Estimated reading time: 7 minutes

There is a lot of confusion about the term ketosis among medical professionals as well as laypeople. It is important to understand when and why nutritional ketosis occurs, and why it should not be confused with the metabolic disorder we call ketoacidosis.


Ketosis is a metabolic state where the liver produces small organic molecules called ketone bodies.

Most cells in the body can use ketone bodies as a source of energy. When there is a limited supply of external energy sources, such as during prolonged fasting or carbohydrate restriction, ketone bodies can provide energy for most organs.

In this situation, ketosis can be regarded as a reasonable, adaptive physiologic response that is essential for life, enabling us to survive periods of famine.

Nutritional ketosis should not be confused with ketoacidosis, a metabolic condition where the blood becomes acidic as a result of the accumulation of ketone bodies.

Ketoacidosis can have serious consequences and may need urgent medical treatment. The most common forms are diabetic ketoacidosis and alcoholic ketoacidosis.

What Is Ketosis?

The human body can be regarded as a biologic machine. Machines need energy to operate. Some use gasoline, others use electricity, and some use other power resources.

Glucose is the primary fuel for most cells and organs in the body. To obtain energy, cells must take up glucose from the blood. Once glucose enters the cells, a series of metabolic reactions break it down into carbon dioxide and water, releasing energy in the process.

The body has an ability to store excess glucose in the form of glycogen. In this way, energy can be stored for later use. Glycogen consists of long chains of glucose molecules and is primarily found in the liver and skeletal muscle. Liver glycogen stores are used to maintain normal levels of glucose in the blood while muscle glycogen stores are mainly used to fuel muscle activity.

Carbohydrates are the body’s primary source of glucose and other sugar molecules such as fructose and galactose. During carbohydrate restriction, both proteins and fats can be used for energy. In fact, most cells can use fatty acids for energy, but brain cells and developing red cells are more dependent on glucose for energy supply. However, brain cells can adapt and use ketones from fat breakdown for its energy needs.

When no carbohydrate is available, the liver will not break down fat completely. Instead, it produces ketone bodies that are used by most cells to provide energy. When ketone bodies are produced more quickly than the body needs, ketone levels build up in the blood, resulting in a condition known as ketosis. Ketosis is most commonly caused by very low carbohydrate consumption or prolonged fasting.

When the body’s glycogen stores become depleted, the breakdown of body fat (mainly triglycerides) results in increased availability of fatty acids. Most cells can use fatty acids for energy production. However, many fatty acids can not pass the blood-brain barrier. Therefore, the brain becomes dependent on ketone bodies produced by the liver.

The breakdown of fatty acids results in the production of an important substance called acetyl CoA. When fat and carbohydrate metabolism is in the balance, most acetyl CoA enters the so-called citric acid cycle (Krebs cycle) where it is used for energy production. When acetyl CoA cannot enter the citric acid cycle, it is shunted to form ketone bodies. This process is called ketogenesis.

Ketone bodies include three compounds: acetone, acetoacetate, and beta-hydroxybutyrate. Acetone can sometimes be smelled from the breaths of people with high levels of ketone bodies in the blood. You may be acquainted with the smell because some nail polish removers contain acetone.

Ketone bodies are not only produced when the glycogen stores become depleted. In fact, ketone bodies are produced by the liver all the time. Research indicates that the heart and kidneys prefer to use ketone bodies rather than glucose as a fuel resource. To dispose of excess ketone bodies, the body uses the kidneys to excrete them in urine, and they are exhaled from the lungs. During ketosis, ketones can easily be detected in the urine.

Nutritional Ketosis

When daily carbohydrate consumption is restricted to below 60 grams per day, the body usually shifts into ketosis and blood levels of ketone bodies start to rise. At this level of consumption, blood levels of ketone bodies may rise above 0.5 mmol/L, which is ten times higher than among individuals who consume 300 grams of carbohydrate per day. With lower consumption blood levels of ketone bodies rise further, and may reach 3 mmol/L.

The hepatic generation of ketone bodies is the normal physiologic response to fasting. Mild ketosis (ketone body concentration of about 1 mmol/L) develops after a 12- to 14-hour fast. If fasting continues, ketone body concentration continues to rise and peaks at a concentration of 8 to 10 mmol/L. Beta-hydroxybutyrate is the major ketone body that accumulates.

