Protocol · graded by use, two uses
Ketogenic diet (very-low-carbohydrate)
Seizure frequency in children and adolescents with drug-resistant epilepsy.
Compare this record with another, side by side →
How this letter has changedLatest: October 2026 · Changed.
The headline showed C on longevity while the record's best use, seizure frequency in children with drug-resistant epilepsy (Neal et al., Lancet Neurol, 2008; PMID PubMed ID: the permanent number the US National Library of Medicine gives a published paper. Each one links to that paper's record on PubMed. 18456557; Lambrechts et al., Acta Neurol Scand, 2017; PMID 27027847; Cochrane, PMID 32588435), is B. The headline now shows B labeled with that use, and longevity and cardiometabolic markers keep C on their own row (Patikorn et al., BMC Med, 2023; PMID 37231411). No evidence changed. Headline policy adopted by the editor in October 2026 (Arnold Sandlow, D.C.): the headline letter is the record's best graded use, labeled with that use, and every other use, including any harm finding, is listed beneath it.PMID 18456557 PMID 27027847 PMID 32588435 PMID 37231411
The decision sheet asked whether drug-resistant epilepsy in children should be A. Two independent randomized trials (Neal et al., Lancet Neurol, 2008; PMID 18456557; Lambrechts et al., Acta Neurol Scand, 2017; PMID 27027847) and a Cochrane pooling (Martin-McGill et al., 2020; PMID 32588435) agree on fewer seizures, but seizure counts come from unblinded diaries kept by families, and the written A admits three endpoint The outcome a trial measures to decide whether the treatment worked, such as heart attacks, blood pressure or a symptom score. kinds only: a hard clinical event, an instrument reading, or a scale scored by a blinded clinician. A diary in an unblinded trial is none of these, so the use is B at the self-report ceiling. The headline stays C on the longevity question: a use in children with drug-resistant epilepsy does not lift the headline of a record read by adults asking about diet and aging, and the record says so. Applied the written grade definition under the editor's standing instruction of October 2026 (Arnold Sandlow, D.C.).PMID 18456557 PMID 27027847 PMID 32588435
The record carried one C across every use while its own umbrella review, Patikorn 2023 (PMID 37231411), rated seizure reduction as high-quality evidence. Drug-resistant epilepsy in children and adolescents is now its own use: Neal 2008 (PMID 18456557) and Lambrechts 2017 (PMID 27027847), two independent positive randomized trials, and Martin-McGill 2020 (PMID 32588435), the Cochrane pool; seizure diaries are unblinded self-report, so B. Schoeler 2023 (PMID 37977712), KIWE, is added so the use excludes infants and the comparison with a further drug. The LDL harm now cites Wang 2024 (PMID 39097343) and Chang 2026 (PMID 41621647), and Żurawski 2026 (PMID 42588148) states the matched-energy weight difference. The description names the epilepsy population. Headline letter unchanged; whether the epilepsy use should be A is on the decision sheet.PMID 37231411 PMID 18456557 PMID 27027847 PMID 32588435 PMID 37977712 PMID 39097343 PMID 41621647 PMID 42588148
Every change to a letter on this site is dated and published. The full evidence log →
A ketogenic diet is very low in carbohydrates. Its best-supported use is in children and teens with epilepsy that drugs have not controlled. In two trials by separate teams, the diet cut seizures compared with usual care. In infants, it did no better than another seizure medicine. In adults, trials found lower weight, blood fats and long-term blood sugar. No trial shows that it helps people live longer. It also raises LDL cholesterol.
The B belongs to one of the two uses graded on this record. The other one carries its own letter.
This record carries a separate letter for each thing it has been measured on. The headline grade is the best of them, and it does not transfer to the others.
Neal et al. (Lancet Neurol, 2008; PMID 18456557): 145 children aged 2 to 16, at least daily seizures and two failed drugs, randomized to the diet or a 3-month delay; seizures 75 percent lower than controls at 3 months. Lambrechts et al. (Acta Neurol Scand, 2017; PMID 27027847): independent Dutch trial, 57 children and adolescents, 50 percent responders against 18 percent with care as usual. Martin-McGill et al. (Cochrane, 2020; PMID 32588435): 13 studies, 932 participants, seizure freedom RR 3.16 and seizure reduction RR 5.80 in children, low to very low certainty for lack of blinding. Two independent positive randomized trials and a Cochrane pool on a clinical outcome: replicated The same finding reproduced by a separate research group. A result seen once can be chance or bias; a replicated one is harder to explain away.. Seizure counts come from unblinded family diaries, a self-reported endpoint, which caps the letter at B. In infants, Schoeler et al. (Lancet Neurol, 2023; PMID 37977712), KIWE, found the diet no better than a further antiseizure medicine, so this use is children and adolescents, not infants, and the comparison is against usual care, not against another drug.
Randomized trials replicate lower triglycerides, weight and HbA1c A blood test that reflects average blood sugar over the previous two to three months. (Patikorn et al., 2023; PMID 37231411; Wang et al., 2024; PMID 39097343) and also a clinically meaningful LDL cholesterol rise, 8.5 mg/dL across 62 trials (Chang et al., 2026; PMID 41621647); weight loss at matched energy is 1.5 kg more, low certainty (Zurawski and Bartosiewicz, 2026; PMID 42588148). No randomized trial reports a longevity or hard clinical outcome, and the mortality association is observational Research that watches what people already do rather than assigning a treatment. It can show that two things travel together, not that one causes the other.. Surrogate benefits offset by a quantified LDL rise: C.
