Ketosis

What Ketosis Is

Ketosis is the metabolic state in which the body produces ketone bodies in sufficient quantities to serve as a major fuel source. It happens when carbohydrate intake is sufficiently low (typically under 50 grams per day) or when the body has been without food for long enough (typically 2-4 days), forcing it to mobilise fat for energy and produce ketones as a byproduct of that mobilisation.

The defining marker is blood beta-hydroxybutyrate (BHB), the primary circulating ketone body. Reference ranges:

• Below 0.2 mmol/L: not in ketosis
• 0.2-0.5 mmol/L: trace ketosis (often seen overnight or after extended exercise)
• 0.5-1.5 mmol/L: nutritional ketosis (the target range for most ketogenic dietary purposes)
• 1.5-3.0 mmol/L: deeper ketosis (typically reached during fasting or strict ketogenic eating)
• 3.0-7.0 mmol/L: starvation ketosis (typical of 3-7 day water fasts)
• Above 10-15 mmol/L: ketoacidosis (a medical emergency, essentially only achievable in type 1 diabetics without insulin)

The crucial distinction: Nutritional ketosis (0.5-7 mmol/L) and diabetic ketoacidosis (typically 15+ mmol/L with severely acidic blood pH) are different physiological states. Nutritional ketosis is regulated by insulin (which is still present, just lower than in the fed state) and the body’s normal acid-base buffering. Ketoacidosis occurs when insulin is absent (type 1 diabetics not on insulin), allowing ketone production to escape regulation and overwhelm the body’s buffering capacity. People with intact insulin production cannot enter ketoacidosis from dietary ketosis alone; the regulatory feedback prevents it. This is medically settled but worth stating clearly because the popular confusion between the two states produces unnecessary fear about ketogenic eating.

Mild ketosis is the normal overnight state: Most healthy adults achieve trace ketosis (0.2-0.4 mmol/L) by morning after a typical 12-hour overnight fast. The body cycles in and out of mild ketosis routinely. What’s unusual is the sustained ketosis (>0.5 mmol/L for hours or days) that ketogenic eating or extended fasting produces.

How the Body Enters Ketosis

The phase-by-phase progression maps onto the metabolic transition described in Fasting Basics, with some specific details worth covering for ketosis. The progression draws on George Cahill’s foundational starvation metabolism work from the 1960s and 70s, plus the substantial keto-adaptation research from Stephen Phinney and Jeff Volek over the past four decades.

• Phase 1 (0-12 hours) – Glycolytic: You burn through circulating glucose and ingested fuel. Duration depends on what and how much you ate.
• Phase 2 (12-24 hours) – Glycogenolysis: The liver releases its glycogen stores (roughly 100-150 grams in a well-fed adult) as glucose to maintain blood sugar. Muscle glycogen (about 300-500 grams) is locally available to the muscle but doesn’t contribute to blood glucose because the muscle lacks the enzyme glucose-6-phosphatase that would release glucose into the blood.
• Phase 3 (24+ hours) – Gluconeogenesis: Liver glycogen depletes. The body shifts to producing new glucose from non-carbohydrate sources, primarily gluconeogenic amino acids (alanine, glutamine, glycine, serine, others) and glycerol (the backbone of triglycerides). The brain still requires glucose at this point because ketone production hasn’t ramped up sufficiently.
• Phase 4 (Days 1-3) – Ketogenesis: Reduced insulin and elevated glucagon shift the liver’s metabolic priorities. Free fatty acids released from adipose tissue arrive at the liver, where they’re broken down through beta-oxidation. The acetyl-CoA produced exceeds what the citric acid cycle can process, particularly when oxaloacetate has been diverted into gluconeogenesis. The surplus acetyl-CoA is converted into acetoacetate, which can be further reduced to beta-hydroxybutyrate or spontaneously broken down to acetone. Ketones are released into circulation and delivered to the brain, muscle, and other tissues.
• Phase 5 (Day 3+) – Sustained ketosis: Glucose-sparing is now substantial. The brain shifts increasingly to ketone metabolism. Muscle preferentially uses free fatty acids rather than ketones (which are spared for the brain). Protein breakdown drops as ketones reduce the demand for gluconeogenic amino acids. The body has settled into the fat-burning state.

In dietary ketosis (eating ketogenically rather than fasting), the same progression occurs but is achieved through carbohydrate restriction rather than total food deprivation. Eating high amounts of fat with very low carbohydrates (and moderate protein) keeps insulin low enough that lipolysis and ketogenesis can proceed, even though calories are being consumed.