This normal physiological response to varying degree of carbohydrate restriction or fasting is termed nutritional ketosis.

This process is dependent on the ability of the pancreas to produce insulin. If insulin production is limited, the acid-base balance of the body may become interrupted.


It is important to differentiate between ketoacidosis and ketosis.

When ketoacidosis is present, the blood becomes acidic (pH level drops). Increased blood acidity can have serious consequences, cause coma and eventually death if untreated. However, although ketone bodies are acidic, the body usually manages to keep the acidity of the blood within normal limits during carbohydrate restriction or prolonged fasting.

Unlike ketoacidosis, nutritional ketosis is by definition a benign metabolic state, enabling our body to react to a shortage of dietary fuels.

The most common causes of ketoacidosis are diabetic ketoacidosis and alcoholic ketoacidosis. The former is primarily associated with type 1 diabetes, due to a shortage of insulin. The latter occurs primarily among alcoholics.

Although nutritional ketosis is considered safe, it is important to realize that people who are unable to secrete basal insulin, such as type 1 diabetics and some type 2 diabetics, are at more risk of entering unsafe levels of ketosis that may require emergency medical treatment.

Ketogenic Diet

Ketogenic diets emphasize foods rich in natural fats and adequate in protein and restrict foods high in carbohydrate. While the standard American diet contains 45-65% of calories from carbohydrate, ketogenic diets usually restrict carbohydrate intake to about 2-4% of calories.

A typical ketogenic meal includes a small amount of protein, a source of natural fats (for example, butter, beef tallow, lard, duck fat, cream, olive oil, or coconut oil) and some green leafy vegetables.
10283570_mFor many of us, a ketogenic diet defies everything we’ve learned about diet and nutrition through the years. That’s why it may be a bit hard for many to accept. Coffee with butter, bacon, steak, cream and fat cheese. Very little fruit. Many will say that such a diet defies common sense. But the fact is that ketogenic diets have been tried and tested, and there is no clear evidence of harm.

Ketogenic diets have been used to treat epilepsy in children. There is some evidence that adults with epilepsy may benefit from such a diet as well. In the 1920’s ketogenic diets were popular due to their efficacy in childhood epilepsy. However, anti-epileptic drugs have largely replaced ketogenic diets as a treatment for epilepsy.

Ketogenic diets usually induce weight loss. Insulin sensitivity is improved, and there may be some other metabolic advances as well. Therefore, low-carbohydrate diets and periodic fasting have become very popular worldwide. Studies indicate that low-carbohydrate diets induce more weight loss, lower triglycerides and raise HDL-cholesterol (the good cholesterol), compared with low-fat diets.

There is some evidence that ketogenic diets may affect the growth of cancer cells and improve the quality of life in cancer patients. However, it is premature to make any recommendations based on current research.

Ketosis Symptoms

Although nutritional ketosis is an adaptive response triggered by lack of external energy resources, it may initially be associated with unpleasant physical symptoms.

Fasting and carbohydrate restriction leading to ketosis can cause headache, nausea, fatigue, dry mouth, bad breath, upset stomach, nausea, frequent urination, and lack of mental clarity.

The symptoms associated with ketosis are most often temporary and will usually pass in a couple of days. Therefore, they are sometimes called  “ketosis flu.” Some of these symptoms are caused by dehydration, and therefore intake of salt and water may be helpful.

25 thoughts on “Ketosis”

  1. The article feels heavily biased towards nutritional ketosis being abnormal rather than either a valid alternative to glucose or superior. It also remains very vague around the difference between nutritional ketosis and ketoacidosis.

    For those who produce healthy levels of insulin, ketones will rarely exceed 3 mmol/L, and almost never 4. Ketoacidosis is usually considered an issue around 15 mmol/L and above. The mechanism lies with the way fats are stored and transported, i.e. triglycerides. When these are metabolized, they form free fatty acids, ketones, and glycerol. Glycerol is converted to glucose. As more trigs are used glycerol levels rise which will trigger more insulin as they are converted to glucose. The presence of insulin inhibits triglycerides from being used as fuel until any circulating glucose is removed/burnt.

