evipedia.ai
← Back

Full Fasting for Health & Longevity

Evidence Review created on 10/30/2026 using AI4L / Opus 5.5

Also known as: Water Fasting, Water-Only Fasting, Complete Fasting, Total Fasting, Zero-Calorie Fasting, Complete Caloric Restriction, Therapeutic Starvation

Motivation

Full fasting means going without all food and caloric drinks for a set period, from a single day to several weeks, while still drinking water. It draws interest because, once stored sugar runs low, the body switches to burning its own fat for fuel, a change studied for its possible effects on body weight, blood pressure and the body’s own repair processes.

Fasting is one of the oldest religious and healing practices. In the mid-twentieth century, hospitals used long supervised fasts to treat severe obesity, and residential clinics in Europe and the United States still offer multi-day water-only fasts. Laboratory research on fasting and aging in animals, together with recent human studies tracking thousands of blood proteins during a week without food, has renewed attention among people pursuing a longer healthy life.

This review examines the human evidence on the benefits and risks of full fasting, from repeated one- to two-day fasts to supervised fasts lasting weeks, how these protocols are carried out, and where full fasting fits within a demanding long-term health and longevity strategy.

Benefits - Risks - Protocol - Conclusion

This section lists expert commentary, interviews and a narrative review that discuss full or prolonged water-only fasting in depth.

No content from Andrew Huberman dedicated to full fasting was found; his Huberman Lab fasting episode addresses time-restricted eating (eating within a daily time window), not zero-calorie fasts lasting a day or longer. Chris Kresser’s fasting content centers on intermittent fasting (regular short fasts of hours to a day) and keto fasting (fasting periods combined with a very-low-carbohydrate diet); his Revolution Health Radio episode with Joseph Mercola touches multi-day water fasting only briefly.

Grokipedia

  • Water Fasting

    An AI-written overview of water-only fasting covering protocols, physiological effects, benefits and risks, research and safety guidance; useful for orientation, though its claims need checking against primary studies.

Examine

  • Prolonged Fasting

    Examine’s independent evidence page on abstaining from food for 2 to 21 days; its database grades one outcome, cognition (no effect, grade D, 31 studies), with study details reserved for members.

ConsumerLab

No ConsumerLab article on full, water-only or prolonged fasting exists. Its fasting-related items are a CL Answer, Are there supplements I should take, or take differently, when intermittent fasting?, and a brief Clinical Update on thiamin deficiency after extreme water-only fasting.

Systematic Reviews

This section lists systematic reviews and meta-analyses of full or prolonged fasting covering body composition, blood lipids, chemotherapy tolerance, rheumatoid arthritis and type 2 diabetes.

Mechanism of Action

Full fasting acts through a staged shift in fuel use. Liver glycogen (stored glucose) is largely used up within about a day; the liver then makes glucose from amino acids and glycerol (gluconeogenesis) while fat tissue releases fatty acids that the liver turns into ketone bodies (fat-derived fuels) such as beta-hydroxybutyrate. After several weeks, ketones supply most of the brain’s fuel, which slows muscle protein breakdown (Owen 1967; a review by Cahill 2006).

  • Hormonal signals: Insulin and IGF-1 (insulin-like growth factor 1, a growth-promoting hormone) fall; IGF-1 dropped about 30% in 24- and 48-hour fasts in a trial co-authored by Valter Longo, founder of L-Nutra (Dorff 2016). Low insulin increases urinary sodium and water loss, a proposed cause of falling blood pressure.
  • Cellular stress pathways: Low nutrient signaling is proposed to suppress mTOR (mechanistic target of rapamycin, a nutrient-sensing growth regulator) and activate AMPK (AMP-activated protein kinase, a cellular energy sensor), promoting autophagy (cellular self-recycling); this rests mostly on animal data, as summarized in a review (Longo 2014).
  • Competing explanations: Much of the benefit may reflect the energy deficit itself, since zero-calorie alternate-day fasting (eating normally one day and nothing the next) matched daily calorie restriction for weight and insulin sensitivity (Catenacci 2016). Fasting also raises LDL (low-density lipoprotein, the main cholesterol carrier linked to artery disease) cholesterol, per a meta-analysis (Çamli 2026) and triggers an inflammatory response (Commissati 2025).

Historical Context & Evolution

Fasting was originally a religious and spiritual practice and an old healing tradition; in the 1920s German physician Otto Buchinger popularized supervised “modified” fasts with small amounts of juice and broth.

Medical full fasting peaked in the 1960s, when hospitals used zero-calorie fasts lasting weeks to months to treat severe obesity (Duncan 1962; Thomson 1966), including one supervised fast of 382 days (Stewart 1973). These reports documented large weight losses, but deaths and heart-muscle damage during therapeutic starvation (Spencer 1968; Garnett 1969) and frequent weight regain (Innes 1974) moved clinicians toward protein-containing very-low-calorie diets. The approach was never disproven; its hazards were judged to outweigh its gains when fasts ran for months.

