12 claims that cleared review. Each one shows what has to be true for it to hold and how strongly it is supported — including the places the sources contradict each other.
A very-low-calorie ketogenic diet (VLCKD), as used in medically supervised clinical protocols, combines severe caloric restriction (roughly 600-800 kcal/day) with a low carbohydrate intake (roughly 20-60 g/day, up to 30-50 g or 13-25% of energy) -- distinct from a generic very-low-calorie diet (400-800 kcal/day), which is only ketogenic if carbohydrate is also substantially restricted.
A very-low-calorie ketogenic diet is a strict, doctor-supervised plan of about 600-800 calories and very little carbohydrate a day. It's different from a generic very-low-calorie diet, which isn't automatically ketogenic.
Safety. Needs medical supervision — do not act on this without a clinician.
Only when
VLCKD trial protocol also specified 1.2-1.5 g protein/kg ideal body weight from high-biological-value sources, and 15-30 g fat mainly from extra virgin olive oil and omega-3 PUFAs (rondanelli-mariangela)
these are medically supervised protocols, not self-directed dieting
a very-low-calorie diet (VLCD) that does not also restrict carbohydrate substantially is not ketogenic (watanabe-m)
Who
overweight or obese adults undertaking a medically supervised VLCKD · Not general population attempting self-directed severe calorie/carbohydrate restriction without clinical supervision
Moderate confidence
Three independent persons, strongest evidence is a meta-analysis (rondanelli-mariangela), but that source's claim carries a disclosed conflict of interest: a co-author on that paper (sajoux-ignacio) is a Scientific Officer at PronoKal Group, a commercial VLCKD program. The remaining two sources are lower-tier (unclear, author_assertion). Named per editorial policy rule 3c; still counted as support but discounted in confidence.
3 independent people · 3 sources · strongest evidence: meta analysis
Sources
Effects of ketogenic diet on health outcomes: an umbrella review of meta-analyses of randomized clinical trials. · BMC medicine · 2023 · doi:10.1186/s12916-023-02874-y
Does the Ketogenic Diet Mediate Inflammation Markers in Obese and Overweight Adults? A Systematic Review and Meta-Analysis of Randomized Clinical Trials. · Nutrients · 2024 · doi:10.3390/nu16234002
Scientific evidence underlying contraindications to the ketogenic diet: An update. · Obesity reviews : an official journal of the International Association for the Study of Obesity · 2020 · doi:10.1111/obr.13053
Major diabetes-care guidelines, including the American Diabetes Association's Standards of Care, recognize low-carbohydrate and very-low-carbohydrate diets as an appropriate option for adults with diabetes or prediabetes, improving glycemic control (including HbA1c) and potentially allowing reduced glucose-lowering medication. Anyone taking glucose-lowering or blood-pressure medication should make this change only with their clinician, since doses may need adjusting.
Diabetes guidelines say a low-carb diet is a reasonable option for many people with diabetes or prediabetes -- it can improve blood sugar and sometimes let you reduce medication, but only with your doctor's involvement.
Safety. Needs medical supervision — do not act on this without a clinician.
Only when
applies to low-carbohydrate diets generally, not specifically the ketogenic diet (malinowska-d's paper is explicit on this point)
medication dose reduction is a potential outcome requiring clinician oversight, not a guaranteed or self-directed result
"selected individuals" -- not a blanket recommendation for everyone with diabetes
Who
adults with type 2 diabetes or prediabetes, selected per clinician judgment
Moderate confidence
Three independent papers converge on the same underlying ADA guideline position, but none is primary trial evidence -- this principle reports guideline consensus, not independently replicated trial data. 'guideline' (dynka-d) is not on the formal evidence ladder (meta_analysis>rct>observational>mechanistic>clinical_experience>author_assertion); treated cautiously as at or above clinical_experience tier, which caps overall confidence at moderate.