    This very elegant feedback mechanism automatically maintains ketones at a healthy and non toxic level. For those perfectly keto-adapted, where the brain is using some 90% plus ketones as fuel, we can see that the glucose produced from glycerol is sufficient to fill the remaining need for glucose to the brain. Note also that such a keto adapted state results in significant physiological insulin resistance to most cells in the body except the brain. This is essential so that all glucose produced is directed to the brain while all other cells use fats and/or ketones.

    A better approach would be to consider glucose a toxin outside of a very narrow range and it is given priority over fats as an urgent necessity to remove it and minimize its damage. The perception is that glucose is the normal and preferred fuel, when it is the entire opposite. We need very little glucose for optimum health and when the elegance of the triglyceride mechanism is understood then true priorities become much clearer.

    • Thanks for the comment Cris.

      If there is bias, it was unintentional, maybe because of my “upbringing”. In fact one of the main purposes of my article was to highlight that nutritional ketosis is a completely normal physiological phenomenon. Of course, whether carbs or fat is the preferred fuel is not for me to judge. In fact, the body’s ability to adapt to different conditions indicates a certain flexibility when it comes to choosing energy resources.

      Thanks for drawing our attention to some of the important feedback mechansims involved in ketosis and acid – base regulation.

      For those who want to dig deeper into the biochemistry it is worth pointing out that there are some other mechanisms involved as well. It is worth mentioning that at a plasma ketone body concentration of 8 to 10 mmol/L, the rate of ketone body synthesis in the liver matches the rate of ketone body utilization. In addition there is a slight ketone body loss into the urine. The net effect is that the plasma bicarbonate concentration usually falls to a certain degree. The fall in plasma bicarbonate is often less than the rise in plasma ketone body concentration due to at least two factors: firstly acetone is not an acid and therefore does not affect the plasma bicarbonate, and secondly some of the hydrogen ions in the ketoacids are buffered in the cells instead of the extracellular fluid.

      The fact that ketone body generation stabilizes at a certain level is due to a few factors. One is a “braking” effect of the ketone bodies on the release of fatty acids from fat tissue, which is mediated by at least three mechanisms: stimulation of insulin release (despite low glucose levels), increased sensitivity of fat tissue to insulin’s inhibitory effect on fatty acid release, and direct inhibition of lipolysis (breakdown of fat) by the ketone bodies themselves. Furthermore, there is an increased rate of central nervous system ketoacid uptake when the brain switches from glucose to ketones as its source of fuel. Finally, there is increased ketone utilization in peripheral tissue.

      Main references:
      Fasting ketosis and alcoholic ketosis,
      Ketone body metabolism and cardiovascular dsiease

    • I really didn’t appreciate any bias. I felt it very evenly covered the topic and didn’t subvert nutritional ketosis at all.


    In 2003, the authors reported that severely obese adults lost more weight and had better serum lipid patterns after 6 months of a low-carbohydrate diet rather than a conventional low-fat diet.

    After 1 year, these same patients still had more favorable triglyceride and high-density lipoprotein cholesterol levels on the low-carbohydrate diet than on the conventional diet. However, weight loss and the other metabolic parameters were similar in the 2 diet groups.

    The effect of the modest improvements in high-density lipoprotein cholesterol and triglyceride levels on the development of diabetes and cardiovascular disease is unknown.

  3. Cris:

    “The article feels heavily biased towards nutritional ketosis being abnormal rather than either a valid alternative to glucose or superior.”

    I agree: “abnormal” is a wrong word here. “Adaptation” is better. But as a permanent diet, a ketogenic diet is quite restrictive (needlessly so) and there’s really no need for an “alternative to glucose” in the sense of being in ketosis.

    “A better approach would be to consider glucose a toxin outside of a very narrow range and it is given priority over fats as an urgent necessity to remove it and minimize its damage. The perception is that glucose is the normal and preferred fuel, when it is the entire opposite. We need very little glucose for optimum health and when the elegance of the triglyceride mechanism is understood then true priorities become much clearer.”

    Once again, I agree partially and disagree partially. There certainly is no “preferred fuel” as a whole as the human body is quite adapt in utilizing both fat and carbs. However, there’s nothing (apart from anecdotes) to indicate that glucose is a toxin as such, unless you define “very narrow range” in a way that has very little significance in human physiology and life.