The Ketone Bodies

Three molecules are produced and used during ketosis:

Acetoacetate (AcAc): The first ketone body produced in the liver. Made from two molecules of acetyl-CoA through several enzymatic steps. AcAc can be:

• Converted to BHB (which is more chemically stable and the major circulating form)
• Used directly for fuel by extrahepatic tissues
• Spontaneously broken down to acetone

Beta-hydroxybutyrate (BHB): Despite its name, BHB isn’t technically a ketone (it’s a carboxylic acid with a hydroxyl group). It’s the dominant circulating ketone body during ketosis, accounting for roughly 78% of total ketones in the bloodstream. BHB is what blood ketone meters measure. It’s converted back to AcAc inside cells before being used for fuel.

Acetone: The breakdown product of AcAc, accounting for a small fraction of total ketones. Acetone is volatile and excreted primarily through the lungs (the characteristic “ketogenic breath” or “fasting breath”) and to a lesser extent through urine. Breath ketone meters measure acetone as a proxy for BHB.

BHB as a signalling molecule: Eric Verdin and John Newman’s research at the Buck Institute and UCSF has established that BHB does more than provide fuel. BHB acts as a signalling molecule with several documented effects:

• HDAC inhibition. BHB inhibits histone deacetylase 1 and 2 (HDAC1, HDAC2), enzymes that normally remove acetyl groups from histones (the proteins around which DNA winds). HDAC inhibition produces sustained changes in gene expression, particularly upregulating genes involved in oxidative stress resistance.
• NLRP3 inflammasome inhibition. BHB inhibits the NLRP3 inflammasome, a major pro-inflammatory protein complex implicated in autoimmune diseases, atherosclerosis, and neurodegeneration. This is part of why ketosis tends to have anti-inflammatory effects beyond what the absence of food alone would produce.
• G-protein coupled receptor activation. BHB acts on specific receptors (HCAR2/GPR109A in particular) to regulate lipolysis and inflammation.
• Direct neuroprotection. BHB appears to have direct protective effects on neurons exposed to oxidative or metabolic stress, beyond what the fuel effect alone would provide.

Ketones aren’t just an alternative fuel for when glucose is scarce. They’re also signalling molecules that change gene expression and cellular behaviour. Some of the benefits attributed to ketogenic eating and extended fasting may trace specifically to BHB’s signalling effects rather than to fat oxidation per se. This is part of why the popular research interest in exogenous ketone supplements has grown, though whether supplementing exogenous BHB produces the full range of endogenous effects remains an open question.

Insulin, Glucagon, and the Hormonal Shift

The hormonal driver of ketosis is the ratio of glucagon to insulin. Ketosis requires glucagon-dominant signalling, which requires sufficiently low insulin.

• Insulin effects: Insulin increases glucose uptake, glycolysis, glycogen synthesis, adipogenesis, and protein synthesis. Insulin suppresses gluconeogenesis, glycogenolysis, lipolysis, ketogenesis, and protein breakdown. Insulin is the master switch between fed-state and fasted-state metabolism. As long as insulin is elevated, ketogenesis is suppressed.
• Glucagon effects: Glucagon does largely the opposite: stimulates glycogenolysis, gluconeogenesis, lipolysis, and ketogenesis. The insulin-to-glucagon ratio (rather than either hormone’s absolute level) determines metabolic direction.

How different macronutrients affect the hormones

Carbohydrates:

• Glucose response: dose-dependent and rapid
• Insulin response: elevated (varies with fibre content)
• Glucagon response: suppressed

When you eat carbohydrates, blood glucose rises quickly, insulin is released to manage it, glucagon drops, and the body enters fed-state metabolism. Ketogenesis is shut down.

Protein:

• Glucose response: moderate and delayed (up to 5 hours, especially with fat and fibre)
• Insulin response: dose-dependent (more protein = more insulin)
• Glucagon response: elevated (this is important)

Protein is more complex than people often assume. It does raise insulin, but it also raises glucagon, with the relative response depending on the carbohydrate context. Protein consumed without carbohydrates produces a balanced insulin-glucagon response that doesn’t shut down ketogenesis. Protein consumed with carbohydrates produces a substantial insulin response with suppressed glucagon, shutting down ketosis. This is why “high protein” and “ketogenic” can coexist when carbohydrates are restricted, but become incompatible when carbohydrates are added.

Fat:

• Glucose response: none
• Insulin response: minimal
• Glucagon response: elevated (in response to low insulin)

Fat has the least disruptive effect on ketosis. Long-chain triglycerides are packaged into chylomicrons and routed through the lymphatic system, eventually delivered to adipose tissue. Medium-chain triglycerides (MCT) bypass this process and are delivered directly to the liver, where they can be rapidly converted to ketones. MCTs are part of why MCT oil supplementation can accelerate the entry into ketosis.

