Ketogenic Diet

Fasting and the first dietary therapy

A historical review in Epilepsia states that fasting and other dietary regimens have been used to treat epilepsy since at least 500 BC. The ketogenic diet was introduced by modern physicians in the 1920s as a way to mimic the metabolism of fasting. A 2018 international consensus guideline and a 2026 review both give 1921 as the year the diet began to be used for epilepsy, and the guideline describes the classic form as in continuous use since then.

Decline and revival

According to the historical review, the diet was widely used for about two decades before the arrival of modern antiepileptic drugs, after which its use fell sharply. By the end of the twentieth century it was available in only a small number of children's hospitals. The review describes a rapid growth in both use and scientific interest over the following 15 years or so.

The 2018 consensus guideline notes that parent-created support groups helped drive that revival: The Charlie Foundation in the United States in 1994 and Matthew's Friends in the United Kingdom in 2004. The guideline says these groups raised awareness and helped treatment centers recruit patients and carry out research. The sources reviewed for this page do not name the individual physicians who first developed the diet, so this page does not attribute it to a person.

Spread beyond epilepsy

A 2026 review reports that publications on the diet have increased sharply in recent years as researchers look for uses outside epilepsy, and that most people who follow it today do so for weight loss, with figures of up to 77 percent cited. That shift from a hospital-supervised therapy to a widely adopted eating pattern is the reason this page separates the medical use from elective uses.

Names and variants

The sources use several terms. Ketogenic dietary therapy refers to the family of medical diets for epilepsy, and the 2018 guideline lists four major forms: the classic ketogenic diet, the medium-chain triglyceride diet, the modified Atkins diet and the low glycemic index treatment. A 2024 meta-analysis of diabetes trials uses the term very low carbohydrate ketogenic diet, abbreviated VLCKD. In popular use the shorthand is keto diet.

Ketosis and carbohydrate restriction

Ketosis refers to a metabolic state in which the body relies more on fat and on ketone bodies for energy because carbohydrate supply is restricted. The 2026 review describes the diet as a way of shifting energy supply from carbohydrate to fat metabolism, which is why people with disorders of fat metabolism are advised against it. A broader umbrella review treats ketogenic diets as one end of a range of low-carbohydrate diets, which differ in how much carbohydrate they allow.

Related Machipedia pages

The Ketogenic Diet belongs to the wider family described on the Dietary Approaches page. Related Machipedia pages include Low Carbohydrate Nutrition, which covers carbohydrate-restricted patterns in general, Intermittent Fasting, Nutritional Psychiatry, Dietetics, the Paleo Diet and the Carnivore Diet. The 2026 review notes that a carnivore-style ketogenic diet appears in a small case series, so the two diets can overlap in practice.

General features

In general terms the diet restricts carbohydrate heavily and relies on fat for most of its energy. The 2018 guideline states that the classic and medium-chain triglyceride forms are typically started in hospital by a dietitian and a neurologist, while the modified Atkins diet and the low glycemic index treatment are started as outpatient programs and need less precise weighing of food. This page does not give food quantities, ratios or meal plans, because the guideline treats these as matters for the treating team.

Clinical delivery for epilepsy

For epilepsy the guideline describes a pre-treatment clinic visit to identify seizure types, rule out metabolic disorders that are contraindications, and check for conditions such as kidney stones, swallowing difficulty or high cholesterol. It recommends laboratory screening before the diet starts and regular follow-up by both a dietitian and a neurologist familiar with the therapy, with visits during the first year at set intervals.

Training and professional standards

The sources reviewed describe dietitians as core members of ketogenic diet teams, and the field of Dietetics has its own professional bodies, registers and codes of practice. No source reviewed for this page describes a separate qualification, register or training standard specific to the ketogenic diet, and none was found. The guideline's description of the team is a statement by clinical experts, not a licensing requirement.

Established medical use

The established medical use is as a treatment for drug-resistant epilepsy, mainly in children. The 2018 guideline calls ketogenic dietary therapies well-established nonpharmacologic treatments for children and adults whose seizures do not respond to medication, and lists specific epilepsy conditions in which it suggests the diet be considered early.

Elective wellness and other uses

Outside epilepsy, people and promoters use the diet for weight loss, type 2 diabetes, cancer, athletic performance and a range of neurological and hormonal conditions. A 2026 review reports weight loss as the most common reason. These are claims about elective use, and the evidence for each is covered in the next section. Evidence from epilepsy treatment cannot be carried over to these uses.