3 independent people · 3 sources · strongest evidence: guideline
Sources
Ketogenic Diet in Obesity and Diabetes: A Narrative Review. · Nutrients · 2026 · doi:10.3390/nu18122004
Ketogenic Diet: A Review of Composition Diversity, Mechanism of Action and Clinical Application. · Journal of nutrition and metabolism · 2024 · doi:10.1155/2024/6666171
Ketogenic dietary therapy is delivered in several distinct forms -- including the classic ketogenic diet, the modified Atkins diet, a medium-chain-triglyceride ketogenic diet, and a low-glycemic-index approach -- which differ mainly in their fat-to-carbohydrate ratio and dietary flexibility.
There isn't just one 'keto diet' -- there are several versions that differ in how strict the fat-to-carb ratio is and how much flexibility they allow, from the classic strict version to more flexible variants.
Only when
classic ketogenic diet: fat-to-carbohydrate ratio of roughly 3:1 to 4:1
modified Atkins diet: fat-to-carbohydrate ratio of roughly 1:1 to 2:1, more flexible than the classic form
medium-chain-triglyceride (MCT) ketogenic diet: more flexible and higher carbohydrate tolerance than the classic diet, because MCTs yield more ketones per kilocalorie than long-chain triglycerides
low-glycemic-index treatment: related but not a true ketogenic diet -- it limits total daily carbohydrate and favors low-glycemic-index foods rather than targeting a specific fat-to-carbohydrate ratio
Who
general adults; the specific fat-to-carbohydrate ratios are documented in the pediatric/infant epilepsy literature but the taxonomy itself is described generally
Moderate confidence
Two independent review-type sources (chen-s, an umbrella review; newmaster-k, a narrative review) agree on the same four-way taxonomy, though chen-s's evidence_type is marked 'unclear' and newmaster-k's population is pediatric epilepsy rather than the general/metabolic-health population this library targets. Evidence type is not RCT/meta-analysis for this descriptive claim, so confidence is moderate rather than high.
Ketogenic Diet and Multiple Health Outcomes: An Umbrella Review of Meta-Analysis. · Nutrients · 2023 · doi:10.3390/nu15194161
Restricting carbohydrate intake lowers the circulating insulin-to-glucagon ratio, which increases fatty-acid availability for oxidation and drives ketone-body production from fatty acids and ketogenic amino acids.
Eating less carbohydrate lowers insulin relative to glucagon, which frees up more fat to be burned for energy and produces ketones as a byproduct -- the basic mechanism behind ketosis.
Who
general adults restricting carbohydrate intake
Moderate confidence
Two independent people (goday-a, patikorn-c; no shared byline per people.yaml) support this mechanistic explanation, all from papers (no spoken-source cap). Both members' evidence_type is mechanistic, a mid-tier proposed mechanism rather than a measured trial outcome, so confidence is capped at moderate. This is a mechanism explanation, not an instruction, so actionable is false and no coach_rule is set.
2 independent people · 2 sources · strongest evidence: mechanistic
Sources
Short-term safety, tolerability and efficacy of a very low-calorie-ketogenic diet interventional weight loss program versus hypocaloric diet in patients with type 2 diabetes mellitus. · Nutrition & diabetes · 2016 · doi:10.1038/nutd.2016.36
Effects of ketogenic diet on health outcomes: an umbrella review of meta-analyses of randomized clinical trials. · BMC medicine · 2023 · doi:10.1186/s12916-023-02874-y
Long-term adherence to a ketogenic diet is difficult to sustain because of its restrictiveness (dietary monotony, psychosocial burden, and meal-preparation demands), and this is especially pronounced in populations with culturally high habitual fruit and vegetable intake, such as Mediterranean populations, where long-term compliance and effectiveness have been observed to be low.
Keto diets are hard to stick with long-term because they're restrictive and repetitive, especially if fruits and vegetables are a big part of your normal eating. Plan for that difficulty up front rather than assuming willpower will carry you.
Only when
Applies to adults with obesity or type 2 diabetes following a ketogenic diet (KD).
Adherence barriers include dietary monotony, psychosocial burden, meal-preparation demands, and the need to avoid fruit and some fresh vegetables.
Effect is most pronounced in populations/cultures where fruit and vegetable consumption is habitual and culturally valued (e.g. Mediterranean-area populations); one such population showed low long-term compliance and effectiveness.
