evipedia.ai
← Back

Fasting-Mimicking Diet for Health & Longevity

Evidence Review created on 08/31/2026 using AI4L / Opus 5

Also known as: FMD, Fasting Mimicking Diet, Periodic Fasting-Mimicking Diet, ProLon

Motivation

The fasting-mimicking diet is a five-day plant-based eating plan, low in calories, sugar and protein and comparatively high in fat, designed to hold the body in a fasting-like state while a person still eats. It was built so that the biological responses associated with going several days without food could be reached without the difficulty and medical supervision that a water-only fast demands, and it is normally repeated as occasional cycles rather than followed continuously.

The approach grew out of laboratory work on nutrition and ageing, was later packaged and sold as a boxed five-day kit, and has since been tested in people with excess weight, raised blood sugar and inflammatory bowel disease. It draws attention among people who track their own health markers partly because a short, bounded protocol is easier to schedule than permanent calorie restriction, and partly because its developers describe effects reaching beyond weight loss.

This review examines what the human and animal evidence shows: which changes the diet produces, in whom, how large and how durable those changes are, what harms and practical limits have been recorded, and how much weight the evidence base can carry given who funded it.

Benefits - Risks - Protocol - Conclusion

High-level orientation pieces on the fasting-mimicking diet from expert platforms and from the narrative-review literature.

Priority-platform note: hubermanlab.com returned nothing on this diet, and lifeextension.com covers it only inside broader pieces on fasting and calorie restriction rather than in a dedicated overview, so no item from those two is listed.

Grokipedia

Fasting-mimicking diet

An encyclopedic overview of the protocol, its physiological mechanisms, clinical applications, safety limits and commercial packaging, useful as a single orientation page before reading the primary literature.

Examine

Examine.com has no dedicated page for the fasting-mimicking diet; its search returns intervention pages on prolonged, intermittent and alternate-day fasting plus member-only study summaries, none of which is a primary entry for this diet.

ConsumerLab

ConsumerLab.com has no article on the fasting-mimicking diet; a site search returns only unrelated supplement reviews and answers on intermittent fasting, so no primary entry for this diet exists there.

Systematic Reviews

Systematic reviews and meta-analyses covering the fasting-mimicking diet’s metabolic effects, its oncology rationale and its principal trade-off against lean tissue.

Mechanism of Action

The diet works by removing the nutrient signals that keep cells in growth mode. Five days at roughly 750–1,100 kilocalories, with protein held near 10% of energy, lowers blood glucose and insulin, which in turn lowers IGF-1 (insulin-like growth factor 1, a hormone that drives cell growth and division). Falling IGF-1 and amino-acid availability reduce activity in two nutrient-sensing pathways — mTOR (mechanistic target of rapamycin, the cell’s main growth switch) and PKA (protein kinase A, an enzyme that relays hormone signals inside the cell) — while raising AMPK (AMP-activated protein kinase, the cell’s low-energy sensor). The result is a shift from building tissue to maintaining it: autophagy (the cell’s recycling of its own damaged components) increases, and the liver switches to burning fat, raising the ketone beta-hydroxybutyrate.

The refeeding phase is held to matter as much as the fast. In mice, cells depleted during the diet are replaced from progenitor and stem-cell pools once food returns, which is the proposed basis for the regenerative claims.

A competing explanation holds that nothing pathway-specific is happening: the effects track the energy deficit itself, and pooled comparisons of intermittent against continuous calorie restriction find broadly similar cardiometabolic results. A third position, supported by a trial comparing high- and low-protein versions of the diet, is that protein restriction is not the essential ingredient it was assumed to be.

Historical Context & Evolution

Therapeutic fasting was practised in European clinics from the 1920s, notably by Otto Buchinger, but was used for digestive and rheumatic complaints rather than for ageing. The modern line begins in the 1990s with yeast and mouse work at the University of Southern California showing that removing growth signals extended lifespan. Water-only fasting was the obvious human translation, but trials recruited poorly and clinicians were reluctant to supervise multi-day starvation, so a fed protocol delivering a similar metabolic signature was engineered in its place.

The first full report combined mouse longevity data with a small human pilot in 2015; a randomised trial in 100 generally healthy adults followed in 2017. Commercial packaging as a boxed five-day kit began in 2016, and trials expanded into oncology, type 2 diabetes, multiple sclerosis and inflammatory bowel disease. An international expert panel formally separated the term from prolonged fasting and time-restricted eating in 2024.

The original animal findings have not been retracted or overturned; what has changed is the weight placed on them. Critics argue the human effects are explained by the calorie deficit rather than by a distinct fasting programme, that trial samples are small, and that a large share of the literature originates from — or is funded by — the developer and the company he founded, L-Nutra. Supporters point to twelve-month primary-care data and to an inflammatory-bowel-disease trial run at a university medical centre, though the developer co-authored that one too. Both readings remain live.

Expected Benefits

High 🟩 🟩 🟩

Reduced Body Weight and Visceral Fat

Repeated five-day cycles produce weight loss that persists between cycles, with fat drawn preferentially from the abdomen. In a twelve-month primary-care trial in type 2 diabetes, monthly cycles reduced weight against usual care, and magnetic resonance imaging in that trial showed abdominal visceral fat falling while abdominal muscle area was unchanged. A three-month randomised trial in generally healthy adults found the same direction of effect on trunk and total body fat. Losses scale with starting adiposity, so lean, metabolically healthy people see smaller changes.