Interest returned through residential water-only fasting for hypertension at TrueNorth Health Center, a clinic that sells the fasting programs it studies (Goldhamer 2001); mouse studies of fasting and blood stem cells (Cheng 2014); and observational links between religious fasting and heart health (Horne 2008). Recent work tracks thousands of blood proteins during fasting (Pietzner 2024). Because full fasting is free and unpatentable, no manufacturer has an incentive to fund outcome trials; institutional payers would gain from a no-cost alternative to drug therapy, yet no payer-funded trial was found, leaving the field to small academic studies and clinic-run case series.

Expected Benefits

High 🟩 🟩 🟩

Weight and Fat-Mass Loss

Zero energy intake produces rapid weight loss in every protocol studied. Randomized trials from two independent groups, Catenacci (University of Colorado) and Stekovic (University of Graz), found that zero-calorie alternate-day fasting reduced weight and fat mass (Catenacci 2016; Stekovic 2019); in Catenacci’s trial the loss did not exceed daily calorie restriction. In multi-day water-only fasts, a meta-analysis found much of the loss is fat-free mass and water (Ulupınar 2026), and most patients in historical fasting programs were classed as failures at 12–64 months (Innes 1974).

Magnitude: −8.2 kg with zero-calorie alternate-day fasting versus −7.1 kg with daily calorie restriction after 8 weeks; the between-group difference was not statistically significant, and no 95% CI (confidence interval, the range likely to contain the true value) is reported in the abstract (Catenacci 2016).

Medium 🟩 🟩

Blood Pressure Reduction

In a 4-week randomized trial from the University of Graz, zero-calorie alternate-day fasting lowered systolic and diastolic blood pressure compared with unrestricted eating in healthy non-obese adults (Stekovic 2019). Residential case series from TrueNorth Health Center report larger falls after about 10–14 days of water-only fasting (Goldhamer 2001), sustained for six weeks on a plant-based diet (Zeiler 2024); a Chinese five-day trial agrees (Jiang 2021). These multi-day studies had no non-fasting control group, and the randomized evidence comes from one research group.

Magnitude: Average fall of 37/13 mmHg in 174 hypertensive patients after a mean 10–11 days of water-only fasting plus refeeding, no control group (Goldhamer 2001).

Low 🟩

Improved Insulin Sensitivity ⚠️ Conflicted

Repeated 20-hour fasts every other day for two weeks raised insulin-driven glucose uptake in eight healthy men (Halberg 2005); a controlled crossover trial of the same two-week duration found no change (Soeters 2009). A review found gains fade after refeeding and vanish in diabetes (Ezpeleta 2024). Net: no consistent improvement.

Magnitude: Insulin-mediated glucose uptake rose from 6.3 to 7.3 mg/kg/min, no control group (Halberg 2005).

Lower Medication Needs in Type 2 Diabetes

A systematic review found that intermittent energy restriction or periodic fasting allowed glucose-lowering medication reduction, including regimens permitting some calories, an indirect form (van den Burg 2023). A case series from a clinic offering fasting programs reports insulin discontinuation (Furmli 2018).

Magnitude: All 3 patients in a clinic case series stopped insulin, no control group (Furmli 2018); in a systematic review, medication could be reduced in 4 of 13 studies, with benefits generally not sustained at one year or later (van den Burg 2023).

Rheumatoid Arthritis Symptom Relief

A systematic review pooled four controlled studies in which a fast was followed by a vegetarian diet and found lasting symptom benefit (Müller 2001). Fasting was always combined with a diet change, an indirect test, and a co-author is affiliated with a commercial fasting clinic.

Magnitude: Direction favors fasting followed by a vegetarian diet over usual diet at three months or longer; the review reports the pooled effect only as statistically and clinically significant, and the accessible literature gives no outcome figure (Müller 2001).

Chemotherapy Tolerance ⚠️ Conflicted

In two small randomized trials, 48-hour fasts around chemotherapy preserved red cell and platelet counts (de Groot 2015) or reduced dose delays (Riedinger 2020). A meta-analysis of nine studies of mixed fasting-like regimens found no fewer side effects (Ferro 2023). Net: no reliable protection is shown.

Magnitude: Pooled RR (relative risk, the ratio of event rates between groups) of chemotherapy side effects 1.10 (95% CI 0.77–1.59) with fasting regimens versus non-fasting (Ferro 2023).

Less Coronary Artery Disease With Routine Religious Fasting

Among 448 patients undergoing coronary angiography (X-ray imaging of heart arteries), those who routinely fasted from food and drink had less coronary artery disease after adjustment (Horne 2008). These fasts excluded water as well as food, an indirect form, and the design cannot rule out confounding (other differences between groups).

Magnitude: Adjusted OR (odds ratio, a relative measure of association) 0.46 (95% CI 0.27–0.81); disease prevalence 64% in routine fasters versus 76% in non-fasters (Horne 2008).

Lower Perceived Stress

Twelve middle-aged men reported lower perceived stress and a sense of well-being after eight days of water-only fasting (Ogłodek 2021). Without a control group, expectation and setting cannot be separated from the fast itself.