Seizure frequency in children and adolescents with drug-resistant epilepsy (B), on two independent randomized trials and a Cochrane review. Longevity and cardiometabolic markers in adults are C.
Why this gradeThe reasoning, and what the grade does not cover.
Grade B, the record's best use: seizure frequency in children and adolescents with drug-resistant epilepsy, on two independent positive randomized trials and a Cochrane pool. Longevity and cardiometabolic markers in adults are graded C on their own row. Randomized trials replicate lower triglycerides, weight and HbA1c, and also a clinically meaningful LDL cholesterol rise (8.5 mg/dL across 62 trials); no randomized trial reports a longevity or hard clinical outcome, and the mortality association is observational. Drug-resistant epilepsy in children and adolescents is its own use at B: two independent positive randomized trials and a Cochrane pool on seizure frequency, a clinical outcome recorded in unblinded diaries, which caps it at B.
WHAT THIS GRADE DOES NOT COVERinfants, where the KIWE trial found no advantage over a further antiseizure medicine; comparison with another drug rather than usual care; psychosis and cancer, where trials exist and were not used; and any mortality claim.
A grade describes the human evidence for a named use. It is not advice, and it is not a judgment on any individual prescription. How the letters are decided →
The evidence, in fullEight papers, each linked to its PubMed record.
Patikorn et al. (BMC Medicine, 2023; PMID 37231411): umbrella review of 17 meta-analyses of 68 randomized trials; high-quality evidence for lower triglycerides, lower seizure frequency and higher LDL cholesterol, moderate for lower weight and HbA1c and higher total cholesterol; long-term clinical-outcome trials lacking. Wang et al. (Am J Clin Nutr, 2024; PMID 39097343): meta-analysis of 27 randomized trials, 1,278 participants; triglycerides, glucose, insulin, diastolic pressure and BMI lower; total cholesterol 0.36 mmol/L and LDL cholesterol 0.35 mmol/L higher. Chang et al. (Endocr Pract, 2026; PMID 41621647): meta-regression of 62 randomized trials; triglycerides 20 mg/dL lower, LDL cholesterol 8.49 mg/dL and total cholesterol 8.14 mg/dL higher; no long-term outcome data. Żurawski and Bartosiewicz (Nutrients, 2026; PMID 42588148): meta-analysis of 6 trials at matched prescribed energy, 259 participants; 1.49 kg more weight loss than higher-carbohydrate diets, low certainty. Epilepsy, graded as its own use: Neal et al. (Lancet Neurol, 2008; PMID 18456557): 145 children aged 2 to 16 with drug-resistant epilepsy randomized to the diet or a 3-month delay; seizures 75 percent lower than controls at 3 months, 38 percent of diet children with more than 50 percent reduction against 6 percent; unblinded, seizure diaries. Lambrechts et al. (Acta Neurol Scand, 2017; PMID 27027847): 57 children and adolescents randomized to the diet or care as usual; 50 percent against 18 percent responders at 4 months. Martin-McGill et al. (Cochrane Database Syst Rev, 2020; PMID 32588435): 13 studies, 932 participants; seizure freedom and seizure reduction favored ketogenic diets in children, all studies at high risk of performance and detection bias for lack of blinding, evidence low to very low certainty. Schoeler et al. (Lancet Neurol, 2023; PMID 37977712): KIWE, 136 infants aged 1 to 24 months randomized to the classic ketogenic diet or a further antiseizure medicine; seizures per day did not differ (incidence rate ratio 1.33, 95% CI 0.84 to 2.11); several authors report grants from medical-nutrition companies. Headline Grade B on the epilepsy use; the longevity and cardiometabolic use is C: surrogate benefits offset by a quantified LDL rise and no longevity endpoint.
Urinary or blood ketones (beta-hydroxybutyrate) daily for the first 4 weeks to confirm nutritional ketosis (0.5-3.0 mmol/L). Lipid panel including LDL particle number or ApoB Apolipoprotein B, a protein carried once on each cholesterol particle that can enter the artery wall, so it counts those particles directly. at baseline and 12 weeks: LDL elevation occurs in a subset of individuals and requires clinical attention. Lean mass by DEXA at baseline and 12 weeks; Protein intake is at least 1.6 g/kg to preserve muscle.
Evidence notes: randomized trial, meta-analysis, review
Endpoint graded: Seizure frequency in children and adolescents with drug-resistant epilepsy; longevity and cardiometabolic markers graded separately (surrogate)
Last reviewed: July 2026
Educational use only · Not medical advice · EBL grades published human evidence and does not recommend treatment. Stopping a prescription is a decision for your physician.
As studiedDose, form, population and duration for every use.
A grade describes the intervention that was tested. Where the product on the shelf differs from it, this says so.
Evidence Based Longevity, Inc. is owner-operated and edited by Arnold Sandlow, D.C., a California-licensed chiropractor. Evidence Based Longevity publishes in English only · Machine translations are not produced or reviewed by EBL for accuracy
© Evidence Based Longevity, Inc.. All rights reserved
These statements have not been evaluated by the Food and Drug Administration. Any grade on this page describes the strength of published human trial evidence in the populations studied. It is not a recommendation for any individual and is not medical advice. Starting, stopping or changing any supplement or medication is a decision for your physician.
How EBL searched for the trials (1 search)
The search that was run, what it returned, and what was left out and why, so the absence of evidence can be checked as well as its presence.
("ketogenic diet"[tiab]) AND (randomized controlled trial[pt] OR meta-analysis[pt]) AND humans[mesh]Grades move when new trials publish. A free monthly email lists every grade EBL moved, with the reason and the trial behind it: the sign-up is on the Evidence Log →