Magnitude: −3.6 kg body weight versus usual care at twelve months (95% CI, the range in which the true effect most likely lies: −5.2 to −2.1) and −37.9 cm² abdominal visceral fat (95% CI −54.7 to −21.0), reported in van den Burg et al., 2024 and Schoonakker et al., 2025; the healthy-adult trial is Wei et al., 2017.

Improved Blood Glucose Control

Glycaemic markers respond more consistently than any other outcome. The only meta-analysis restricted to this diet pooled eleven randomised trials and found a clear fall in HbA1c (glycated haemoglobin, a measure of average blood sugar over roughly three months). The twelve-month primary-care trial found a smaller HbA1c fall that is more striking than it looks, because it occurred while participants were reducing their glucose-lowering medication. Effects are largest in people with raised baseline values and modest to absent in those already in range.

Magnitude: Pooled HbA1c fall of 8.6 mmol/mol (95% CI −12.4 to −4.8, roughly 0.8 percentage points) across eleven randomised trials in Mohammadzadeh et al., 2025; −3.2 mmol/mol (−0.3 percentage points) at twelve months alongside reduced medication in van den Burg et al., 2024.

Lower Blood Pressure

Blood pressure falls during and after cycles, an effect visible both in the pooled randomised evidence and in the individual trials that reported it. The mechanism is plausibly a combination of sodium and fluid loss during the diet days and the sustained reduction in visceral fat and insulin between them. The pooled effect is modest at population level but clinically meaningful for someone sitting just above threshold, and it is one of the few outcomes replicated across independent research groups.

Magnitude: Pooled systolic reduction 4.1 mmHg (95% CI −7.6 to −0.7) and diastolic 2.3 mmHg (95% CI −4.2 to −0.4) across randomised trials in Mohammadzadeh et al., 2025, consistent with the blood-pressure fall in Wei et al., 2017.

Reduced Liver Fat and Liver Inflammation Markers

Liver fat responds strongly and is measurable non-invasively. In the twelve-month primary-care trial, magnetic resonance imaging showed both the liver fat fraction and an imaging marker of liver inflammation and fibrosis improving against usual care, and more participants crossed from the high-risk to the low-risk band for fatty liver disease. A separate analysis of two earlier trials in adults without diabetes found the same direction, with the largest reductions in those already above the fatty-liver threshold.

Magnitude: Liver fat fraction −2.8 percentage points (95% CI −4.7 to −0.8) and the imaging inflammation/fibrosis marker −29.9 ms (95% CI −51.8 to −8.0) at twelve months in van den Burg et al., 2025, with concordant findings in Brandhorst et al., 2024.

Medium 🟩 🟩

Reduced Need for Glucose-Lowering Medication

The most practically consequential finding for anyone managing early type 2 diabetes comes from a single twelve-month randomised trial in Dutch primary care. Participants using metformin alone or diet alone added monthly diet cycles to usual care; their medication burden fell while their HbA1c improved, and a composite judgement combining both moved decisively in their favour. It is one trial, in one health system, in a population deliberately restricted to low-intensity treatment, and it has not been replicated.

Magnitude: Medication effect score −0.3 (95% CI −0.4 to −0.2) at twelve months; glycaemic management judged improved in 53% of diet participants versus 8% of controls, and deteriorated in 23% versus 59% (van den Burg et al., 2024).

Improved Smell and Taste Function

A 102-person randomised crossover trial in adults with overweight or obesity found broad gains in smell and taste acuity across six monthly cycles, alongside the expected cardiometabolic changes. Chemosensory decline is common with excess weight and tracks reduced diet quality and appetite dysregulation, so the finding is mechanistically coherent rather than incidental. It rests on a single trial with an investigator group linked to the diet’s developer, and the outcome has not been examined elsewhere.

Magnitude: The proportion of participants with impaired smell fell from 38.1% at baseline to 6.4% after six cycles (Micarelli et al., 2025).

Clinical Response in Mild-to-Moderate Crohn’s Disease

An open-label randomised trial at a university medical centre, co-authored by the diet’s developer, tested three monthly cycles against continued baseline diet in adults with mild-to-moderate Crohn’s disease. Both symptom response and remission favoured the diet, and a stool marker of gut inflammation moved in the same direction, which makes the result harder to attribute to reporting bias in an unblinded design. Relevance beyond diagnosed inflammatory bowel disease is untested.

Magnitude: Clinical response in 69.2% of diet participants versus 43.8% of controls (P = 0.03, the probability that a difference this large arose by chance); faecal calprotectin (a stool marker of gut inflammation) fell 22.0% versus an 8.0% rise (Kulkarni et al., 2026).

Low 🟩

Improved Blood Lipids ⚠️ Conflicted

Individual trials report lower triglycerides and total cholesterol, concentrated in participants whose values were abnormal at baseline, but the pooled analysis of randomised trials found no significant effect on any lipid fraction. Net reading: any lipid benefit is confined to people starting outside the reference range.

Magnitude: No pooled effect on total, low-density or high-density lipoprotein cholesterol or on triglycerides across eleven randomised trials (Mohammadzadeh et al., 2025); the literature reports no outcome figure for the at-risk subgroup identified in Wei et al., 2017.

