells. Hence, they are effectively going into starvation. The body does what it would do in anyone – it starts to make ketones out of fat and proteins. Here’s the problem: the diabetic patient in this case can’t produce any insulin, so there is no feedback loop and they continue to produce more and more ketones without stopping. By the time ketone levels (specifically, beta-hydroxybutyrate) approach 15 to 25 mM, the resulting pH imbalance leads to profound metabolic derangement and the patient is critically ill.
But this state of metabolic derangement is not actually possible in a person who can produce insulin, even in small amounts. The reason is that a feedback loop prevents the ketone level from getting high enough to cause the change in pH that leads to the cascade of bad problems. A person who is said to be “keto-adapted,” or in a state of nutritional ketosis, generally has beta-hydroxybutyrate levels between about 0.5 and 3.0 mM. This is far less than the levels required to cause harm through acid-base abnormalities.
Keto-adaption is a state, achieved through significant reduction of carbohydrate intake (typically to less than 50 grams per day) and moderate protein, where the body changes from relying on glycogen as its main source of energy to relying on fat. Specifically, the brain shifts from being primarily dependent on glucose, to being primarily dependent on beta-hydroxybutyrate. This has nothing to do with what a diabetic patient is experiencing in DKA, but does illustrate how poorly informed and quick to react the medical community is. DKA and nutritional ketosis (or keto-adaptation) have as much in common as a house fire and a fireplace.
Photo by Andrew Yardley on Unsplash
What would you say to the idea that the body does not have 1 unified energy economy and that certain systems prefer to use fat for fuel (skeletal muscles) and others prefer glucose (the brain) and that ketosis causes a negative hormonal reactions inside the body and that a diet that provides adequate carbohydrate supports the thyroid? Additionally I’ve read that glucose provides cleaner fuel for the body – and that a diet balanced in carbohydrate protein and fat that provides adequate metabolic cofactors is a better choice than entering a protein and fat staggered diet that forces one into a ketogenic state? I believe the real problem is PUFA and not carbohydrate .
I agree there are likely different “ideal” fuels for different organs. The brain, for example, even in complete starvation when serum ketones exceed serum glucose will derive 1/3 to 1/2 of its energy from glucose and the remaining 2/3 to 1/2 from ketones. Heart and skeletal muscles, under aerobic conditions seem to prefer BHB, then FFA, then glucose. Of course, under aerobic conditions glucose is necessary.
I am not aware of data supporting the glucose-burns-cleaner argument. If anything, more glucose means more AGE, which is far from clean.
You were very kind in your response to this. Based on everything you’ve explained until now, this hypothesis about glucose has nothing to do with carbohydrate intake, as the liver will make glucose from protein as needed. So protein and fat is needed to manufacture both ketones and glucose for the appropriate organs, but carbohydrate is not needed even in this hypothesis.
Peter,
Any sources you might recommend for reading about AGE and fat? There are some materials out there that say fatty animal products are high in AGE when cooked. (for example: https://www.ncbi.nlm.nih.gov/pubmed/20497781).
Caitlin
This may be true for oils, especially PUFA when cooked at high temp. This is why many recommend coconut or palm oil for cooking (SFA –> less chance to violate a carbon-carbon bond).
I would be interested in your take on the Paleo thing of current almost mania crescendo. I think it offers some good points and benefit but see a lot of hype and what seems misinformation from various outlets and sources. Many ascertions that simply eating meat and cutting out white carbs will instantly transform one to a state of health/fitness. Would like your take on the current phenom. I practice some of the concept but am not a full bore Paleo geek ? I have discussed this with others as well like Gary Collins and coming to my own conclusions. Your input would be appreciated.
Simple question, why does it take such a low (50g) amount of carbs to stay in NK ? I would think that once you deplete your carb stores and are cruising at .5 or greater ketones, that you could eat maybe 50g of carbs per meal (x 3 per day) and if you are reasonably active, wipe that out and stay where you are ketosis-wise?
I like to think in terms of calories, so say for lunch you have the sandwich with two 100 Cal slices of bread, and the rest protein/fat (and that’s your typical macro/nutrient ratio x 3). At rest, just sitting around figure about 100 Cals/hour (about 50% fat/50% carbs burned). So, in a typical day (just counting waking hours), your gonna wipe out that 600 cals from carbs (16 hours x 100 cals /2 = 800). And if you have much activity at all, (i.e. – A few hours of walking around, housework, etc – I’d consider that you HAVE to have destroyed that carb intake…)
So once you are already carb depleted and in ketosis, why wouldn’t you stay there on moderate carbs, say 150g/day or so (and not have to go as low as 50g carbs a day to maintain the > .5 nutritional ketosis) ?