Magnitude: Mean Perceived Stress Scale (PSS-10, a validated 0–40 questionnaire) score fell from 23.83 ± 4.73 before to 12.76 ± 3.45 after 8 days of water-only fasting, no control group (Ogłodek 2021).

Speculative 🟨

Autophagy and Immune-Cell Renewal

Cycles of prolonged fasting regenerated blood-forming stem cells and reversed chemotherapy-induced immune suppression in mice (Cheng 2014); a 72-hour human fast raised autophagy gene expression in white cells (Qian 2021). The basis is mechanistic only.

Lifespan Extension

Periodic fasting extends lifespan in yeast, worms and rodents, as summarized in a review (Longo 2014). No human study has measured lifespan after full fasting; the basis is animal evidence only.

Lower Circulating Amyloid-Beta

About ten days of water-only fasting lowered plasma amyloid-beta 40 and 42, proteins that form Alzheimer’s plaques, in 20 adults (Commissati 2025). Effects on brain amyloid or dementia are untested; no controlled data exist.

Benefit-Modifying Factors

  • Genetic polymorphisms: No study has tested whether gene variants change the benefits of full fasting; variants such as APOE4 (a gene variant affecting cholesterol transport and Alzheimer’s risk) remain unexamined in fasting trials.
  • Baseline blood pressure: Higher starting pressure predicted larger falls; the estimated floor below which no further decrease occurred was about 96/67 mmHg at the end of the fast (Goldhamer 2002).
  • Baseline body weight and glucose: In a meta-analysis, weight fell more with obesity than with normal weight (Ulupınar 2026); a review found glucose and insulin improved with normal glucose regulation but not in diabetes (Ezpeleta 2024).
  • Sex: Molecular responses to water-only fasting differed between men and women with metabolic syndrome (clustered high blood pressure, glucose, lipids and waist size) (Jiang 2023); no study shows a sex difference in clinical benefit.
  • Pre-existing conditions: People with hypertension showed the largest blood pressure falls (Goldhamer 2001); a review found no glycemic improvement in type 1 or type 2 diabetes (Ezpeleta 2024).
  • Age: Adults with a median age of 62 regained fat-free mass and kept fat loss after refeeding in a TrueNorth Health Center study (Gabriel 2025); older adults with low muscle reserve have the least margin for the lean-mass decline during the fast.

Potential Risks & Side Effects

High 🟥 🟥 🟥

No risk reaches High: adverse-event data for multi-day full fasting come from uncontrolled case series and single-arm trials, and the only controlled comparisons tested fasts of 24–48 hours or measured lipid markers rather than clinical events.

Medium 🟥 🟥

Rise in LDL and Total Cholesterol During the Fast

Fasting mobilizes stored fat and cholesterol into the blood. In a randomized crossover trial, a single 24-hour water-only fast raised total cholesterol versus a day of usual eating, with values back to baseline 48 hours later (Horne 2013). A meta-analysis of 32 mostly uncontrolled studies found LDL rising for up to about 10 days of fasting (Çamli 2026), and cholesterol normalized after refeeding in a 7–14-day cohort (Fang 2021). Whether repeated transient rises affect artery disease is unknown.

Magnitude: Pooled LDL increase Hedges’ g 0.489 (a standardized effect size, where about 0.5 is moderate; 95% CI 0.286–0.692), from mostly pre-post studies with no control group (Çamli 2026).

Low 🟥

Fatigue, Insomnia, Nausea, Headache and Near-Fainting

These were among the most frequent adverse events across 768 supervised water-only fasting visits, mostly mild and transient (Finnell 2018). A trial of 24–72-hour fasts around chemotherapy reported the same grade 1–2 fatigue, headache and dizziness (Dorff 2016). Neither study had a non-fasting comparison group.

Magnitude: Share of visits with each event: fatigue 48.2%, insomnia 33.5%, nausea 32.2%, headache 30.1%, presyncope (near-fainting) 28.3%; no control group, so no comparison rate is available (Finnell 2018).

Electrolyte Disturbances and Dehydration

Eight days of water-only fasting produced dehydration, hyponatremia (low blood sodium) and falls in calcium and magnesium (Ogłodek 2021). In one cohort a 70-year-old man developed grade 4 (life-threatening) hyponatremia with speech difficulty on day 9 (Finnell 2018). Starvation programs reported electrolyte breakdown (Runcie 1970).

Magnitude: Serious adverse events requiring hospitalization in 2 of 768 visits (one grade 4 hyponatremia, one dehydration; the paper prints this as 0.002%, whereas 2 of 768 is 0.26%); no control group, so no comparison rate is available (Finnell 2018).

Hyperuricemia (High Blood Uric Acid)

Blood uric acid rises during water-only fasting (Ogłodek 2021), a change usually attributed to ketone bodies competing with urate for kidney excretion. In obese patients on total fasting, hyperuricemia did not trigger acute gout, even in one known gout patient (Runcie 1969).