Enhanced Chemotherapy Response and Tolerability ⚠️ Conflicted

A randomised phase 2 breast-cancer trial found more tumour responses when the diet surrounded chemotherapy, without extra toxicity. The pathological result held only on per-protocol analysis, and adherence was poor. Net reading: a genuine signal that no adequately powered trial has confirmed.

Magnitude: Odds ratio (a measure of how much more likely an outcome is in one group than another) 3.17 for radiological complete or partial response (P = 0.039) and 4.11 for 90–100% tumour-cell loss on per-protocol analysis (P = 0.016) in 131 patients (de Groot et al., 2020).

Improved Skin Hydration and Texture

Three monthly cycles raised instrument-measured facial skin hydration and prevented the rise in skin roughness seen in controls, in a single 45-woman randomised trial co-authored by staff of the company selling the diet. Self-reported happiness and confidence also improved. No independent replication exists.

Magnitude: Skin hydration rose at day 11 (P = 0.00013) and day 71 (P = 0.02) against baseline, with roughness maintained versus an increase in controls (P = 0.032) (Maloh et al., 2023).

Reduced Albuminuria (Protein Leaking into the Urine) in Diabetic Kidney Disease

A post-hoc analysis of a six-month randomised trial in type 2 diabetes with kidney involvement found that participants whose urinary protein improved showed a sustained shift towards fat oxidation and amino-acid turnover, which non-responders did not. The responder split was defined after the fact, so the finding is hypothesis-generating.

Magnitude: Not quantified in available studies. The published analysis reports metabolic pathway shifts by responder status rather than an absolute or relative change in urinary albumin for the randomised group as a whole (Sulaj et al., 2025).

Reduced Multiple Sclerosis Symptoms

In mouse models of multiple sclerosis, a disease in which the immune system strips the insulation from nerve fibres, cycles reduced severity in every animal and prompted that insulation to regrow. The human arm was reported only as safe, feasible and linked to better quality of life.

Magnitude: Not quantified in available studies. The human arm was a preliminary pilot in relapsing-remitting patients reporting feasibility and quality-of-life change rather than a relapse or disability effect estimate (Choi et al., 2016).

Speculative 🟨

Lifespan and Healthspan Extension

In mice, cycles begun in middle age extended lifespan, roughly halved tumour incidence and slowed bone-density loss (Brandhorst et al., 2015). No human survival data exist; the basis is animal work only.

Reduction in Biological Age Estimates

Secondary analyses of two trials reported a median 2.5-year fall in a blood-marker age score (Brandhorst et al., 2024). The score predicts mortality observationally but is not validated as a modifiable target.

Immune Cell Turnover and Stem-Cell-Based Regeneration

Mouse work shows white-blood-cell depletion during the diet and stem-cell-driven rebuilding on refeeding (Brandhorst et al., 2015). Human data extend only to a shifted lymphoid-to-myeloid ratio; no clinical immune outcome has been measured.

Improved Cognitive Performance

In old mice, cycles grew new neurons in the brain’s memory centre and improved motor learning and memory (Brandhorst et al., 2015). No human trial has measured cognition, so the basis is animal work only.

Benefit-Modifying Factors

  • Baseline metabolic abnormality: The single strongest modifier. Post-hoc analysis of the healthy-adult trial found blood pressure, glucose, growth factor, lipids and inflammatory markers improved substantially more in participants who began above risk thresholds than in those already in range.
  • Adiposity and fat distribution: Absolute weight and visceral-fat losses scale with starting fat mass. Someone at a body mass index of 22 with little abdominal fat has less to lose and correspondingly smaller metabolic movement.
  • Genetic variation in nutrient handling: Variants in FTO (fat mass and obesity-associated gene, which influences appetite) and TCF7L2 (a gene governing insulin secretion) shift how strongly weight and glucose respond to any energy-restriction protocol, though no trial has stratified this diet by genotype.
  • Sex differences: Trials enrolled both sexes without pre-specified sex analysis. Women carry proportionally more subcutaneous and less visceral fat, so visceral-fat endpoints tend to move less; the skin trial recruited women only, so its result is untested in men.
  • Age: Metabolic gains appear across the adult range, but people past roughly 70 have less anabolic reserve, so the same energy deficit buys the same metabolic benefit at a higher potential cost in lean tissue.
  • Pre-existing conditions: Diagnosed inflammatory bowel disease, fatty liver and type 2 diabetes are the conditions with direct trial evidence of larger gains. Established cardiovascular disease and cancer have only adjunct-setting data.
  • Habitual protein intake and training status: People eating well above protein requirements and resistance-training regularly show smaller relative drops in growth signalling, which may blunt the diet’s mechanistic effect while protecting lean mass.

Potential Risks & Side Effects

High 🟥 🟥 🟥

Transient Fatigue, Weakness and Headache

Tiredness, weakness and headache are the commonest complaints. They cluster on days two and three as the body switches to burning fat, and they resolve within a day of refeeding. Randomised trials in generally healthy adults recorded no serious adverse events across three cycles, and a single-arm study in breast-cancer patients receiving chemotherapy recorded a small proportion of severe events per cycle. Most events are graded mild, but they are sufficient to interfere with demanding cognitive or physical work on the diet days.