Drug-resistant epilepsy

The 2020 Cochrane review included 13 trials with 932 participants, 711 children and 221 adults. All were rated at high risk of performance and detection bias because blinding was not possible, and the certainty of evidence for every outcome was low to very low. In children, seizure freedom favored the diet over usual care (risk ratio 3.16, four studies, 385 participants, very low certainty), as did a 50 percent or greater seizure reduction (risk ratio 5.80, low certainty). In adults the effect was uncertain.

The review concluded that the diets could be effective in children but that evidence in adults remained uncertain, and that the modified Atkins diet may have a similar effect with fewer adverse effects. It described the diets as a valid option for people with drug-resistant epilepsy who are unsuitable for surgery. The 2018 expert guideline recommends delivery by a specialist team.

Weight loss

A 2023 meta-analysis of 110 randomized trials of carbohydrate-restricted diets in adults with overweight or obesity found greater weight loss with lower carbohydrate intake, with high-certainty evidence at 6 months and moderate certainty at 12 months. Beyond 12 months the relationship was not linear. The trials covered carbohydrate restriction broadly, not only ketogenic diets.

A 2026 meta-analysis of trials comparing ketogenic with Mediterranean diets found greater short-term reductions in weight and body mass index with the ketogenic diet, with no significant differences in blood glucose, HbA1c or LDL cholesterol. The authors suggested the Mediterranean diet may remain more sustainable long term.

Type 2 diabetes

Reviews disagree. A 2023 meta-analysis of 11 randomized trials in adults with type 2 diabetes found no significant difference from control diets in glycemic control or body weight, though triglycerides fell and HDL rose, with low to moderate quality evidence. A 2024 meta-analysis of 29 trials of very low carbohydrate ketogenic diets, assessed with GRADE, reported lower HbA1c, fasting glucose and triglycerides, and no significant difference in LDL or total cholesterol.

Cancer, sport and other uses

A 2026 systematic review of the diet in malignant glioma included 23 studies and 306 patients, of which only two were randomized trials. Survival results were inconsistent, the randomized trials showed no significant differences in survival, and dropout was about 20 percent. The authors concluded that survival or quality-of-life benefits remain unproven. The umbrella review of low-carbohydrate diets likewise found limited evidence for adult cancer.

A 2026 meta-analysis of 13 studies in trained athletes found effects on anaerobic performance varied by domain: single maximal efforts were sometimes maintained, while repeated high-intensity efforts were more often impaired. The authors called their pooled estimates exploratory because durations ranged from 2 days to 12 weeks. Sources reviewed do not cover other neurological conditions.

Adverse effects

A 2026 review reports that adverse effects often appear in the first days and usually resolve within weeks. These include the cluster sometimes called keto flu, fatigue, nausea, headache, dizziness and constipation, with constipation reported in 1 to 68 percent of adults across studies. In children the 2018 guideline reports gastrointestinal effects in up to half, raised blood lipids in 14 to 59 percent of children on the classic diet, and kidney stones historically in 3 to 7 percent. The 2026 review cites 2.5 to 4 percent.

On growth, the guideline reports mixed data but that all six studies longer than 6 months showed negative effects of the classic diet on growth in children, with younger children affected more. It also reports that long-term complications in children on the diet beyond 2 years have not been reviewed systematically, and that long-term vascular outcomes of the diet are unknown.

Contraindications

The 2026 review lists absolute contraindications as rare disorders of fat metabolism, including primary carnitine deficiency and pyruvate carboxylase deficiency, and porphyria, because restricting carbohydrate could cause serious metabolic complications. It lists relative contraindications as acute pancreatitis, acute liver failure, advanced chronic kidney disease, propofol use and familial hypercholesterolemia. The 2018 guideline likewise advises screening to rule out metabolic disorders before epilepsy treatment.

Medicines, pregnancy and eating disorders

The 2026 review reports that people taking insulin or other glucose-lowering medicines may develop low blood sugar if doses are not adjusted. It also cites case reports of euglycemic diabetic ketoacidosis in people who combined a carbohydrate-restricted diet with SGLT2 inhibitors, including one after a single dose. It states that ketone bodies cross the placenta, that studies in pregnancy are too few to judge the diet's effect, and that more research is needed on breastfeeding.

For eating disorders, the same review notes that the diet suppresses appetite and can cause rapid early weight loss, which is undesirable in underweight people. It also mentions preliminary reports of benefit in anorexia nervosa, drawn from a case series of three adults and six prospective cases, which are too small to guide care. The review states the diet is not right for everyone and that risks should be assessed individually.

Care boundary

People with epilepsy, diabetes, cancer, kidney or liver disease, pregnancy or a history of disordered eating who are considering the diet are advised in the sources to do so only with a clinician. The diet should not delay or replace necessary medical care, and a change in medication should be made with the prescriber.