The macronutrient hormonal responses: A ketogenic diet works because the absence of carbohydrate keeps insulin low. Adding a small amount of carbohydrate (a piece of bread, a glass of orange juice) can shut down ketogenesis for hours, even though the total calories are small. This is why ketogenic eating is sensitive to compliance in a way that other diets aren’t: small carbohydrate “cheats” produce disproportionate metabolic disruption.

Keto-Adaptation: The Six-Stage Process

Being in ketosis is not the same as being keto-adapted. The distinction is important and frequently elided in popular discussions.

• Ketosis is the metabolic state with elevated blood ketones. You can be in ketosis after 24-48 hours of fasting or strict ketogenic eating. The ketones are present in circulation.
• Keto-adaptation is the broader physiological adaptation that allows the body to efficiently use ketones and fatty acids as primary fuels. It involves changes in mitochondrial biogenesis, enzyme expression, fat oxidation capacity, ketone uptake by tissues, and overall metabolic flexibility. Keto-adaptation takes substantially longer than entering ketosis, typically 2-4 weeks for basic adaptation and 3-6 months for full adaptation.

The six-stage progression that Sim Land’s Metabolic Autophagy articulates (drawing on the Phinney/Volek metabolic ward research) tracks how the body progressively shifts toward fat-based metabolism:

• Stage 1: Carb Withdrawal: Days 1-3. You restrict carbohydrates and remove most or all glucose from your diet. Glycogen depletes. The body is still primarily glucose-dependent. Energy may feel inconsistent.
• Stage 2: Keto Flu: Days 3-14 (or longer). The body is in ketosis, but tissues aren’t yet efficient at using ketones. Symptoms include fatigue, brain fog, headaches, irritability, electrolyte imbalances, sleep disruption, and reduced exercise performance. This is the phase most people quit if they’re going to quit. Adequate sodium, magnesium, and potassium intake (covered in the Cheatsheet) substantially reduces keto flu severity.
• Stage 3: Getting Used to Ketones: Weeks 2-4. The brain begins extracting ketones more efficiently. Energy stabilises. Mental clarity improves. Some hunger reduction. Initial energy improvements appear.
• Stage 4: Fat Burning Mode: Weeks 4-8. Mitochondrial machinery for fat oxidation is upregulated. Exercise performance recovers and may exceed the previous baseline for low-intensity work. Recovery from workouts improves. Sustained energy across the day. Hunger drops substantially.
• Stage 5: Keto Adaptation: Months 2-6. The body runs efficiently on dietary fat and stored fat alike. You can fast for extended periods without significant performance drops. You can exercise at moderate intensity without needing carbohydrates. Mental focus is strong. The state feels normal rather than restrictive.
• Stage 6: Metabolic Flexibility: Months 3+. The body can readily switch between ketone and glucose metabolism. You can eat a carbohydrate-containing meal without crashing back into glucose dependence; the body briefly uses the glucose, then returns to fat oxidation. This is the ultimate goal of keto-adaptation: not perpetual ketosis, but the capacity to flexibly use whichever fuel is appropriate to the situation.

The timeline varies substantially between individuals. Younger, more metabolically healthy people typically progress faster. Older adults, particularly those with insulin resistance, can take substantially longer. Athletes adapt faster than sedentary people because their metabolic machinery is more responsive in general. Individual genetic variation in ketone uptake and fatty acid oxidation also affects progression.

Variations of the Ketogenic Diet

The original ketogenic diet, developed in the 1920s at the Mayo Clinic for refractory pediatric epilepsy, called for approximately 90% of calories from fat, 6-9% from protein, and minimal carbohydrate. This “classical” ketogenic diet remains the standard for epilepsy treatment. Variations have evolved for non-medical applications.

Classical ketogenic diet:

• Fat: ~90% of calories
• Protein: 6-9% of calories
• Carbohydrate: 0-4% of calories
• Used clinically for refractory epilepsy

Modified ketogenic diet (the most common contemporary form):

• Fat: 65-85% of calories
• Protein: 15-35% of calories
• Carbohydrate: 0-10% of calories (ideally from fibrous vegetables)

Daily targets for most people on a modified ketogenic diet:

• Carbohydrate: no more than 50 grams total per day (often 20-30 grams for stricter approaches)
• Protein: 1.2-1.6 g/kg of lean body mass (1.6-2.2 g/kg may be more appropriate for active people building muscle)
• Fat: remaining calories, or to satiety

The fat-protein-carbohydrate balance affects how readily and deeply ketosis develops. Higher protein at the expense of fat tends to produce shallower ketosis because some protein converts to glucose via gluconeogenesis. Higher carbohydrate (even with adequate fat) prevents ketosis from establishing. The classical 90% fat ratio achieves the deepest ketosis but is difficult to maintain long-term and isn’t necessary for most non-medical purposes.