Magnitude: Hyperuricemia occurred in 42 obese patients treated by total fasting, none of whom developed acute gout; no control group, so no comparison rate is available (Runcie 1969).

Heart-Rhythm Disturbances and Death in Very Long Fasts

During zero-calorie fasts averaging 97 days, the QTc interval (heart-rate-corrected electrical recovery time on an electrocardiogram, ECG) lengthened, and one patient had two cardiac arrests from torsade de pointes (a dangerous ventricular rhythm) (Pringle 1983). Deaths occurred during 1960s therapeutic starvation (Spencer 1968).

Magnitude: QTc exceeded normal by the end of the fast in 7 of 13 patients fasting 97 ± 25 days; no control group, so no comparison rate is available (Pringle 1983).

Starvation Ketoacidosis

Prolonged full fasting can tip fasting ketosis into ketoacidosis (blood acid buildup from ketones) without diabetes. A 25-year-old woman needed intravenous fluids for vomiting and ketoacidosis after a 26-day water-only fast (Lefevre 2020). A review of prolonged-fasting trials also lists metabolic acidosis (excess blood acid) among adverse events (Ezpeleta 2024).

Magnitude: Not quantified in available studies. Only a case report and a review’s list of adverse events describe ketoacidosis after full fasting, so no incidence can be estimated.

Refeeding Syndrome and Wernicke Encephalopathy

Refeeding syndrome (dangerous falls in phosphate, potassium and magnesium when eating resumes) threatens anyone with little intake for over 10 days, as a review describes (Mehanna 2008). Wernicke encephalopathy (brain injury from thiamine deficiency) followed 40-day water-only fasts in two case reports (Hutcheon 2015; Aweer 2025).

Magnitude: Not quantified in available studies. Only case reports and clinical reviews describe these events after full fasting, so no incidence can be estimated.

Lean (Fat-Free) Mass Loss ⚠️ Conflicted

A meta-analysis found fat-free mass falls during prolonged fasts (Ulupınar 2026). However, it largely recovered after refeeding in one cohort (Gabriel 2025), and zero-calorie alternate-day fasting gave more favorable lean-mass change than calorie restriction at 24 weeks (Catenacci 2016). Net: acute loss is real but partly recoverable.

Magnitude: Fat-free mass made up 74% of the weight lost by the end of fasting; after refeeding and maintenance the median fat-free mass change from baseline was −4.4%, no control group (Gabriel 2025).

Speculative 🟨

Inflammatory and Platelet Activation ⚠️ Conflicted

Ten-day water-only fasts raised CRP (C-reactive protein, an inflammation marker) and platelet-activation markers in 20 adults (Commissati 2025). Fasting for 7–14 days reduced platelet activation (Fang 2021). Net: uncontrolled biomarker data only; consequences unknown.

Organ-Level Protein Shifts of Unknown Consequence

A 7-day water-only fast changed over 1,000 plasma proteins, including a brain-specific matrix protein, and genetic analyses predicted both beneficial and adverse effects (Pietzner 2024). The basis is mechanistic only.

Risk-Modifying Factors

  • Genetic polymorphisms: Variants in ABCG2 (a gut and kidney uric acid exporter) and SLC2A9 (a kidney uric acid transporter) raise baseline uric acid and could amplify fasting hyperuricemia; congenital long-QT variants could add to fasting QT prolongation. Both are inferred, not studied.
  • Baseline biomarkers: Low potassium, phosphate or magnesium before refeeding and a BMI (body mass index) under 16 kg/m² are high-risk criteria for refeeding syndrome in NICE (UK National Institute for Health and Care Excellence) guidance (NICE CG32), as summarized in a review (Mehanna 2008).
  • Sex: Adverse-event severity did not differ by sex across 768 supervised fasting visits by 652 patients, 63% of them women (Finnell 2018).
  • Pre-existing conditions: Drug-treated diabetes (Corley 2018), heart-rhythm disorders (Pringle 1983), and depleted nutritional reserves (a review, Mehanna 2008) raise risk; gout and kidney disease plausibly add uric acid and fluid hazards.
  • Age: Adverse-event severity rose weakly with age and fast length, and both serious events in one cohort occurred in men aged 70 and 73 (Finnell 2018).

Key Interactions & Contraindications

Documented interactions:

  • Insulin and sulfonylureas (insulin-releasing drugs: glyburide, glipizide, gliclazide): Hypoglycemia (low blood sugar); very-low-calorie fasting days raised hypoglycemia despite medication reduction, RR 2.05 (95% CI 1.17–3.52) (Corley 2018). Dose reduction before fasting and frequent glucose checks mitigate.
  • Metformin (a first-line glucose-lowering drug): Volume depletion, low blood pressure and kidney impairment; the label says to stop it temporarily while food and fluid intake is restricted (Metformin FDA label). Pausing it with the prescriber during multi-day fasts mitigates.
  • SGLT2 inhibitors (sodium-glucose cotransporter-2 inhibitors, diabetes drugs: empagliflozin, dapagliflozin): Avoid during multi-day fasts. Ketoacidosis (blood acid buildup) with near-normal glucose; the label lists reduced caloric intake as a precipitant and advises withholding 3 days before prolonged-fasting procedures (Jardiance FDA label; Mendelsohn 2020).
  • Acetaminophen (paracetamol), over the counter: Liver injury; recent fasting was associated with hepatotoxicity at 4–10 g per day (Whitcomb 1994), and a 36-hour fast raised exposure 20% in a randomized crossover study (Achterbergh 2019). Keeping doses well below 4 g daily, or avoiding it, mitigates.
  • Alcohol: Hypoglycemia, because ethanol blocks glucose production when glycogen is depleted; the syndrome was reproduced experimentally with pure ethanol (Freinkel 1963). Abstaining during the fast and early refeeding prevents it.
  • Warfarin: Higher warfarin exposure and bleeding risk; a 36-hour fast cut S-warfarin clearance 19% in a randomized crossover pharmacokinetic study (Lammers 2017), and abrupt vitamin K and weight changes add instability. INR (international normalized ratio, a clotting-time measure) checks during and after the fast mitigate.
  • Caffeine, metoprolol, midazolam and omeprazole: Slightly shorter drug action; a 36-hour fast raised caffeine clearance 17–20%, metoprolol 13% and intravenous midazolam 12%, and left omeprazole unchanged, in randomized crossover studies (Lammers 2015; Lammers 2017; Lammers 2018). Watching response to usual doses mitigates.
  • Lithium: Rising lithium levels and toxicity; the label lists volume depletion, dehydration and electrolyte changes, especially in sodium, as raising toxicity risk (Lithobid FDA label). Lithium-level checks during the fast mitigate.
  • NSAIDs (nonsteroidal anti-inflammatory drugs: ibuprofen, naproxen, celecoxib), over the counter and prescription: Reduced kidney blood flow and kidney injury; NSAID labeling names dehydration and hypovolemia (low blood volume) among the states of greatest kidney risk (Celebrex FDA label). Avoiding them during multi-day fasts mitigates.
  • Antihypertensives (blood-pressure-lowering drugs: lisinopril, amlodipine, metoprolol): Additive blood pressure fall and presyncope; the lisinopril label names volume or salt depletion and dehydration as causes of excessive hypotension (low blood pressure) (Zestril FDA label). Supervised tapering with standing blood pressure checks mitigates.
  • Thiazide and loop diuretics (drugs increasing urine output: hydrochlorothiazide, furosemide): Compounded dehydration, hyponatremia and hypokalemia (low blood potassium); the furosemide label warns of water and electrolyte depletion (Lasix FDA label). Electrolyte checks every 2–3 days mitigate.

Theoretical interactions (inferred from mechanism, not studied in humans):

  • Glucose-lowering supplements (berberine, chromium, cinnamon extract): Additive hypoglycemia, inferred from shared glucose-lowering action. Pausing them during the fast mitigates.
  • Blood-pressure-lowering supplements (garlic extract, hawthorn, beetroot nitrate): Additive hypotension and presyncope, inferred from shared blood-pressure-lowering action. Pausing them during the fast mitigates.
  • Fat-soluble vitamins and fish oil (vitamin D, vitamin E, omega-3 capsules): Poor absorption without dietary fat, inferred from absorption physiology. Pausing them until refeeding avoids waste.
  • Sauna and hot baths: Additive dehydration and presyncope, inferred from sweat loss and blood-vessel dilation. Avoiding heat exposure during multi-day fasts mitigates.

Populations who should avoid Full Fasting:

  • People meeting NICE refeeding high-risk criteria (NICE CG32), such as BMI under 16 kg/m², unintentional weight loss over 15% in 3–6 months, or low potassium, phosphate or magnesium (a review, Mehanna 2008).
  • People taking SGLT2 inhibitors who cannot withhold them (Jardiance FDA label).
  • People with a prolonged QTc or prior torsade de pointes, for long fasts (Pringle 1983).
  • Pregnant or breastfeeding women: inferred from the fetus’s and infant’s dependence on continuous maternal nutrient supply; no human data.
  • People with type 1 diabetes: inferred from absent insulin reserve and ketoacidosis risk; no human data on full fasting.
  • People with a current or past eating disorder: inferred from the restrictive pattern full fasting rehearses; no human data.
  • People with advanced kidney or liver disease: inferred from impaired electrolyte, urate and glucose regulation; no human data.
  • Children and adolescents: inferred from growth demands; no human data.

Risk Mitigation Strategies

Durations, intervals and thresholds below follow common practice unless cited.