Who
Adults with obesity or type 2 diabetes considering or following a ketogenic diet, particularly in dietary cultures with high habitual fruit/vegetable intake
Moderate confidence
Two independent authors of review articles (a narrative review and a systematic review) agree; both describe adherence challenges qualitatively rather than with a strong causal effect size, and one source's evidence_type is "unclear", so this stays at moderate rather than high.
2 independent people · 2 sources · strongest evidence: observational
Sources
Ketogenic Diet in Obesity and Diabetes: A Narrative Review. · Nutrients · 2026 · doi:10.3390/nu18122004
Efficacy of Ketogenic Diets on Type 2 Diabetes: a Systematic Review. · Current diabetes reports · 2021 · doi:10.1007/s11892-021-01399-z
Compared with refined grain foods, whole grain foods acutely reduce hunger and desire-to-eat and increase fullness and satiety, but this appetite benefit only translates into meaningfully lower energy intake when at least about 90 g of whole grain is consumed per day.
Whole grain foods make you feel less hungry and fuller than refined grain right after eating. But eating less overall only happens with a decent daily amount of whole grain, not a small swap.
Only when
acute or same-day comparison of whole grain vs refined grain foods
the appetite benefit was significant mainly in studies matched by available carbohydrate rather than by calories or by calories and volume, which may partly reflect a larger food volume in the whole-grain arm rather than a whole-grain-specific effect
pooled reduction in total energy intake across all trials was small and not statistically significant; a significant reduction appeared only above roughly 90 g whole grain per day
rye and barley showed a significant satiety benefit over refined grain; wheat did not
a next-meal appetite effect was still detectable at least 11 hours after consumption in some trials
Who
general adults
Moderate confidence
Single systematic review/meta-analysis (one voice, sanders-l) pooling many RCTs, but some members flag contradicts_mainstream: the pooled effect on actual energy intake is small and not significant, and the appetite benefit may be a volume-matching artifact rather than a true whole-grain effect.
1 independent person · 12 sources · strongest evidence: meta analysis
Sources
Effects of Whole Grain Intake, Compared with Refined Grain, on Appetite and Energy Intake: A Systematic Review and Meta-Analysis. · Advances in nutrition (Bethesda, Md.) · 2021 · doi:10.1093/advances/nmaa178
Over 3 years in adults with overweight or obesity, a higher dietary glycemic index predicted worse glycemic control markers and greater waist-circumference gain more consistently than total or added sugar intake did, even though higher sugar intake still tracked with greater gains in body fat and BMI.
How fast a food raises blood sugar (its glycemic index) tracked more closely with blood sugar control and belly fat than the total amount of sugar eaten did, though more sugar still meant more body fat gained.
Only when
findings come from the PREVIEW weight-loss-maintenance cohort study (Sydney, Helsinki, Nottingham centres), 8-156 week follow-up, energy-adjusted intake models
added sugar intake showed no association with fasting insulin, fasting glucose, HbA1c, or C-peptide, but was positively associated with body fat percentage and BMI
free and total sugar intake were, if anything, inversely associated with some glycemic markers (fasting glucose, HbA1c, fasting insulin, C-peptide) in at least one cohort
total sugar intake was positively associated with body fat percentage
glycaemic index was positively associated with fasting insulin, HbA1c, C-peptide, and waist circumference across cohorts
these are observational associations from one cohort study; they do not establish causation
Who
adults 25-70y with overweight/obesity, with or without prediabetes, in a weight-loss-maintenance phase
Moderate confidence
Single observational cohort study (one research voice); results are directionally consistent across three independent cohorts within the same study, which supports internal consistency, but this is one paper's dataset, not independently replicated elsewhere. Several members carry contradicts_mainstream: true (the sugar-vs-diabetes-risk direction runs against common assumption), which routes this to review.
1 independent person · 7 sources · strongest evidence: observational
Sources
Longitudinal Associations of Dietary Sugars and Glycaemic Index with Indices of Glucose Metabolism and Body Fatness during 3-Year Weight Loss Maintenance: A PREVIEW Sub-Study. · Nutrients · 2023 · doi:10.3390/nu15092083
Total and added sugar intake are not consistently linked to higher type 2 diabetes risk, but sugar consumed in liquid form, such as sugar-sweetened beverages, is consistently linked to higher type 2 diabetes risk.