What to Eat on a Ketogenic Diet

Carbohydrates (0-10% of calories), primarily from low-starch vegetables:

• All leafy greens and lettuces
• Cruciferous vegetables (broccoli, cauliflower, Brussels sprouts, cabbage)
• Celery, cucumber, zucchini
• Bok choy, mushrooms, asparagus
• Avocado (technically a fruit but functionally a fat)
• Small amounts of low-sugar berries (raspberries, blackberries, blueberries)

Protein (15-35% of calories):

• Fatty cuts of red meat (ribeye, chuck, brisket, lamb)
• Poultry with skin on (chicken thighs, duck)
• Fatty fish (salmon, mackerel, sardines, anchovies, herring)
• Eggs (including yolks)
• Organ meats (liver, heart, kidney) have exceptional nutrient density
• Nuts and seeds (macadamia, pecans, walnuts, hemp seeds, pumpkin seeds)
• Aged cheeses (if tolerated)

Fats (60-85% of calories), from saturated and monounsaturated sources, with emphasis on omega-3:

• Whole foods: avocados, eggs, fatty cuts of meat, coconut, olives, nuts
• Oils: coconut oil, MCT oil, ghee, butter, extra-virgin olive oil, avocado oil, macadamia nut oil
• Animal fats: tallow, lard, schmaltz

Minimise or avoid:

• All grains and grain-based foods
• All sugars and sweeteners (including most fruits)
• Starchy vegetables (potatoes, sweet potatoes, corn, beets)
• Most legumes and beans
• Industrial seed oils (soybean, corn, cottonseed, sunflower, safflower, canola, grapeseed, peanut)
• Anything labelled “low-fat” is typically high in carbohydrates or seed oils

The Carnivore Diet

The carnivore diet represents the extreme end of the low-carbohydrate spectrum: eating only animal foods, with most variants excluding all plants entirely. The diet emerged from relative obscurity to substantial popularity in the late 2010s through the work of practitioners and self-experimenters, including Shawn Baker (an orthopaedic surgeon who’s been one of the most visible advocates), Jordan and Mikhaila Peterson (whose substantial autoimmune disease remission on carnivore eating drew major attention), and a growing community of self-reporters documenting their experiences.

By removing all plant foods, you eliminate:

• All plant compounds (lectins, oxalates, phytates, polyphenols, salicylates, glutamates, FODMAPs, alkaloids, defensive chemistry)
• All fiber
• All carbohydrates beyond the trace amounts in animal foods
• All anti-nutrients that bind minerals

The diet typically maintains sustained ketosis because carbohydrate intake is essentially zero. Protein is high. Fat ratios vary depending on which animal foods are emphasised (ribeye-heavy carnivore produces high-fat ratios; lean meat-heavy variants produce higher protein ratios).

Experiential reports: The most striking aspect of carnivore eating in the recent practitioner literature has been the consistency of certain reported effects:

• Substantial autoimmune disease remission (especially rheumatoid arthritis, lupus, ulcerative colitis, psoriasis, atopic dermatitis, Hashimoto’s thyroiditis)
• Resolution of various digestive issues (IBS, IBD symptoms, food sensitivities)
• Mood improvements (substantial reports of depression and anxiety remission)
• Substantial weight loss in obese individuals
• Improvements in chronic pain conditions
• Sleep quality improvements

These are experiential reports rather than clinical trial data. The current state of the clinical evidence base is limited:

• Lennerz et al. 2021 Current Developments in Nutrition surveyed 2,029 self-identified carnivore dieters and found substantial self-reported improvements in health markers and minimal adverse effects across the cohort. This is a survey study, not a controlled trial, and is subject to selection bias (people having bad outcomes wouldn’t be on the diet to be surveyed), but the size and consistency of the reported effects are striking.
• Limited case study literature documenting specific autoimmune remissions.
• A growing but still limited body of clinical observation from practitioners working with carnivore patients.

The vitamin C question: One of the more interesting mechanistic stories around carnivore eating. The conventional understanding is that humans need to consume vitamin C because we can’t synthesise it and severe deficiency produces scurvy. Plants are the typical dietary source. Meat contains small amounts of vitamin C, but the amounts in muscle meat are typically considered inadequate for vitamin C requirements, as commonly stated.