  • Supervision for fasts beyond 3 days: Mitigates electrolyte, blood pressure and refeeding hazards; the largest safety series comes from residential fasts with daily clinical checks (Finnell 2018).
  • Refeeding for half the fast length: Prevents refeeding syndrome; the supervised protocol refeeds for half as many days as the fast (Finnell 2018), and fasts over 10 days meet NICE high-risk criteria (a review, Mehanna 2008).
  • Thiamine before refeeding long fasts: Prevents Wernicke encephalopathy; NICE-based guidance in a review gives thiamine before and during refeeding of high-risk patients (Mehanna 2008).
  • Electrolyte checks: Detect hyponatremia and hypokalemia; sodium and potassium every 2–3 days during multi-day fasts, ending the fast if sodium falls below 130 mmol/L.
  • Medication plan before starting: Prevents hypoglycemia, ketoacidosis and hypotension: SGLT2 inhibitors withheld 3 days, as labeled before prolonged-fasting procedures (Jardiance FDA label), insulin and sulfonylureas reduced (Corley 2018), antihypertensives tapered (Goldhamer 2001).
  • Slow position changes, no strenuous exercise: Reduces presyncope, which occurred in 28.3% of supervised fasting visits (Finnell 2018).
  • ECG monitoring in very long fasts: Detects QTc prolongation before arrhythmia; regular in-patient ECG monitoring was advised after torsade de pointes in a fasting patient (Pringle 1983).
  • Protein-adequate refeeding and resistance training: Counters lean-mass loss; fat-free mass recovered after refeeding in a supervised cohort (Gabriel 2025).
  • Lipid recheck after refeeding: Separates the transient LDL rise from persistent change; lipids measured 2–4 weeks after refeeding, since fasting values normalize after refeeding (Fang 2021).
  • Hydration of 2–3 liters per day: Limits dehydration and uric acid concentration during multi-day fasts.

Therapeutic Protocol

Timing, titration and repetition intervals without a citation reflect common practice.

  • Short full fast (24–36 hours): A 24-hour water-only fast was tested in a randomized crossover trial (Horne 2013); alternate-day protocols repeat zero-calorie days (Stekovic 2019); repeated 24-hour fasts are used by Jason Fung’s clinic (Furmli 2018).
  • Five-day water-only fast: Studied in normal-weight adults with water, salt and minerals allowed, then refeeding at 30%, 60% and 100% of energy needs on days 6–8 (Jiang 2021).
  • Supervised prolonged water-only fast: TrueNorth Health Center (Alan Goldhamer): short fruit-and-vegetable prefast, water-only fasting (median 7 days, range 2–41), refeeding for half the fast length on a plant-based diet (Finnell 2018; Goldhamer 2001).
  • Modified-fasting alternatives: Buchinger fasting supplies 200–250 kcal per day for 4–21 days at the Buchinger Wilhelmi clinic, per the clinic’s own cohort report (Wilhelmi de Toledo 2019); Longo’s fasting-mimicking diet is a five-day low-calorie plan. Neither is full fasting.
  • Fluids and electrolytes: Some protocols allow water only (Finnell 2018); others add salt and minerals (Jiang 2021); a 10-day study concluded that sodium chloride, not vitamins, should be supplemented (Dai 2022).
  • Repetition interval: No trial has compared intervals; a review found metabolic gains faded 3–4 months after prolonged fasts (Ezpeleta 2024), and practitioners repeat prolonged fasts every few months to yearly.
  • Time of day: Fasts usually start after an evening meal, so the first night is spent asleep, and end with a light midday meal, which eases refeeding.
  • Genetic polymorphisms: No pharmacogenetic data guide fast length; people with familial hypercholesterolemia (inherited very high LDL) or high-urate genotypes may see larger LDL or uric acid rises (inferred).
  • Sex: Women made up 63% of the 652 patients in a 768-visit supervised cohort with no sex difference in adverse-event severity (Finnell 2018); no sex-specific protocols exist.
  • Age: Adverse-event severity rose weakly with age (Finnell 2018); older adults often use shorter fasts with protein-adequate refeeding to protect muscle.
  • Baseline biomarkers: Higher baseline blood pressure predicted larger falls (Goldhamer 2002); low electrolytes or BMI under 16 kg/m² signal refeeding risk (a review, Mehanna 2008).
  • Pre-existing conditions: Drug-treated diabetes or hypertension requires medication changes before fasting (Corley 2018; Zeiler 2024); people with cancer have fasted around treatment only within trials (de Groot 2015).

Discontinuation & Cycling

  • Short-term by design: Full fasting is episodic; single fasts last from one day to several weeks and may be repeated, but no protocol is continuous or lifelong.
  • Withdrawal effects: There is no withdrawal syndrome; the hazard of ending a long fast is refeeding syndrome from shifts in phosphate, potassium, magnesium and fluid, as a review describes (Mehanna 2008).
  • Tapering (refeeding): Ending is graded: 30%, 60%, then 100% of energy needs over three days after a 5-day fast (Jiang 2021), or refeeding for half the fast length (Finnell 2018).
  • Cycling: No data define an optimal cycle; a review found metabolic gains disappeared 3–4 months after prolonged fasts even when weight loss was maintained (Ezpeleta 2024), the usual rationale for periodic repetition.
  • Ending a fast early: Fasts are stopped for symptomatic low blood pressure, confusion, palpitations, persistent vomiting or abnormal electrolytes.