I’m just wondering what the mechanism is…
See post on carbs and ketones co-existing.
Dr. Attia,
Could you please respond to this article reporting on a non-diabetic woman who entered ketoacidosis that the doctors attributed to ketogenic diet?
https://www.nejm.org/doi/full/10.1056/NEJMc052709#article
I am on a forum for a medical condition that I have, and another member has posted this along with the assertion that people run the risk of death from ketogenic diet. The poster also says that there are studies showing ketogenic elevates inflammation, and the study she sites says it elevates all cholesterol and heart disease! I am not familiar with any of that. I would like to know how to respond to these assertions.
Thank you very much for your site – I have learned a tremendous amount.
Seems odd. I’ve read one other case study on ketoacidosis in someone who was neither T1D/T2D *or* on a carb-restricted diet. In this case, I wonder what triggered the event? Perhaps an underlying metabolic stress, such as a viral illness?
I don’t know what triggered it, but I, too, suspect that more is going on there than the average, sane person who might attempt a ketogenic diet. For instance, it appears from the article that there were multiple events and that the lady wasn’t entirely forthcoming with her doctors about her diet. Once she finally “came clean”, what she revealed seems like a pretty extreme carb restricted for a very long time . It also appears that she wasn’t monitoring her ketone levels, or would have seen the problem coming. Finally, the fact that they had to give her supplemental insulin to reign the episodes in would surely mean she had some kind of insulin dysregulation, wouldn’t it? Not being a doctor myself, I can’t really evaluate those things. What do you think?
Mainly I just PLEASE need to know what to say on my forum! I definitely want to defend the ketogenic diet, which is helping me immensely. Based on this article should I concede that it IS possible to go into ketoacidosis through ketogenic diet? If so, how is one to know in advance that they are that rare individual for whom it is not safe to even try? I’m rather baffled.
I have been in ketosis for quit a while. Short periods I lost it as social circumstances let me eat more carbs than I wished for. I went into this ketogenic life style because I had been diagnosed with type 2 diabetes at the age of 54 after a life long of eating everything or anything, staying slim. I started with Dr Cousen’s raw diet to reverse my diabetes, but found it hard to keep up in Spain (where I live) and changed after about 1 month and a half to the LCHF diet. I feel great during ketosis without using medicines (metformina) but noticed that when out of that for just a short period I regained some to all the symptoms of my diabetes.
Does the ketogenic diet finally reverses diabetes type 2 like Dr Cousens claims with his Life Food vegan raw diet?
Yes.
Great, thank you, I understood so much but like to hear it from an expert. Can you tell how long is the time it takes to reverse this disease? I have been on a ketogenic lifestyle now over 1 ½ year. I was a professional golf teacher and do still a lot of golf and cycle 3 to 4 times a week down the mountain I live on and back up (prox. 15 min. over 8 km down and up with an elevation of 450 meters).
Hi,
My daughter, age nine, has been placed on a modified Atkins diet for epilepsy. Her carbohydrate count is limited to 15 a day. She has been on this diet for about a week now and seems to be doing very well. I have started using Morton’s Lite Salt for added potassium in the foods that I cook. I would like to know if this is safe for her as well as my other children, ages six and two? It was emphasized to me that I needed to make sure she was getting adequate electrolytes, especially potassium, and most foods high in potassium, except avocados, are two high in carbohydrates. I know there is a danger with too much potassium and wondering if lite salt would cause this. For example I made sausage balls with one 2lb ground beef and added 2 tsp. of lite salt. This was divided between 5 people, the two adults consuming the most, and eaten throughout the day. There is about 350 mg. of potassium in 1/4 tsp. of lite salt. How much is too much for children under age 10? I have always been on the low end of potassium and have been prescribed potassium pills on three different occasions so I know what low potassium feels like and I don’t want that to happen to her, however, I don’t want to over do it either. It is amazing to see improvements in just one week. I have seen drastic improvement in her handwriting which I didn’t expect. We have always been a whole food eating family without processed sugars so eating this way wasn’t a huge jump for her but I am shocked at the improvements already. Great website by the way.