  4. Doc,

    this recent study looked directly to the arteries of the children treated with low-carb diets. Didn’t look too promising. The controls were on the typical American diet. In fact, this paper will add to tremendous amount of data showing that artery clogging saturated fats….clogs arteries, surprise. Although, not a perfect randomized control trial, like the upcoming studies done on PCSKY9 antibodies, we should not have problems with the results given the Darwinian foundation of our biomedical research paradigm (Stamler 2010).

    The impact of the ketogenic diet on arterial morphology and endothelial function in children and young adults with epilepsy: A case-control study.

    Results: The population study was composed by 43 epilepsy patients (23 males), aged between 19 months and 31 years (mean 11 years). Twenty-three patients were or had been treated with ketogenic diet, and 20 had never been on it (control group). Subjects treated with the ketogenic diet had higher arterial stiffness parameters, including AIx and β-index and higher serum levels of cholesterol or triglycerides compared to those who had never been on the diet (control group) (p<0.001).

    Arterial stiffness is increased in children and young adults treated with the ketogenic diet, before the increase of the intima media thickness. This supports that arterial stiffness is an early marker of vascular damage.

    • I think it should be noted that this study recruited people who HAD BEEN or currently were on a ketogenic diet. Furthermore, there was no control over what the studied group actually ate because the study was done on their current states rather than giving them all an identical structured diet over time. One doctor’s “ketogenic-friendly” food list doesn’t necessarily match another doctor’s. I find the results suspect due to too many unknowns in what the study group actually had been eating.

  5. Richard, the study you mention had a very, very small number of participants and there was a LOT of variation in their ages (19 months – 31 years). This alone effectively renders the study more or less meaningless.

  6. Mie,

    I disagree. I think you are being overly skeptical. All participants in the ketosis arm were or had been treated with ketosis. The results were most likely mediated via the cholesterol mechanism. The adolescents who were or had been treated with ketosis had less desirable blood lipids. Again, the results were perfectly in line given the preponderance of evidence. Nearly all long-term low-carbers who have had their lipid panel taken know that their LDL (and/or LDL-P) is at the high end. This is uncontroversial. Besides, I don’t think that studies should be looked in isolation, but instead trying to see how what’s their role in the larger picture.

    • I’m wondering about vitamin C intake, given its importance to vascular health and the inability of humans to synthesize it in-house.

      People on very strict ketogenic diets may well have deficiencies in vitamin C and other antioxidants, especially if their fats are coming predominantly from CAFO animals. The lipid profiles of grass-fed and pastured animals are very different to those of the agribusiness products. There appears to be adequate vitamin C in raw meat, but cooking the meat destroys the vitamin C. Fat is not a good source of vitamin C. Isn’t there some evidence that the much-studied traditional Maasai warrior group had high levels of endothelial fat streaks and even atherosclerosis, but that their arteries enlarged with age? Traditionally, the cow milk was mixed with blood, and blood is a good source of vitamin C. The warrior group’s diet was not restricted to milk and blood; they also ate varying amounts of plant foods. Exercise alone may not explain the contradictory aspects of the Maasai findings.

      I found no indication of whether the study group was assessed for vitamin C status, or was adequately supplemented during the trials. I found one study where a 9-year-old epileptic girl on a ketogenic diet developed scurvy, so vitamin C status seems worthy of consideration.

    Interview With Jerry Stamler, MD

    DR. MOSER: Why does the public so readily accept very low-carb or Atkins diets when data suggest that that may not be the best way to lose weight and reduce the risk of heart disease?

    DR. STAMLER: An essential fact about diet is: If you reduce calories, you reduce weight. So it is with Atkins-type diets that reduce carb intake. But long-term success, taking weight off and keeping it off? And long-term health effects? Unknown—likely adverse. No population in the world of any size, under any more or less ordinary circumstances (leaving out special limited and peculiar circumstances, e.g., small Inuit groups above the Arctic Circle) has ever subsisted for any length of time on a nutritional pattern such as recommended by Atkins originally or recently—under pressure from criticisms—with modifications. It is not possible to regard this diet as optimal in caloric density, ratio of nutrients to calories, or in key essential nutrients (minerals, vitamins, antioxidants, fiber) that—as we learn more and more—are protective. This diet also tends to be, unless you pick your fats very carefully and knowledgeably, excessive in saturated fats and cholesterol—dietary items that we have known for years are harmful. It’s a bill of goods sold to the public without decent long-term scientific assessment.
    Effect of a high-fat ketogenic diet on plasma levels of lipids, lipoproteins, and apolipoproteins in children.