Magnitude: Grade III or worse diet-related adverse events in 5.95% (5 of 84) of cycles in a 30-patient chemotherapy study (Xue et al., 2025); no serious adverse events in the 100-person healthy-adult trial (Wei et al., 2017).

Gastrointestinal Symptoms and Hunger

Hunger, nausea, abdominal discomfort and constipation are reported during the diet days, driven by low food volume, the fibre profile of the packaged version and the slowing of gut motility that accompanies low energy intake. Hormonal data give the mechanism a floor: fasting regimens lower leptin and ghrelin without a matching rise in adiponectin, so appetite signalling is disrupted rather than suppressed. Symptoms are graded mild in most participants and settle within a day of refeeding.

Magnitude: Mild complaints predominate and cluster on diet days two and three across randomised trials (Wei et al., 2017; Micarelli et al., 2025), which report these as tolerability observations and give no pooled incidence figure.

Medium 🟥 🟥

No risk sits at Medium: every adverse event this diet has produced was documented in more than one randomised trial and therefore grades High, while the remaining risks rest on uncontrolled, indirect or conflicting human data.

Low 🟥

Loss of Fat-Free Mass ⚠️ Conflicted

A large energy deficit can cost lean tissue: a meta-analysis of intermittent versus continuous restriction found greater fat-free mass loss. Diet-specific trials found abdominal muscle area and maximal force unchanged. Net reading: loss is not established for this protocol but cannot be excluded in untested sarcopenic (muscle-depleted) older people.

Magnitude: −0.20 kg fat-free mass for intermittent versus continuous energy restriction (95% CI −0.39 to −0.01) in Schroor et al., 2024; abdominal muscle area change with this diet −1.6 cm² (95% CI −4.6 to 1.4), not distinguishable from zero (Schoonakker et al., 2025), with muscle function preserved in Nardon et al., 2022.

Hypoglycaemia in People Taking Insulin or Sulfonylureas

Trials in type 2 diabetes enrolled only participants on metformin or diet alone, precisely because five days near 750 kilocalories alongside insulin or a sulfonylurea (a drug class that forces insulin release) can drive blood sugar dangerously low. The risk is inferred, not measured.

Magnitude: Not quantified in available studies. No trial of this diet has enrolled participants using insulin or sulfonylureas, so no event rate exists for that combination (van den Burg et al., 2024).

Symptomatic Low Blood Pressure and Light-Headedness

The blood-pressure fall that counts as a benefit in hypertensive participants can produce dizziness on standing in people whose blood pressure is already normal or who take antihypertensive (blood-pressure-lowering) drugs. Sodium and fluid loss over the first two days compounds it, and the effect resolves on refeeding.

Magnitude: Pooled systolic blood pressure falls 4.1 mmHg (95% CI −7.6 to −0.7) across randomised trials (Mohammadzadeh et al., 2025); the literature reports no incidence figure for symptomatic light-headedness.

Gallstone Formation with Repeated Rapid Weight Loss

Sustained very-low-calorie dieting raises gallstone incidence sharply through bile stasis and cholesterol supersaturation. Whether five-day cycles separated by weeks of normal eating carry the same risk is untested; the concern applies mainly to people cycling frequently for continuous weight loss.

Magnitude: New gallstones form within four weeks at rates 15–25 times those in the comparable population with obesity during sustained very-low-calorie weight loss, becoming symptomatic in about one third of those affected (Weinsier & Ullmann, 1993).

Trigger or Worsening of Disordered Eating

A rigid, rule-bound five-day protocol with a defined refeeding day can entrench restrictive patterns or restrict-then-binge cycles in people with a history of eating disorders. A review of restrictive anticancer diets found no diet-related distress in the trials it examined, but noted that long-term psychological assessment is missing.

Magnitude: Not quantified in available studies. People with current or past eating disorders are excluded from fasting trials, so no controlled incidence data have been generated (Da Prat et al., 2024).

Speculative 🟨

Blunted Immune Response to Infection During the Diet Window

Mouse work shows a large transient fall in circulating white blood cells during the diet. Whether that creates a real window of vulnerability to infection in humans has never been measured.

Risk-Modifying Factors

  • Glucose-lowering and antihypertensive medication: The single largest risk modifier. Insulin, sulfonylureas and multi-drug blood-pressure regimens turn an expected metabolic effect into a hypoglycaemic or hypotensive event during the five diet days.
  • Baseline lean mass and grip strength: Someone with low appendicular lean mass or reduced grip strength has less reserve to lose, making the contested fat-free-mass signal materially more consequential than it is in a resistance-trained adult.
  • Sex differences: Women enter with lower absolute lean mass and lower bone mineral density, so the same relative loss carries greater functional cost; men more often carry the visceral fat that drives the benefit side.
  • Genetic variation in bile handling: Variants in ABCG8 (a transporter that pumps cholesterol into bile) raise gallstone risk independently and would compound the risk of repeated rapid weight loss.
  • Age: Beyond roughly 70, anabolic resistance means refeeding rebuilds lean tissue less completely, and the fall in blood pressure is more likely to cause falls through orthostatic symptoms (dizziness on standing).
  • Pre-existing conditions: Prior gallstones, chronic kidney disease, a history of eating disorders, active cancer cachexia (severe wasting) and pregnancy each shift the risk profile, and each was an exclusion criterion in at least one published trial.
  • Baseline body mass index below 20: Participants below this threshold were excluded from the oncology trials because the expected weight loss carries no benefit and a measurable cost.