Eating sugar in food doesn't clearly raise diabetes risk, but drinking it in sugary beverages does.
Only when
runs counter to a common assumption among health professionals that added sugar intake broadly increases type 2 diabetes risk
a cited meta-analysis of 15 prospective cohort studies found no association between total sugar intake and type 2 diabetes risk, and higher sucrose consumption was associated with decreased risk
the liquid-vs-solid distinction is the key qualifier: sugar-sweetened beverages specifically, not solid-food sugar sources, are the form consistently linked to higher risk
Who
general adults
Moderate confidence
Strongest evidence present is a meta-analysis of 15 cohort studies cited within this single review (one voice); confidence is moderate rather than high because this is a narrative synthesis of others' epidemiology rather than new primary data from multiple independent voices.
1 independent person · 3 sources · strongest evidence: meta analysis
Sources
Longitudinal Associations of Dietary Sugars and Glycaemic Index with Indices of Glucose Metabolism and Body Fatness during 3-Year Weight Loss Maintenance: A PREVIEW Sub-Study. · Nutrients · 2023 · doi:10.3390/nu15092083
The apparent weight-loss advantage of very-low-carbohydrate diets in type 2 diabetes disappears in the trials at low risk of bias, and survives only in the trials at high risk of bias.
Very low-carb diets look better for weight loss only in the weaker studies. In the well-run ones they lose no more weight than higher-carb diets. Choose what you can stick to.
Only when
in low-risk-of-bias randomised trials the difference was 0.9 kg favouring higher-carbohydrate diets, 95% CI -1.9 to 3.6, not significant
in high-risk-of-bias trials the difference was -1.8 kg, 95% CI -2.8 to -0.7, favouring the low-carbohydrate arm
this is a risk-of-bias subgroup analysis within an umbrella review, so it describes the shape of the evidence rather than a treatment effect
Who
adults with type 2 diabetes choosing a weight-loss diet
Moderate confidence
Meta-analytic evidence from an umbrella review, with high extraction confidence (0.8-0.85) and a risk-of-bias split that is exactly the analysis that should decide this question. Held at moderate rather than high because it is one voice (churuangsuk-c; per people.yaml co-author reynolds-a shares the same voice, so the byline collapses to one).
1 independent person · 1 source · strongest evidence: meta analysis
Sources
Diets for weight management in adults with type 2 diabetes: an umbrella review of published meta-analyses and systematic review of trials of diets for diabetes remission. · Diabetologia · 2022 · doi:10.1007/s00125-021-05577-2
Muscle adaptation to a high-fat, low-carbohydrate diet -- a shift toward burning more fat and less carbohydrate as exercise fuel -- is substantially complete after about 5 days, and this down-regulation of carbohydrate use persists for at least 1-2 days after carbohydrate intake is restored, even when muscle glycogen is fully restored and carbohydrate is consumed during exercise.
Your muscles adapt to burning more fat and less carb for fuel after about 5 days on a low-carb, high-fat diet. That shift persists for a day or two even after you resume eating carbs normally.
Only when
the 5-day adaptation timeline is for a non-ketogenic high-fat diet (roughly 60-70% fat energy, restricted carbohydrate, no ketosis induced); little further change was seen up to 15 days
persistence of the fat-adapted fuel shift was measured at 1-2 days after carbohydrate intake resumed to normal or high levels, following 5-6 days of adaptation
the down-regulated carbohydrate use persists even under conditions designed to reverse it (glycogen restored, carbohydrate fed during exercise), per the source's summary of its own research programme
a related, excluded caveat from this same cluster: apparent reductions in muscle glycogen use during exercise after fat-adaptation partly reflect lower starting glycogen stores rather than a true fuel-sparing effect, so only studies matching pre-exercise glycogen levels can isolate a genuine fat-adaptation effect -- this is a methodological caveat about interpreting the literature, not part of the core timeline claim, so it is left out of the principle itself
Who
general adults undergoing deliberate high-fat, low-carbohydrate dietary adaptation, in the context of exercise fuel use
Moderate confidence
All support is a single-author narrative review (burke-louise) summarizing this author's own and others' RCT-based findings; there is no independent second voice in this cluster, so despite the underlying evidence including RCT data, confidence is capped at moderate pending independent confirmation. Per people.yaml, burke-louise shares_voice_with maughan-ronald (second author on an unrelated IOC consensus statement), which is irrelevant here since maughan-ronald is not in this cluster, but is noted to prevent future double-counting.