    “Design, Setting, and Patients
    A 6-month prospective cohort study of 141 children (mean [SD] age, 5.2 [3.8] years for 70 boys and 6.1 [4.4] years for 71 girls) with difficult-to-treat seizures who were hospitalized for initiation of a high-fat ketogenic diet and followed up as outpatients. This cohort constituted a subgroup of the 371 patients accepted into the ketogenic diet program between 1994 and 2001. A subset of the cohort was also studied after 12 (n = 59) and 24 (n = 27) months.

    A ketogenic diet consisting of a high ratio of fat to carbohydrate and protein combined (4:1 [n = 102], 3.5:1 [n = 7], or 3:1 [n = 32]). After diet initiation, the calories and ratio were adjusted to maintain ideal body weight for height and maximal urinary ketosis for seizure control.

    Main Outcome Measures
    Differences at baseline and 6-month follow-up for levels of total, VLDL, LDL, HDL, and non-HDL cholesterol; triglycerides; total apoB; and apoA-I.

    At 6 months, the high-fat ketogenic diet significantly increased the mean plasma levels of total (58 mg/dL [1.50 mmol/L]), LDL (50 mg/dL [1.30 mmol/L]), VLDL (8 mg/dL [0.21 mmol/L]), and non-HDL cholesterol (63 mg/dL [1.63 mmol/L]) (P<.001 vs baseline for each); triglycerides (58 mg/dL [0.66 mmol/L]) (P<.001); and total apoB (49 mg/dL) (P<.001). Mean HDL cholesterol decreased significantly (P<.001), although apoA-I increased (4 mg/dL) (P = .23). Significant but less marked changes persisted in children observed after 12 and 24 months.

    A high-fat ketogenic diet produced significant increases in the atherogenic apoB–containing lipoproteins and a decrease in the antiatherogenic HDL cholesterol. Further studies are necessary to determine if such a diet adversely affects endothelial vascular function and promotes inflammation and formation of atherosclerotic lesions.

    • We have just “known” for “years” that these things are harmful. Guess what? What you thought you knew was completely wrong. You’ve been duped, Dr, Stamler. Read a book that isn’t more than 30 years old, maybe.

  8. The length of the treatment before the study is not the key point here: the key points are the study design (a case-control study), the heterogeneity of the participants and the small n. At best, this kind of study can be somewhat suggestive of need for further research.

    Just like I mentioned above, I don’t see the point in staying in ketosis, intentional/nutritional or not, for prolonged periods of time. However, I don’t think it reasonable to use very poor “evidence” to argue for or against it, either. If you want to let your standards slide, be my guest. I’d rather not.

    • Not all ketogenic diets are the same. Early versions for epileptics were dominated by vegetable oils, high in omega-6, and we know the oxidation effects here on lipids can be devastating. Trans-fats and other hydrogenated fats were also used. In current results using grass fed animal protein and fats, coconut oil, etc, all show healthy lipid profiles. In the high carb world we might compare an all candy bar diet with one dominated by fruits and veggies – all carbs, right – does that mean a fair comparison.

  9. Mie,

    again, I am not sure what is your problem. The study looked at the vessels of the neck of a group adolescents who either were being treated or had been treated with ketosis and published the results. They did a comparison to a group of adolescents who have not been treated with ketosis. The sample was more than adequate for a study like this. Why are the results so hard for you to accept given the preponderance of evidence? What do you suggest instead, that people who have been fed with high-fats diets, possible containing more saturated animal fats, do not face adverse prognosis for developing CHD at later in life? It’s seems it’s too easy to be overly skeptical and in denial to the preponderance of evidence (Stamler 2010). Hard to claim that ketosis have not been shown to do harm when you actually have a study that looked directly to the arteries of young people having been or being treated with ketosis. No surprises with the results provided.