Key Interactions & Contraindications

  • Insulin and sulfonylureas (glibenclamide, gliclazide, glimepiride): Absolute contraindication without physician-supervised dose reduction. Consequence: severe hypoglycaemia. Mitigation: trials excluded these agents; clinical programmes reduce or suspend the dose across diet days with capillary glucose monitoring.
  • SGLT2 inhibitors (sodium-glucose cotransporter 2 inhibitors, which make the kidney excrete glucose; empagliflozin, dapagliflozin, canagliflozin): Caution. Combined with the diet’s ketogenic shift they raise the risk of euglycaemic ketoacidosis (a dangerous acid build-up at normal blood sugar). Mitigation: suspension across diet days.
  • Antihypertensives (lisinopril, amlodipine, hydrochlorothiazide): Caution. Consequence: symptomatic hypotension and falls, especially with diuretics given the fluid loss on days one and two. Mitigation: blood pressure measured before and after each cycle, with dose review.
  • Metformin: Monitor only. It carries little hypoglycaemia risk alone and was the sole permitted glucose-lowering drug in the twelve-month primary-care trial. Gastrointestinal side effects can compound the diet’s own nausea; splitting doses with the largest meal helps.
  • Levothyroxine and narrow-therapeutic-index drugs (warfarin, lithium, digoxin): Caution. Consequence: altered absorption and altered volume of distribution as fluid and food intake change. Mitigation: consistent dosing times and a check of the relevant level after the first cycle.
  • Over-the-counter NSAIDs (non-steroidal anti-inflammatory drugs; ibuprofen, naproxen, aspirin): Caution. Consequence: gastric irritation and, with reduced fluid intake, added kidney strain. Mitigation: taken with the largest meal of the diet day, or deferred until refeeding.
  • Over-the-counter caffeine and stimulant preparations: Monitor. Consequence: amplified palpitations, jitteriness and light-headedness against a falling blood pressure. Mitigation: habitual intake is maintained rather than raised, since abrupt withdrawal itself causes headache.
  • Glucose-lowering supplements (berberine, chromium picolinate, alpha-lipoic acid, cinnamon extract): Caution, additive. Consequence: additive hypoglycaemia with an already low carbohydrate intake. Mitigation: suspended across the five diet days.
  • Blood-pressure-lowering supplements (beetroot nitrate, magnesium, potassium, hibiscus, omega-3): Caution, additive. Consequence: additive hypotension with the diet’s own 4 mmHg systolic fall. Mitigation: separation from the diet window or dose reduction.
  • GLP-1 receptor agonists (glucagon-like peptide-1 mimics of a gut hormone that suppresses appetite; semaglutide, tirzepatide): Caution. Consequence: compounded nausea, further reduced intake and accelerated lean-mass loss. Mitigation: cycles are not run during dose escalation.
  • Endurance and resistance training: Caution rather than contraindication. Consequence: reduced glycogen availability lowers high-intensity performance and can blunt training adaptation. Mitigation: heavy sessions are scheduled outside diet days.

Populations who should avoid Fasting-Mimicking Diet:

  • Pregnancy and breastfeeding
  • Body mass index below 18.5, or documented sarcopenia or frailty
  • Current or past eating disorder
  • Type 1 diabetes, or type 2 diabetes managed with insulin or a sulfonylurea without physician supervision
  • Chronic kidney disease at stage 4 or worse (eGFR — estimated glomerular filtration rate, a measure of kidney function — below 30 mL/min/1.73 m²)
  • Active cancer with weight loss or cachexia, outside a supervised trial
  • Age under 18 or over 75 without individual clinical assessment
  • Liver impairment at Child-Pugh Class C (the most severe grade of liver-function loss), or heart failure at NYHA Class IV (New York Heart Association’s most severe grade, with symptoms at rest)
  • Recent myocardial infarction or stroke (<90 days)

Risk Mitigation Strategies

  • Medication review before the first cycle: Prevents hypoglycaemia and symptomatic hypotension by identifying insulin, sulfonylureas, SGLT2 inhibitors and diuretics that need suspension or dose reduction across the five diet days.
  • Capillary glucose checks twice daily during diet days: For anyone on any glucose-lowering agent, catches the fall before it becomes symptomatic; a reading below 3.9 mmol/L (70 mg/dL) is the conventional threshold for stopping.
  • Blood pressure measured seated and standing before and after each cycle: Detects the orthostatic drop that causes falls, particularly relevant for anyone on two or more antihypertensive drugs.
  • Protein and resistance training in the refeeding window: Mitigates the contested fat-free-mass loss. Protocols typically resume 1.2–1.6 g of protein per kilogram of body weight from day six, with two resistance sessions in the following week.
  • Cycle frequency capped at monthly, and at quarterly once targets are met: Limits cumulative energy deficit and reduces the gallstone risk associated with sustained rapid weight loss; trials used three to twelve monthly cycles, not continuous use.
  • Fluid and electrolyte intake of 2–2.5 litres daily with added sodium: Prevents the headache, dizziness and constipation that dominate the adverse-event profile, most of which reflect volume depletion rather than energy restriction.
  • Avoidance of high-intensity training and driving-intensive days during days two to four: Mitigates the fatigue, weakness and light-headedness that peak as ketone production rises.
  • Screening for eating-disorder history before starting: Prevents entrenchment of restrictive or binge-refeed patterns; a validated screening questionnaire takes minutes and no trial has enrolled this group.