The mechanism that may resolve this apparent paradox involves GLUT-1, the glucose transporter that also transports vitamin C. Vitamin C and glucose share the same uptake pathway into cells, and they compete for it. On a high-carbohydrate diet with elevated blood glucose, the GLUT-1 transporters are largely occupied by glucose, limiting vitamin C uptake. On a very low-carbohydrate or ketogenic diet with much lower blood glucose, more GLUT-1 transporters are available for vitamin C, potentially allowing the smaller dietary intake to suffice. This may explain why long-term carnivore eaters don’t develop scurvy despite intakes that would be inadequate on a standard diet.

The Arctic explorer literature supports the practical claim: Vilhjalmur Stefansson and Karsten Anderson lived under medical supervision in New York’s Bellevue Hospital from 1928 to 1929 on a meat-only diet for a year, with no vitamin C supplementation. They developed no scurvy and maintained good health on the protocol. The Inuit traditionally consumed diets with minimal plant content (often less than 5% of calories) and showed no vitamin C deficiency. The mechanism wasn’t well understood at the time; the glucose-vitamin C competition framework is more recent.

The clinical evidence for this mechanism is mechanistically plausible but not extensively validated in randomised trials. It’s reasonable to engage with it as a hypothesis that explains observed phenomena rather than as settled science.

The autoimmunity question: The consistency of autoimmune disease remission reports on carnivore eating is one of the more intriguing patterns in the practitioner literature. Several potential mechanisms:

• Plant compound elimination. Many plants produce defensive chemistry that triggers immune responses in some individuals. Lectins (especially wheat germ agglutinin and similar), oxalates, salicylates, and other compounds have been implicated in immune reactivity for some people. Complete elimination removes these triggers entirely.
• Gut healing. The absence of fibre and plant compounds may allow the gut lining to heal in people with compromised intestinal barrier (“leaky gut”), which triggers systemic immune responses.
• Ketosis as an anti-inflammatory state. BHB’s NLRP3 inflammasome inhibition (covered above) reduces a major pro-inflammatory signalling pathway implicated in autoimmune conditions.
• Histamine reduction. Many plants are histamine liberators or contain histamine themselves; their elimination reduces total histamine burden.
• Elimination diet effect. Any sufficiently strict elimination diet produces some autoimmune improvement in people whose triggers are plant compounds; carnivore happens to eliminate all of them essentially simultaneously.

The experiential reports are consistent and substantial enough to take seriously. The clinical evidence base is limited. The mechanistic hypotheses are plausible but not all proven. People with autoimmune conditions who try carnivore protocols often report substantial improvement; this doesn’t mean carnivore is the optimal long-term diet for everyone or that the mechanism is fully understood.

What we don’t know about long-term carnivore eating:

• The longest sustained populations on a near-carnivore diet (Inuit, Maasai, Mongolian pastoralists historically) had genetic adaptations and food preparation practices that differ from modern carnivore practitioners. Direct comparisons are imperfect.
• The microbiome effects of zero-fibre eating are not fully characterised. The dominant view in microbiome research has been that fibre is essential for microbial diversity; carnivore eating directly contradicts this assumption. Long-term studies of microbiome health in carnivore eaters are limited.
• Cardiovascular risk markers of carnivore eating are mixed. LDL cholesterol often rises substantially, sometimes dramatically. Whether this represents elevated cardiovascular risk in the metabolic context of low insulin and elevated HDL is contested (the “lean mass hyper-responder” phenomenon has produced substantial discussion).
• Long-term cancer outcomes are unknown. The mechanistic concerns around mTOR elevation from high protein intake are real; the clinical evidence either way is limited.
• Mineral and micronutrient status vary substantially depending on whether organ meats are consumed. People eating only muscle meat may develop deficiencies that those eating nose-to-tail carnivore wouldn’t.

For people interested in trying carnivore: The general approach in the practitioner community is to start with an elimination phase (30-90 days of strictly animal foods, often called the “Lion Diet” in its strictest version) to assess effects, then potentially reintroduce specific foods to identify triggers. Most practitioners recommend including organ meats (especially liver, weekly), eggs, and fatty cuts rather than relying on lean muscle meat alone. Adequate sodium intake matters more than on most diets due to substantially reduced retention in low-insulin states.

Carnivore eating is one of several legitimate approaches with substantial experiential support and limited but emerging clinical evidence. It works dramatically well for some people, particularly those with autoimmune conditions and metabolic dysfunction. It isn’t a universal optimal diet, and the long-term data are genuinely limited. Anyone considering a carnivore diet for therapeutic reasons should engage with the substantial experiential literature, consider working with a practitioner experienced in the approach, and approach it as a structured experiment rather than a permanent commitment until they understand how their body responds.