Jennifer, have you been in touch with Charlie Foundation? https://www.charliefoundation.org/
All of your questions will be answered. Yes, it’s remarkable when I child responds so well to something so “simple”…
Thank you for your great blog. I have been trying to find answer for KCA versus ketosis. I do however have a question I have not been able to find a satisfactory answer to. I am a type 1 diabetic. I have maintained an AIC of below 6 for over 15 years. I am active and exercise regularly . I have been in KCA twice thru the years. At the time, as aforementioned, my sugars were good and I was physically active. My question is what other reason could that have happened? The answers I have received were late stage and poor control.
Secondly, what reading should I be looking for while using keto sticks.
Thank you!
Julie
DKA in T1D (well-controlled T1D, which it sounds like you are) is often precipitated by a systemic stress, such as infection. While not always the case, there is usually something that gets the patient out of sorts.
Hi Peter,
Im very interested in going ketogenic at the moment im eating 50g carbs, 140g protein & 170g fat, however im 165lbs and do crossfit type workouts and Olympic weightlifting. How will this make me perform better and would it?? Also I would appreciate if you could let me know about fat to protein ratio? This is all self experiment and also to move away from eating carbs.
Hope to hearcfrom you soon
Mike.S
Sorry . DKA
Hi Peter,
I would have a question and would be happy if you could give me an answer please. I am on low carb diet for about a week (carbs intake as low aprox 25g a day, only from vegetables). During that week I had 3 high intensity trainings. I am meassuring my ketone levels via urine tests. Are these reliable? It showed me ketones in urine on 5th day of LCHF, now after 8 days it shows 8 mmol/l. Does this say something? Should I increase carbs intake from 25 to about 50? Not sure how exact and reliable these urine tests are.
Many thanks in advance for your answer and also for this excellent page, which is the main source of information on LCHF for me.
Greetings from Slovakia.
Tomas
Urine cannot accurately predict/reflect serum ketone level.
Peter,
I have been on a ketogenic diet for about a year, <20g of carbs, approx. 50g protein, and the rest fat in the form of nuts and coconut oil. No dairy or grains either. I'm 22 years old, 130 lbs, 5ft 3in, and 19%-20% body fat. I workout for about 45min everyday doing weightlifting and interval sprints 1min:1min. What I would like to hear your thoughts on is why my ketones are consistently higher than average. My average ketone level has recently been 5mM, but can easily get to 8mM, and even to the point where my Nova Max meter just flashes "high". My average fasting CBG is usually somewhere around 70 mmol/L, but occasionally it can be in the 80's or 90's and still have a ketone level over 4mM. Why do you think this is?
Thanks Alot,
Rebakah
Hi,
I was on ketosis for a long time (about 1-2 years). Eventually, I got really sensitive to carb. Sugar from 1 apple would put me in a food coma (basically I fall asleep) afterwards. I was worried about insulin insensitivity. Eventually I got off the low carb diet and explore other options. I do miss being on the low carb diet and thinking about going back on it again after reading your article.
I’m just worried that the extreme sensitivity to carb was a bad sign. What would you recommend to be the right balance for this? Perhaps carb re-introduction is necessary?
Thank you,
Gift
hi
I have blood ketones between 4.7-5.9 but don’t seem to be dropping the weight.
What am I doing wrong.
My macros are 85% fat, 10% protein, 5% carbs.
Any advice would be appreciated.
I am new to the keto lifestyle been doing it for 6- 8 weeks.
> … but don’t seem to be dropping the weight.
You might work down the checklist at:
https://www.wheatbellyblog.com/2012/10/i-lost-the-wheat-but-didnt-lose-weight-2/
and watch the 3 related videos that Dr. Davis posted to his blog in Feb 2014.
You already have items #1 and 2 nailed, so it’s apt to be something further down.
I’m assuming you have item “0” nailed as well (eliminate gluten-bearing grains 100%). FYI, Wheat Belly isn’t just about wheat, and it recommends a macronutrient balance that is only borderline keto.
Sarina,
I had a similar experience after I first found this site awhile back, and maintained high ketones but no fat loss.
I finally figured out that I was too enthusiastic about eating very high fat — the ketones were coming more from the fat I was eating, than from the fat I wanted to lose. I cut back some, while still staying high-fat in a percentage sense, and was never hungry but the weight started to come off. It just took experimentation to find my best balance. Ketones stayed high all along, but now were coming more from my own fat.