  10. Cris:

    “Early versions for epileptics were dominated by vegetable oils, high in omega-6, and we know the oxidation effects here on lipids can be devastating.”

    Depending on the intake level and the context, quite possibly yes.

    “In current results using grass fed animal protein and fats, coconut oil, etc, all show healthy lipid profiles. In the high carb world we might compare an all candy bar diet with one dominated by fruits and veggies – all carbs, right – does that mean a fair comparison.”

    There’s very little evidence (apart from a handful of risk marker studies) to suggest that grass feeding makes any real difference. Or that coconut oil and other hyped alternatives are superior.

    And the comparison too is a BIT far-fetched, don’t you think? 🙂


    I already explained what the problem is. The fact remains that this matter requires more high quality epidemiological data and/or RCT data, as this study is woefully inadequate – and referring to evidence NOT directly related to ketogenic diets is of little use here. In addition, I already stated why I’m skeptical of ketogenic diets: their extreme nature doesn’t exactly sound promising in terms of adherence and compliance in the long term. I see no need for such a drastic measures.

    Unless Cris has evidence indicating otherwise.

  11. Mie,

    I hear you very well. However, I see this more straightforward. Ketogenic diet or any diet containing high amounts of fat, especially saturated animal fats tend to increase LDL (and LDL-P) in nearly all people. There exists convincing evidence that elevated LDL alone promotes CHD on nearly every mammalian independent of the mechanism used.

    The strengths of the study that I referred to is that it tested a hypothesis on a sample of young people. The results are not skewed by attempts to loose weight, cardiovascular drugs or any underlying chronic disease affecting lipid metabolism. These are issues that make the interpretation of the results more difficult if the underlying study population consist mainly of obese, middle-aged people in standard care. These problems often persists in lifestyle medicine even if superior study methodology is used.

  12. Interesting findings on Nicoyan longevity diet (
    “The limited available information on frequency of food consumption suggests a
    prosaic Nicoya diet that does not fit clearly into any of the popular theories about
    diets conducive to living long, such us the Mediterranean diet (Willett 2001), the
    calorie-restriction diet (Walford andWalford 2005), the Okinawan diet (Hiroko 2001)
    or the low animal protein diet (Campbell and Campbell II 2005). The data do not
    show that Nicoyans eat more of reputedly healthy foods like salads, fruits, olive
    oil or nuts, nor that they eat less fat, animal protein and sugars among other foods
    associated with worse health. Compared to other Costa Ricans, they take up slightly
    but significantly more calories, proteins and carbohydrates (although with higher
    fibre content and lower glycemic index) as part of a plain diet that includes rice,
    beans, fish, beef and chicken. However, the association of a lower glycemic index
    diet in Nicoya is consistent with the most recent nutritional evidence suggesting the
    primary importance of this as a determinant of cardiovascular health (Jenkins et al.

  13. I am failing to completely understand some of the theory here.
    Suppose a person is on a calorie-balanced low carb diet, say 40% protein, 60% fat. Calorie consumption is the same as calorie expenditure.

    Is all the fat being utilized for energy? Is this fat energy utilization via ketosis or is there some other metabolic pathway? If there is, is there something special about ketosis compared to this alternate pathway, and is it better, worse, or no different? If no different, why do people seem so concerned with ketosis?

    If there is no alternate pathway for the conversion of fat to energy, then in a balanced diet, this would mean that the fat is handled via ketosis. Or does ketosis simply mean that most of the energy comes via fat processed into ketones etc. What about excess protein being converted to glycogen, if there are no carbohydrates being consumed?

  14. Greetings I am so grateful I found your webpage, I really found you by error, while I was loking on Google for something else, Anyhow I am here now and would just like to say thanks for a marvelous post and a all round exciting blog (I
    also love the theme/design), I don’t have time to read through it all at the minute but I have
    bookmarked it and also included your RSS feeds, so hen I have time I will bee
    back to read more, Please do keep up the awesome work.

  15. Magnificent beat ! I wish to apprentice even as you
    amend your website, how can i subscribe for a weblog web site?
    The account aided me a acceptable deal. I have been a
    little bit familiar of this your broadcast provided bright transparent idea


Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.