Metabolic Flexibility

The goal of working with ketosis isn’t necessarily to remain in strict ketosis indefinitely. The goal is metabolic flexibility: the capacity to efficiently use whichever fuel is appropriate to the situation, switching between ketones, fatty acids, and glucose as conditions require.

A metabolically flexible person can:

• Fast for 24+ hours without significant performance impairment
• Eat a carbohydrate-containing meal without crashing back into glucose dependence
• Exercise at moderate intensity without needing pre-workout carbohydrates
• Maintain stable energy across the day without snacking
• Switch fuel sources based on metabolic demand rather than being locked into one

Several protocols use ketosis as a base while allowing strategic carbohydrate use:

• Targeted Ketogenic Diet (TKD): Consume small amounts of easily digestible carbohydrates (5-30 grams) before or during intense workouts. The carbohydrates support performance during the workout while maintaining ketosis the rest of the time. Suited to people doing intense resistance training or high-intensity intervals 4+ times per week.
• Cyclical Ketogenic Diet (CKD): Maintain ketogenic eating 5-6 days per week, with 1-2 days of higher-carbohydrate refeeding. The carbohydrate days replenish muscle glycogen for high-intensity training while the ketogenic days maintain fat-burning adaptations. Suited to bodybuilders, powerlifters, CrossFit athletes, and competitive sports requiring substantial glycogen.
• Carb Backloading: Eat ketogenically during the day, then consume carbohydrates after evening resistance training. The post-workout window is the most insulin-sensitive part of the day; carbohydrates consumed then are preferentially routed toward muscle glycogen rather than fat storage. Suited to people training in the evenings who want some carbohydrate without disrupting overall metabolic flexibility.
• Seasonal Cycling: Eat more carbohydrates in summer (when fruit, vegetables, and grain are seasonally appropriate in temperate climates) and more ketogenically in winter (when historically the diet would have been more meat and stored fat). This connects to the broader framework around seasonal eating covered in the Nutrition section.

The choice between strict ketosis and metabolic flexibility approaches depends on goals. Therapeutic ketosis (cancer, epilepsy, neurodegenerative disease) typically requires sustained strict ketosis. Performance and general health goals are usually better served by metabolic flexibility, with periodic ketogenic phases to maintain fat-oxidation capacity.

Ketosis and the Brain

The clinical application that established ketogenic eating as a recognised medical treatment was epilepsy.

Epilepsy as the foundational application: The ketogenic diet was developed at the Mayo Clinic in 1921 by Russell Wilder, based on observations that fasting reduced seizure frequency in epileptic patients. Wilder reasoned that a diet that mimicked the metabolic state of fasting (high fat, low carbohydrate, sufficient protein) might produce similar therapeutic effects without the constraints of total food deprivation. The ketogenic diet became the standard treatment for pediatric epilepsy throughout the 1920s and 30s, fell out of favour with the introduction of anticonvulsant medications in the 1940s-70s, and was rediscovered in the 1990s for refractory epilepsy (cases that don’t respond to medication).

Roughly half of children with refractory epilepsy who go on a ketogenic diet experience >50% reduction in seizure frequency, with about a third experiencing >90% reduction. The mechanism likely involves multiple pathways: stable energy supply from ketones reduces neuronal hyperexcitability, BHB has direct anti-seizure effects, and altered amino acid metabolism (lower glutamate, higher GABA) shifts the brain toward inhibition rather than excitation.

Alzheimer’s disease and ketones: The “type 3 diabetes” framing for Alzheimer’s (covered in Fasting Basics) suggests that the disease involves substantial metabolic dysfunction, with insulin resistance in the brain contributing to the energy crisis that precedes cognitive decline. Ketones provide an alternative fuel that doesn’t require insulin signalling, potentially bypassing the metabolic block.

Clinical evidence is preliminary but growing:

• MCT supplementation (which raises blood BHB) has shown modest cognitive improvements in mild-to-moderate Alzheimer’s patients in several small trials
• Ketogenic diet protocols have shown some benefit in MCI (mild cognitive impairment) and early Alzheimer’s
• The mechanism through BHB-mediated reduction of beta-amyloid and tau pathology is being actively investigated

Promising preliminary work, not yet established as a standard treatment, worth taking seriously as an experimental approach.

Parkinson’s disease and ketones: Similar mechanistic logic applies. Several small trials have shown ketogenic diet protocols producing improvements in motor and non-motor symptoms in Parkinson’s. The evidence base is preliminary.

Migraine: A substantial portion of chronic migraine patients respond to ketogenic diets, often dramatically. Mechanism likely involves stable energy supply, reduced inflammation, and possibly mitochondrial improvements. Evidence base includes several small controlled trials and substantial clinical observation.