Therapeutic Protocol

  • Standard five-day cycle: The protocol used in nearly every trial delivers about 1,100 kilocalories on day one and 725–800 on days two to five, plant-based, with protein near 10% of energy and fat supplying 44–55%.
  • Packaged versus food-based delivery: The boxed kit, developed at the University of Southern California Longevity Institute and sold by L-Nutra, is the version used in most trials. Recipe-based equivalents matched on macronutrients performed comparably in a Chinese oncology cohort.
  • Competing approach — supervised modified fasting: European clinics in the Buchinger tradition run 200–400 kilocalorie broth-and-juice regimens for 5–14 days under inpatient supervision. Neither approach has been shown superior; they differ in cost, duration and medical oversight.
  • Competing approach — continuous calorie restriction: Pooled trials find intermittent and continuous restriction broadly equivalent on cardiometabolic markers, so the choice turns on adherence and scheduling rather than on demonstrated superiority.
  • Cycle frequency: Trials used three monthly cycles in healthy adults, twelve monthly cycles in type 2 diabetes, and three in Crohn’s disease. Maintenance schedules of one cycle every three to six months are common in practice.
  • Time of day: Daily rations are distributed across the waking day rather than compressed, typically breakfast, lunch, an afternoon soup and an evening tea, which limits the overnight hypoglycaemic dip and improves sleep on diet nights.
  • Metabolic time course rather than drug half-life: The diet is not a compound with a half-life. Ketone levels rise from about hour 36, peak on days four to five, and return to baseline within 24–48 hours of refeeding.
  • Single versus divided intake: Divided intake is standard. No trial has tested the same energy delivered as one meal per day; the divided pattern is what all efficacy and safety data describe.
  • Genetic considerations: No pharmacogenetic testing applies, since no drug is involved. Variants in FTO and TCF7L2 modify weight and glucose response to energy restriction generally, but no trial has selected or stratified participants by genotype.
  • Sex-based considerations: Trials enrolled both sexes without pre-specified sex analysis. Women may need the lower end of any recipe-based energy target for equivalent relative restriction, given lower average resting energy expenditure.
  • Age-based considerations: Above roughly 70, practitioners typically extend the interval between cycles to three months and raise refeeding protein, to offset the reduced capacity to rebuild lean tissue.
  • Baseline biomarker considerations: Response scales with baseline abnormality. Fasting glucose, HbA1c, blood pressure and liver fat above threshold predict the largest movement; values already in range predict little.
  • Pre-existing condition considerations: Type 2 diabetes on metformin alone, fatty liver disease and mild-to-moderate Crohn’s disease have direct trial protocols. Every other condition is extrapolation from those populations.
  • Refeeding day: Day six is a transition day of soups, vegetables and modest complex carbohydrate before normal eating resumes, which trials used to limit gastrointestinal upset and the sharp glucose rebound.

Discontinuation & Cycling

  • Not a lifelong regimen: The protocol is periodic by design. Every trial applied a fixed number of five-day cycles separated by unrestricted eating; no study has tested continuous or indefinite use.
  • Cycling is the intervention, not an optimisation: Benefits in the trials accrued across three to twelve cycles. A single cycle produces the acute metabolic shift but not the sustained weight, liver-fat or HbA1c changes.
  • No withdrawal syndrome: Stopping produces no rebound phenomenon. Metabolic markers drift back toward baseline over months in proportion to whether weight is regained, not through any dependence mechanism.
  • No taper required: The five-day format ends at a defined point. The only transition step is the refeeding day, which limits gastrointestinal upset rather than preventing withdrawal effects.
  • Between-cycle diet determines durability: Twelve-month benefits in the primary-care trial coincided with participants making self-initiated improvements to their everyday eating; cycles alone against an unchanged background diet have not been shown to hold gains.
  • Discontinuation triggers: Persistent fatigue lasting beyond the refeeding day, unintended loss of lean mass on repeat body composition scanning, new gallbladder symptoms, or any hypoglycaemic episode are the practical stopping points.
  • Maintenance spacing: Once metabolic targets are reached, spacing cycles to every three to six months is the common practice pattern; the twelve-month trial data cover monthly use, so wider spacing is extrapolation.

Sourcing and Quality

  • Packaged kit versus self-assembled: The boxed kit standardises the macronutrient split and micronutrient content across five days. Self-assembly is far cheaper but reproduces the calorie target more easily than the 10% protein and micronutrient completeness.
  • What to look for in a packaged product: A published macronutrient breakdown per day, added micronutrients and omega-3, absence of animal protein, and manufacture under current Good Manufacturing Practice. Kits without a per-day breakdown cannot be matched to the trial protocol.
  • Third-party testing: Food kits are not subject to the identity and potency testing applied to supplements. Certification worth checking is the manufacturing standard and allergen control rather than assay of an active ingredient.
  • Named products: ProLon, from L-Nutra, is the kit used in the great majority of published trials and is the only version with direct efficacy data. Generic five-day fasting kits have no trial evidence.
  • Compounding pharmacies: Not applicable — no compounded preparation exists. Some pharmacies and clinics resell the branded kit, which does not alter its composition.
  • Self-assembled ingredient quality: Where recipes are used, vegetable soups, olives, nuts, olive oil and herbal teas are the trial-consistent components; a recipe-based version matched on energy and macronutrients performed comparably in an oncology cohort.
  • Micronutrient adequacy: Five days at 750 kilocalories cannot meet reference intakes for several micronutrients. Trial kits address this with an added multivitamin and mineral sachet, which self-assembled versions routinely omit.