Sourcing and Quality

  • No product involved: Full fasting requires no product, so purity, formulation and third-party testing considerations largely do not apply.
  • Supervised programs: TrueNorth Health Center (water-only) and the Buchinger Wilhelmi clinic (modified fasting) publish safety data from their own programs (Finnell 2018; Wilhelmi de Toledo 2019) while selling those programs, a direct financial interest.
  • Physician-society standards: German fasting-therapy guidelines require physicians trained and certified in fasting therapy (Wilhelmi de Toledo 2013); the issuing medical association’s members run fasting clinics and earn revenue from the fasting they endorse.
  • Electrolyte products: Where salt or minerals are used, plain sodium chloride and third-party-tested magnesium or potassium products avoid sugars, sweeteners and proprietary blends.
  • Home meters: Finger-stick blood ketone and glucose meters with test strips matched to the device give more reliable readings than urine ketone strips, which fall as the kidney adapts.

Practical Considerations

  • Time to effect: The switch from glucose to ketone fuel stabilized between days 3 and 6 (Dai 2022); systemic protein changes appeared only after 3 days (Pietzner 2024); weight and blood pressure fall within days.
  • Common pitfalls: Refeeding too quickly, keeping diabetes or blood pressure drugs unchanged, strenuous exercise, unsupervised long fasts, and expecting lasting metabolic change without a later diet change, which a review found fades (Ezpeleta 2024).
  • Regulatory status: Full fasting carries no product label; the regulatory text bearing on it is drug labeling, such as the instruction to withhold SGLT2 inhibitors before procedures with prolonged fasting (Jardiance FDA label).
  • Cost and access: Short self-directed fasts cost nothing; supervised residential programs for prolonged fasts are offered by few clinics and are usually paid privately.
  • Daily-life demands: Fatigue, insomnia and presyncope in many visits (Finnell 2018) make multi-day fasts hard to combine with demanding work, driving or travel.

Interaction with Foundational Habits

  • Sleep: Disrupting (direct). Insomnia occurred in 33.5% of supervised fasting visits (Finnell 2018), plausibly from hunger signaling and raised alertness hormones; protocols build in daytime rest and avoid late caffeine.
  • Nutrition: Potentiating (indirect). The post-fast diet determines whether gains persist: TrueNorth pairs fasts with a whole-plant-food diet free of added salt, oil and sugar, with blood pressure and fat loss sustained for weeks (Zeiler 2024; Gabriel 2025).
  • Exercise: Blunting for muscle (direct). A meta-analysis found fat-free mass falls during fasts (Ulupınar 2026) and near-fainting limits activity; light walking during the fast and resistance training plus adequate protein after refeeding support lean-mass recovery.
  • Stress management: Mixed (direct). Perceived stress fell after an 8-day water-only fast (Ogłodek 2021), but fasting is itself a physiological stressor that triggers an inflammatory response (Commissati 2025); fasts are usually scheduled away from periods of heavy strain.

Monitoring Protocol & Defining Success

Baseline testing: Before a multi-day fast, baseline tests establish safety and a reference for later change: sodium, potassium, magnesium and phosphate, creatinine, uric acid, glucose, a lipid panel, lying and standing blood pressure, body composition, and an ECG for fasts beyond about a week or for anyone taking QT-prolonging drugs. Medications are reviewed with the prescribing clinician before starting.

Ongoing monitoring: During fasts longer than 3 days, blood pressure and symptoms are checked daily and electrolytes every 2–3 days, with regular ECGs during very long fasts (Pringle 1983). After refeeding, lipids, uric acid and body composition are rechecked at 2–4 weeks and again at 3 months, when a review found metabolic changes fade (Ezpeleta 2024).

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Sodium 135–145 mmol/L (standard reference range) Safety: hyponatremia stops the fast Grade 4 hyponatremia occurred on day 9 of a supervised fast (Finnell 2018); pair with chloride
Potassium 3.5–5.0 mmol/L (standard reference range) Safety: low values signal refeeding risk Low potassium before refeeding is a NICE high-risk criterion (review: Mehanna 2008); pair with magnesium
Phosphate 2.5–4.5 mg/dL (standard reference range) Safety: falls on refeeding Check before and during the first days of refeeding (review: Mehanna 2008)
Magnesium 1.7–2.2 mg/dL (standard reference range) Safety: falls during long fasts Plasma magnesium decreased from the first month of a 382-day fast (Stewart 1973)
Uric acid 3.5–7.2 mg/dL (standard reference range) Safety: expected to rise; matters with gout Rose in total-fasting patients (Runcie 1969); dehydration raises values
Creatinine 0.6–1.3 mg/dL (standard reference range) Safety: kidney function A slight kidney-function decrease occurred during a 10-day fast (Dai 2022)
Glucose 70–99 mg/dL fasting (standard reference range) Expected to change (falls); safety with diabetes drugs Values below this range are expected during a fast; track against the pre-fast baseline and symptoms; values near 30 mg/dL without symptoms occurred in a 382-day fast (Stewart 1973)
Beta-hydroxybutyrate Below 0.4 mmol/L (standard reference range) Expected to change (rises) Values above this range are expected during a fast and mark adherence; finger-stick meter; the fuel switch stabilized on days 3–6 (Dai 2022)
LDL cholesterol No established target during a fast; compare the post-refeeding value with the individual’s own baseline Expected to change (rises) Draw 2–4 weeks after refeeding; rise shown in a meta-analysis (Çamli 2026); morning sample after an overnight fast
Blood pressure Below 120/80 mmHg (“normal” in the American College of Cardiology and American Heart Association guideline) Expected to change (falls); safety for hypotension Source: Whelton 2018; the issuing societies’ cardiologist members treat high blood pressure, so lower thresholds can enlarge their patient base; conventional reference range about 90/60 to 120/80 mmHg; measure lying and standing
QTc on ECG 450 ms or less in men, 460 ms or less in women (standard reference range) Safety: prolongation precedes arrhythmia Progressive QTc lengthening during long zero-calorie fasts (Pringle 1983)
Fat-free mass No established target; track change from the individual’s own baseline Expected to change (falls, then recovers) Measured by dual-energy X-ray absorptiometry (a low-dose body-composition scan) before and after (Gabriel 2025)