I never counted anything, just played with it until it started to work. Peter has a “fat flux” article that addresses some of that.
Hope this is of use to you.
Bob.
Hi, Dr. Attia! First, I highly respect you and your research. Thank you for all that you are doing! I am just graduating with a degree in nutrition from University of Houston. I want to be a dietitian. I have learned so much from you among others like Gary Taubes, Dr. Jeff Volek, and Dr. Stephen Phinney. It is really sad that all this valuable science you talk of in your articles and lectures are not even mentioned in my nutrition degree program at Uh. Actually I did take a class called research in obesity and the professor had us read “Good Calories Bad Calories” which is what got me researching all this. But the professor didn’t talk much about the science the book was based on. Anyways, I wanted to ask you: my mom has late stage type 2 diabetes where I believe her pancreas burned out (she was on long acting insulin but now on both long and short acting). Would you think it wise to put a late stage type 2 diabetic on a ketogenic diet as long as insulin and ketone levels are watched? Would ketones even need to be watched if blood glucose was kept within normal range. I know you cannot give medical advice online, but theoretically is there a proper protocol for late stage type 2 diabetics on a ketogenic diet? If you cannot answer in a comment I completely understand.
Thank you,
Jerry
Hi,
I am a dietetic intern going through my rotation, I have had a lot of preceptors tell me that ketosis could still cause ketoacidosis, and they told me to go to our hospitals UpToDate website and search for journal articles showing ketoacidosis in some patients. Is this really a possibility? was I looking at bad studies? What would be a great comeback?
Hello,
I’ve stumbled across a comment here about fasting blood glucose rising on low-carb/keto diets; the response was that you hope to write about this topic soon and that was back in 2012. I couldn’t find it by doing a regular search, so I’m wondering if you’ve written about this topic since? I’ve been experimenting with the ketogenic diet since January 2014 (which a Paleo-friendly cheat day once a week) and recently got a fasting blood glucose reading of 109 which really freaked me out, hence my deep investigating and questions now. I’ve found evidence that this is okay, and is apparently even expected on low-carb diets, but I haven’t found anything that makes me feel really confident that I’m not doing any long-term harm on my body. Any information would be greatly appreciated! Thanks so much in advance,
Lynette
Peter at Hyperlipid wrote something about this. I think I’ve referenced it a few times in comments, but his blog will be quickest way to find it.
> Peter at Hyperlipid wrote something about this.
Hyperlipid is a bit clumsy to navigate. Might it be:
“Physiological insulin resistance”
at:
https://high-fat-nutrition.blogspot.com/2007/10/physiological-insulin-resistance.html
… or not, that was 7 years ago.
I’ve seen a number of sources (e.g. Jaminets/PHD) claim that FBG creep occurs on VLC or keto diets. The Jaminets advance the solution of increasing consumption of what they call “safe starches” (which is not the same as resistant starches, which they also endorse).
The questions that arise are:
1. Is VLC/keto FBG creep real?
2. Is it universal, or just affecting some people?
3. What is the cause?
4. What is the (safe) fix?
Regarding “universal”, assuming the effect is real, I’ve been closely following the Wheat Belly Blog since its incep. WB advocates a diet that I characterize as VLC/borderline-keto. That blog seems to hear from everyone who doesn’t get textbook results, but no one has to my recollection complained about FBG creep. I’d don’t measure my own BG [yet], so have no personal anecdote to contribute here.
Peter,
My daughter has type 1 and now has been diagnosed with epilepsy. I thought her seizures were from hypoglycemia,now shes on yucky epilepsy meds. My gut tells me that the seizures are related to her diabetes and nutrition. The endo and neuro docs don’t think so.
Any studies or docs that I can read up on to help coach her?
Not sure of link between T1D and epilepsy, though, as you probably know, KD is treatment of choice for drug-resistant seizures in children.
Reta,
The book that got me looking at Ketosis in general might be interesting to you:
https://www.drperlmutter.com/about/grain-brain-by-david-perlmutter/
I’m not a scientist, nor a doctor, nor any sort of an expert in any way, shape or form. But, the book points towards inflammation as a potential cause for things like that. Specifically, that book is pretty anti-gluten as it apparently causes inflammation in a lot of people (even if they don’t test positive for being a celiac). Anyway, take it or leave it :).
Dose 5.3 on my sd freecodec means i’m in ketosiss ?