Practical Considerations

  • Time to effect: Ketone rise and glucose fall appear within 36–48 hours of a single cycle. Weight, liver fat and HbA1c changes reported in trials required three cycles, and the medication-reduction result required twelve months.
  • Common pitfall — treating one cycle as sufficient: The trial benefits are cumulative across cycles. A single five-day round produces the acute metabolic signature and little of the durable outcome.
  • Common pitfall — compensatory overeating on refeeding: Leptin and ghrelin both fall during fasting regimens, disrupting satiety signalling. Weight regained in the days after a cycle erases most of the between-cycle benefit.
  • Common pitfall — ignoring the background diet: Twelve-month gains coincided with everyday dietary change. Cycles layered on an unchanged high-energy diet have not been shown to sustain any outcome.
  • Regulatory status: The packaged kit is sold as food, not as a drug or approved medical device. It has not been evaluated by regulators for efficacy, and the marketed claims are structure-function claims rather than approved indications.
  • Cost and accessibility: A branded five-day kit runs roughly $150–250, so monthly cycles cost more than most generic chronic medications over a year. Self-assembly cuts this substantially at the price of protocol fidelity.
  • Payer incentives and structural bias: Insurers have an evident incentive to favour generic metformin, at cents per day, over a repeated paid kit, and an equally evident one to favour the kit over costlier GLP-1 receptor agonists. Both pressures shape which comparisons get funded.
  • Scheduling burden: Five days of markedly reduced energy is incompatible with heavy physical work, long-haul travel and high-stakes cognitive tasks, which is the practical reason adherence in trials was imperfect.

Interaction with Foundational Habits

  • Sleep: Direct and mostly negative during the cycle. Falling glucose and rising noradrenaline fragment sleep on nights two and three, and hunger delays sleep onset. Distributing the daily ration so that a portion falls in the evening, rather than front-loading it, reduces this. Sleep normalises within a night of refeeding.
  • Nutrition: Direct and defining — the diet is a nutrition intervention. It supplies inadequate protein and several micronutrients across the five days, so the surrounding diet carries the whole nutritional load. Trial kits include a micronutrient sachet. Alcohol is excluded during cycles, since it displaces the limited energy budget.
  • Exercise: Blunting for high-intensity and resistance work, neutral for low-intensity activity. Glycogen depletion reduces power output and may attenuate training adaptation, though a randomised trial in young men found no loss of muscle volume or maximal force after three cycles. Heavy sessions are conventionally scheduled outside diet days; walking is well tolerated.
  • Stress management: Indirect and bidirectional. Energy restriction raises cortisol and lowers stress tolerance during the cycle, while the reported gains in mood and self-reported confidence appear after refeeding. Running cycles during periods of high psychological load compounds both the perceived difficulty and the dropout rate seen in trials.

Monitoring Protocol & Defining Success

Baseline testing is taken before the first cycle and is what makes every later reading interpretable, because the size of the response tracks how abnormal the starting values were. A fasting panel — glucose, insulin, HbA1c, a full lipid panel, liver enzymes, kidney function and electrolytes — together with high-sensitivity C-reactive protein and seated blood pressure gives the working baseline; body composition by DEXA (dual-energy X-ray absorptiometry, a scan that separates fat, lean tissue and bone) adds a lean-mass reference that matters if cycles are repeated often. Ongoing monitoring runs on a slower clock than the diet itself: the core panel is repeated after the third cycle, then every six months while cycling continues, with blood pressure measured before and after each cycle in anyone taking antihypertensive medication and capillary glucose measured twice daily during the diet in anyone on glucose-lowering medication.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Fasting glucose 75–86 mg/dL (4.2–4.8 mmol/L) Primary responsive marker; falls within 48 hours Conventional range extends to 99 mg/dL; requires a 12-hour fast
HbA1c 4.8–5.3% Average blood sugar over ~3 months; the endpoint with pooled trial evidence Conventional cut-off for prediabetes is 5.7%; no fasting needed; unreliable in anaemia
Fasting insulin 2–5 µIU/mL Detects insulin resistance earlier than glucose Most labs report no functional target at all; best paired with glucose to compute HOMA-IR (a calculated index of insulin resistance)
hs-CRP <1.0 mg/L Tracks body-wide inflammation, which falls with visceral fat hs-CRP is high-sensitivity C-reactive protein, a general inflammation marker; conventional threshold is <3.0 mg/L; invalid within two weeks of infection or injury
IGF-1 Mid-normal for age (roughly 100–160 ng/mL at 40–70 years) Insulin-like growth factor 1, the growth signal the diet is designed to lower Falls sharply during cycles and rebounds on refeeding; measure at least 10 days post-cycle
Triglycerides <80 mg/dL Tracks liver fat and carbohydrate handling Conventional cut-off is <150 mg/dL; requires a 12-hour fast; highly sensitive to the prior evening’s meal
LDL-C <100 mg/dL The main artery-damaging cholesterol fraction LDL-C is low-density lipoprotein cholesterol; conventional target is <130 mg/dL in low-risk adults; can rise transiently during active fat mobilisation
ALT <25 U/L (men), <20 U/L (women) Liver enzyme that tracks liver fat, the outcome with the strongest trial signal ALT is alanine aminotransferase; conventional upper limits are roughly double these; best paired with GGT (gamma-glutamyl transferase, another liver enzyme)
eGFR >90 mL/min/1.73 m² Estimated glomerular filtration rate; screens for the kidney impairment that contraindicates cycling Conventional threshold for concern is <60; creatinine-based estimates are distorted by high muscle mass
Potassium and sodium Potassium 4.0–4.5 mmol/L; sodium 138–142 mmol/L Detects the depletion that drives headache, cramp and light-headedness Conventional ranges are wider; most informative if drawn on day four or five of a cycle
Beta-hydroxybutyrate 0.5–2.0 mmol/L on days 3–5 Confirms the intended metabolic switch actually occurred Capillary meter at home; morning reading before the first ration; not a conventional clinical test
Appendicular lean mass (DEXA) No established target for this diet; tracked as change from the individual’s own baseline The contested fat-free-mass outcome; the only way to settle it individually Repeat on the same machine; hydration status shifts readings, so measure at least 5 days post-cycle
Blood pressure <120/80 mmHg seated Both a benefit endpoint and the source of the orthostatic risk Also measured standing after 3 minutes to catch the postural drop
25-hydroxyvitamin D 40–60 ng/mL Low status compounds fatigue and bone loss over repeated cycles Conventional sufficiency starts at 30 ng/mL; not fasting-dependent