1 independent person · 1 source · strongest evidence: rct
Sources
Re-Examining High-Fat Diets for Sports Performance: Did We Call the 'Nail in the Coffin' Too Soon? · Sports Medicine (Auckland, N.Z.) · 2015 · doi:10.1007/s40279-015-0393-9
In type 2 diabetes, very-low-carbohydrate eating produced no greater weight loss than higher-carbohydrate diets over three to thirty-six months, and the strictest versions were the hardest to keep to within the first year.
Cutting carbohydrate very low does not produce more weight loss than a moderate reduction, and the strictest versions are harder to keep up past a year. Pick a level you can live with.
Only when
very low carbohydrate here means 21-70 g per day; the pooled difference was -0.7 kg, 95% CI -2.0 to 0.7, I2 46%, p=0.10
adherence assessed up to one year was poorer below about 50 g per day than under a less restrictive 130 g per day approach
the authors note many low-carbohydrate trials allowed carbohydrate back in later for maintenance, which may itself explain part of the adherence difference
the same umbrella review's risk-of-bias subgroup analysis points the same way -- the weight-loss advantage for very-low-carbohydrate eating survives only in the trials at high risk of bias (-1.8 kg, 95% CI -2.8 to -0.7) and disappears in those at low risk of bias (0.9 kg favouring higher-carbohydrate diets, 95% CI -1.9 to 3.6, not significant) -- merged in from carbohydrate_quality--1b02, which this principle supersedes
Who
adults with type 2 diabetes choosing how far to restrict carbohydrate
Moderate confidence
Meta-analytic evidence from an umbrella review at extraction confidence 0.7-0.85, with a clinically useful pairing of effect and adherence. Capped at moderate because it is one voice (churuangsuk-c; co-author reynolds-a is recorded in people.yaml as sharing that voice), and because the adherence comparison is observed across trials rather than randomised.
1 independent person · 1 source · strongest evidence: meta analysis
Sources
Diets for weight management in adults with type 2 diabetes: an umbrella review of published meta-analyses and systematic review of trials of diets for diabetes remission. · Diabetologia · 2022 · doi:10.1007/s00125-021-05577-2
Replacing a dessert's sugar with xylitol and raising its fibre, while keeping available carbohydrate unchanged, lowers its glycemic index and load and produces a better glucose, insulin and fullness response than the conventional version.
Swapping a dessert's sugar for xylitol and adding fibre gives a flatter blood-sugar and insulin response and leaves you fuller, with the same carbohydrate. Useful when dessert is happening regardless.
Only when
the reformulated dessert tested reached a glycemic index of 42 and a glycemic load of 8.8 per 150 g serving
the comparison is against a conventional sugar-sweetened dessert, and is not an endorsement of eating dessert
xylitol in quantity causes digestive upset in some people, which this source does not address
Who
adults with type 2 diabetes, and adults managing body weight
Moderate confidence
Randomised trial evidence at extraction confidence 0.75-0.85, directly measuring the glucose, insulin and appetite response. Held at moderate because it is one voice (argiana-v) and one small reformulation trial of one recipe, so it should not be generalised to every low-glycemic dessert on a shelf.
1 independent person · 1 source · strongest evidence: rct
Sources
Low-Glycemic-Index/Load Desserts Decrease Glycemic and Insulinemic Response in Patients with Type 2 Diabetes Mellitus. · Nutrients · 2020 · doi:10.3390/nu12072153
This is a summary of published research, not medical advice. It cannot account for your medications, conditions or history. Talk to a clinician before acting on any of it.