Qualitative markers:

  • Energy level and fatigue across fasting days
  • Sleep quality and night waking
  • Hunger intensity, which typically fades after the first days
  • Dizziness or near-fainting on standing
  • Mood and perceived stress
  • Mental clarity and concentration
  • Palpitations or irregular heartbeat

Emerging Research

  • Seven-day water-only fast in prostate cancer: A non-randomized Johns Hopkins study of about 40 men, fasting and ketone-supplement cohorts combined, (NCT06826924) tests safety and feasibility of a 7-day water-only fast with 3-day refeeding; estimated completion June 2027. Positive results extend safety data to older men with cancer; null results would weaken multi-day fasting in illness.
  • Water-only fasting around breast-cancer chemotherapy: The single-arm GAMMER study (NCT06536881) enrolls about 30 women for 24–48-hour fasts per cycle, with feasibility primary and blood toxicity secondary; estimated completion 2029. Positive toxicity data would revive the chemotherapy item; null data would support the meta-analysis that found no protection (Ferro 2023).
  • 36-hour fast and immune profiling: A UC Davis single-arm study (NCT05104502) of a 36-hour water-only fast in up to 4 healthy men has passed its registered September 2026 completion date; it is still listed as recruiting, with no results posted. Either result would describe only short-term immune-cell changes.
  • Exercise plus five-day modified fasting: The randomized EMPOWER-BAT trial (NCT07837752) plans 60 postmenopausal women on about 600 kcal per day, not full fasting, with Buchinger Wilhelmi as collaborator; not yet recruiting, completion 2029. Positive results would support pairing exercise with fasting; null results would favor exercise alone.
  • Completed TrueNorth trials: NCT04515095 (hypertension) and NCT04514146 (body composition) completed in 2021–2022, with no registry results posted; their findings appeared as single-arm reports (Zeiler 2024; Gabriel 2025).
  • Head-to-head with calorie restriction: No randomized trial has compared multi-day water-only fasting with equal-calorie restriction; zero-calorie alternate-day fasting showed no weight advantage (Catenacci 2016). Such a trial would show whether fasting acts beyond the calorie deficit.
  • Persistence of lipid and inflammatory changes: LDL rises in a meta-analysis (Çamli 2026) and CRP and platelet activation (Commissati 2025) need follow-up after refeeding; persistence would weaken the cardiovascular case, while normalization, as reported by Fang 2021, would strengthen it.
  • Protein signatures as outcome predictors: Pietzner 2024 predicted both beneficial and adverse consequences from fasting-induced protein changes; outcome studies could confirm or refute these predictions.

Conclusion

Full fasting, going without all food for a day to several weeks while drinking water, reliably produces rapid weight loss, and repeated short fasts perform about as well as ordinary daily calorie cutting. For a health-focused adult willing to accept a demanding practice, the other promised benefits remain uncertain. Falls in blood pressure, better blood sugar control and symptom relief in some chronic conditions come mostly from studies without comparison groups, and gains on blood tests tend to fade within a few months after eating resumes. Claims about cellular cleanup, immune renewal and longer life rest on animal research.

The costs are concrete. Multi-day fasts bring fatigue, poor sleep, nausea and near-fainting in many people, a loss of muscle as well as fat, a temporary rise in the type of cholesterol tied to artery disease, and occasional dangerous drops in blood salts. Very long fasts carry rare but serious heart-rhythm, nerve-damage and eating-resumption hazards, and the danger rises sharply for people taking diabetes or blood pressure drugs.

The evidence base is thin and tilted. Much of the human data on long water-only fasts comes from a clinic that sells such fasts, some leading laboratory researchers also sell related diet products, and neither drug makers nor insurers have paid for large long-term studies. On current evidence, full fasting is a powerful short-term change in metabolism whose long-term value for healthy aging remains unproven.

Top - Benefits - Risks - Protocol