Qualitative markers worth tracking alongside the labs:

  • Energy and fatigue on days two to four, and how quickly they resolve after refeeding
  • Sleep continuity on diet nights compared with normal nights
  • Hunger intensity and whether refeeding tips into compensatory overeating
  • Cognitive clarity and tolerance for demanding work during the cycle
  • Grip strength and perceived training performance in the week after each cycle
  • Mood, irritability and stress tolerance across the cycle
  • Gastrointestinal comfort, including constipation during and after the diet days
  • Waist circumference measured at the same point after each third cycle

Emerging Research

  • Ulcerative colitis trial: NCT03615690, Stanford University, recruiting, 75 participants, primary completion December 2027. Tests whether monthly cycles induce response in ulcerative colitis, extending the Crohn’s disease result to the other main form of inflammatory bowel disease.
  • Colorectal cancer prognosis trial: NCT05384444, Fudan University, 602 participants, primary completion October 2027 — the largest oncology trial of this diet, with long-term prognosis endpoints after radical resection. The protocol is published as Guo et al., 2026.
  • Independent biological-age replication: NCT07354620, The Christ Hospital REVERSE study, early phase 1, 52 participants, randomising six monthly cycles against low-dose rapamycin and the combination on a methylation age clock. Tests whether the signal in Brandhorst et al., 2024 reproduces outside the developer’s group.
  • Protein content may not need to be low: Burns et al., 2025 found a high-protein version outperformed the standard low-protein version on visceral fat, triglycerides, heart-rate variability and gut microbiome diversity, which would undercut the growth-signalling rationale for keeping protein at 10%.
  • Surgical preconditioning outside metabolic medicine: NCT05709600, University of Cologne, recruiting, 80 participants. Five days of the diet before living-donor kidney removal, against ketogenic and amino-acid-restricted comparators, with molecular endpoints in pre-transplant kidney biopsies rather than clinical outcomes.
  • Evidence that could weaken the case: pooled comparisons of intermittent against continuous energy restriction (Schroor et al., 2024) find no cardiometabolic advantage for intermittent patterns and slightly greater fat-free-mass loss, so calorie-matched trials may show the protocol adds nothing beyond its energy deficit.
  • Independent oncology safety data: Xue et al., 2025 delivered a recipe-based version without the commercial kit and reported tolerability and body-composition outcomes, a template for testing the protocol free of the manufacturer’s supply chain.

Conclusion

The fasting-mimicking diet is a short, repeatable eating plan rather than a continuous way of eating, and the evidence reflects that design. The most consistent findings across controlled human studies are reductions in body weight and abdominal fat, lower blood sugar, lower blood pressure and less fat in the liver, with the largest changes in people who started with the worst readings and little movement in those already in a healthy range. Findings on blood fats are mixed. Results for inflammatory bowel disease, chemotherapy tolerance and skin condition rest on single or conflicting trials. Claims about slowing ageing itself rest on animal work and on composite scores built from blood markers, not on human survival data.

Harms recorded in trials have been mostly short-lived: tiredness, weakness, headache and hunger during the diet days. Loss of muscle is the contested question, with pooled analyses of related eating patterns pointing one way and diet-specific measurements the other. Risk concentrates in people taking blood-sugar or blood-pressure medication, in those already thin or frail, and in anyone with a history of disordered eating.

The evidence base is unusual in that one commercial developer funded, supplied or co-authored a large share of the human trials, and the scientist who devised the plan holds a financial stake in it. Independent replication outside that network remains limited, which bears on how confidently modest average effects can be read against a paid, repeated protocol.

Top - Benefits - Risks - Protocol