Mood disorders: Emerging research on the ketogenic diet for bipolar disorder, treatment-resistant depression, and schizophrenia spectrum disorders has been promising. Several case series and small trials have shown substantial improvements. Christopher Palmer at Harvard has been one of the more visible academic voices articulating “metabolic psychiatry” as a working framework. The clinical evidence base is preliminary but rapidly developing.

Ketosis and Cancer

This is one of the more contested clinical applications, and because cancer is big business and pharmaceutical companies have spent billions creating anti-cancer marketing material that attributes it to bad luck, we need to be careful.

The metabolic theory of cancer: Cancer cells, as Otto Warburg first documented in the 1920s, preferentially use glycolysis even in the presence of oxygen (“the Warburg effect”). This metabolic shift is one of the defining features of most cancer cells and was being studied adamantly until research grants were diverted to genetic studies. The metabolic theory of cancer, articulated most thoroughly by Thomas Seyfried at Boston College, holds that this metabolic dysfunction is central to cancer rather than incidental to it. The therapeutic implication: depriving cancer cells of glucose while providing ketones for healthy tissue might selectively starve cancer cells.

The clinical state of the evidence:

• Mechanistic plausibility is strong (cancer cells do depend more heavily on glucose than most healthy cells)
• Animal model evidence consistently shows that ketogenic diets slow tumour growth across various cancer types
• Case studies and small clinical series have documented some dramatic responses to the ketogenic diet combined with standard treatment
• The combination of the ketogenic diet with hyperbaric oxygen (Seyfried’s “press-pulse” protocol) has shown promising results in case studies

What the evidence doesn’t yet support:

• Ketogenic diet as a standalone cancer treatment
• Specific protocols as established standard of care
• The GKI of 0.7-2.0 as a clinically validated target (Seyfried’s recommendation is based on the mechanistic case, not on clinical trial validation)

Ongoing clinical trials are evaluating the ketogenic diet as an adjuvant therapy alongside chemotherapy and radiation. Early results suggest the ketogenic diet may improve tolerance of standard treatments (potentially reducing side effects) and possibly enhance their effectiveness against certain cancers (glioblastoma has been a particular focus). The picture is genuinely promising but not yet conclusive.

Cancer is a serious disease, and metabolic management is one of several legitimate lines of investigation. Anyone using a ketogenic diet for cancer should work with an oncologist familiar with the approach rather than pursuing it as a sole intervention. The mechanistic case justifies serious investigation; the clinical evidence doesn’t yet justify treating the ketogenic diet as an established cancer therapy. The interesting research is happening; the verdict isn’t yet in. Engage with it as an active investigation rather than a settled answer.

Other Clinical Applications

• Type 2 Diabetes Reversal: This is one of the strongest clinical applications of ketogenic eating. Sarah Hallberg’s work with Virta Health has demonstrated that ketogenic diet protocols can reverse type 2 diabetes (defined as HbA1c below diabetic range without medication) in a substantial percentage of patients. The 2018 Diabetes Therapy paper showed 60% of patients reduced or eliminated diabetes medications, with HbA1c reductions averaging 1.3 percentage points over a year. Two-year follow-up showed sustained results. The mechanism is straightforward: removing the dietary trigger (carbohydrate) for the insulin-resistant state allows the underlying physiology to recalibrate.
• Polycystic Ovary Syndrome (PCOS): PCOS has substantial insulin resistance components. Ketogenic protocols have shown improvements in insulin sensitivity, ovulation regularity, weight, and androgen excess in several small trials. Clinical evidence is preliminary, but the mechanistic case is strong.
• Cardiovascular Disease Risk Markers: Ketogenic diets generally improve triglycerides substantially (often 50%+ reductions), raise HDL, improve insulin sensitivity, and reduce inflammation markers. LDL cholesterol responses are heterogeneous: most people see modest changes, some see substantial elevations (the “lean mass hyper-responder” pattern), and the cardiovascular risk implications of elevated LDL in the metabolic context of low insulin and elevated HDL are contested. The overall cardiometabolic profile typically improves substantially even when LDL rises.
• Fatty Liver Disease: Non-alcoholic fatty liver disease (NAFLD) is the most common chronic liver condition globally and has strong associations with insulin resistance. Ketogenic protocols consistently reduce liver fat in clinical trials, often dramatically. Mechanism involves reduced de novo lipogenesis (the liver makes less fat when insulin is low) and increased fatty acid oxidation.

Risks, Caveats, and Who Shouldn’t Do Keto

Ketogenic eating isn’t appropriate for everyone. Specific contraindications:

• Type 1 Diabetes: Requires careful medical supervision. The risk of diabetic ketoacidosis is serious, and the management of insulin dosing with carbohydrate restriction requires expertise. Some type 1 diabetics use ketogenic approaches successfully with appropriate medical guidance; doing it without supervision is genuinely dangerous.
• Pregnancy and Breastfeeding: The evidence base for ketogenic eating during pregnancy is limited, and the available animal research has produced some concerning findings around developmental effects. The mainstream medical recommendation is to avoid a ketogenic diet during pregnancy. We don’t know enough to recommend ketogenic eating during pregnancy, and there are reasonable mechanistic reasons (fetal development uses substantial glucose; metabolic shifts may be unpredictable) to be cautious.
• Children: Outside of medical applications (refractory epilepsy under specialist supervision), routine ketogenic eating for children isn’t recommended. Children have different metabolic and developmental needs, and the long-term effects of sustained ketogenic eating during growth and development are insufficiently studied.
• Eating Disorder History: The restrictive nature of ketogenic eating can interact badly with eating disorder histories, potentially triggering relapse or reinforcing disordered patterns. Anyone with a history should approach ketogenic eating cautiously and ideally with mental health support.
• Pancreatic Insufficiency: People with substantially reduced pancreatic function (chronic pancreatitis, pancreatic cancer, certain genetic conditions) may not tolerate the high-fat intake required for ketogenic eating.
• Certain Liver Conditions: People with substantial liver dysfunction may have difficulty managing the metabolic demands of ketogenic eating.
• Gallbladder Removal: People who’ve had their gallbladder removed can manage ketogenic eating but typically need to ramp up fat intake gradually and may benefit from bile acid supplementation.
• Older Adults at Risk of Sarcopenia: Older adults who are losing muscle mass need adequate protein intake to maintain muscle. Ketogenic protocols that prioritise fat over protein can exacerbate sarcopenia risk. Appropriate ketogenic eating for older adults should emphasise adequate protein.
• Athletes Doing Substantial High-Intensity Work: Strict ketogenic eating impairs maximum power output and high-intensity performance. Athletes who require glycolytic capacity (sprinters, powerlifters, weightlifters, team sport athletes) typically perform better on approaches that include carbohydrate, even if they use ketogenic phases at other times.

Common side effects during keto-adaptation:

• Keto flu symptoms (covered in keto-adaptation section above)
• Initial water weight loss (substantial; this isn’t fat loss)
• Constipation (manageable with adequate hydration, magnesium, and fibre from low-carb vegetables)
• Bad breath (“keto breath” from acetone excretion)
• Hair shedding (transient, usually resolves within 2-3 months)
• Sleep disruption (usually transient)
• Reduced exercise performance during adaptation (usually transient)

Testing Ketones

• Blood ketone meters (measuring BHB). Most accurate. Test strips are expensive ($1-2 per test). Best for monitoring during the adaptation phase or specific therapeutic protocols. Once adapted, regular testing typically isn’t necessary. Common models: Keto-Mojo, Precision Xtra, KetoBM.
• Breath ketone meters (measuring acetone). Less expensive over time (no consumables). Slightly less precise but adequate for tracking trends. Common models: Ketonix, LEVL.
• Urine ketone strips (measuring acetoacetate). Cheapest. Useful initially, but becomes unreliable once keto-adapted because adapted bodies excrete less unused acetoacetate. Not recommended for long-term monitoring.

Summary

Ketosis is a normal metabolic state that humans evolved to inhabit periodically. The body produces ketones overnight, during exercise, and during any sustained period without food. Sustained ketosis (achieved through ketogenic eating or extended fasting) accesses adaptations that mild overnight ketosis doesn’t, including substantial neurological, immune, and metabolic effects.

The clinical applications with the strongest evidence base are epilepsy (the foundational application), type 2 diabetes reversal, and certain metabolic syndromes. Promising preliminary evidence supports applications to neurodegenerative disease, certain cancers (as an adjuvant), mood disorders, migraine, and PCOS. Experimental and personal experimentation with carnivore eating has produced substantial improvements in autoimmune conditions and metabolic dysfunction for many people, though the clinical evidence base is limited.

The goal for most people isn’t perpetual strict ketosis but metabolic flexibility: the capacity to use whichever fuel is appropriate to the situation. Periodic ketogenic phases (combined with the time-restricted eating and longer fasting protocols covered in Fasting Basics and the Cheatsheet) maintain fat-oxidation capacity without requiring permanent dietary restriction.

The biology is clear enough that the question for most people isn’t whether ketosis is real or useful, but rather what role periodic ketosis should play in your specific context. The Game Plan covers how to think about that question for different goals and life stages.