Fasting

Fasting is a somatic practice of deliberate abstention from food — and in some traditions water — for a defined period. It appears independently in virtually every major religious tradition and forms the basis for a large modern clinical literature on metabolic and cardiovascular health. It alters physiology primarily through interoceptive engagement and a metabolic switch from glucose to ketone oxidation.

The practitioner abstains from caloric intake for a defined window, typically while continuing to drink water and electrolytes. Protocols range from a 12–16 hour daily eating-window restriction (time-restricted eating) to multi-day water fasts. Sessions are self-directed in most lay contexts and physician-supervised at the prolonged end.

A typical fast runs from 12 hours (overnight time-restricted eating) to 21 days under medical supervision. At introductory levels, effects include reduced post-meal energy fluctuations, mild hunger waves, and improved sleep continuity. Established 24–72 hour fasts produce hunger peaks, headache, irritability, and mild ketosis, followed by diminished hunger and elevated mental clarity. Prolonged fasts (3+ days) intensify interoceptive sensitivity and, for some, produce vivid dreaming, perceptual shifts, and visionary experiences.

The primary risks are refeeding syndrome (potentially fatal electrolyte collapse on reintroduction of food after prolonged fasts),^[1] hypoglycemia in those on insulin or sulfonylureas, and elevated risk of eating disorder onset, particularly among adolescents.^[2] Eating disorder history, type 1 diabetes, pregnancy, lactation, and underweight status are absolute contraindications. The IF literature includes hundreds of RCTs and a 2024 umbrella review across cardiometabolic outcomes;^[3] evidence quality is rated moderate, with substantial protocol-dependent heterogeneity.

Origins

Fasting appears independently across virtually every major religious tradition, with documented practice extending to at least the Vedic period (~1500 BCE). Its recurrence across geographically isolated cultures suggests convergent development rather than singular diffusion.

The Torah prescribes a complete fast (no food or water) on Yom Kippur, framed as teshuvah (repentance).^[4] The Quran prescribes sawm throughout Ramadan, the ninth lunar month, as one of the Five Pillars of Islam; the fast runs from pre-dawn fajr to sunset maghrib.^[5] The Gospel of Matthew records Jesus fasting forty days in the desert,^[6] a pattern echoed by accounts of Moses and Elijah; early Christian practice (Didache, 2nd century) prescribed weekly fasts, and Eastern Orthodox traditions maintain some of the most extensive contemporary fasting calendars in any religion.

Vedic texts describe upavasa ("dwelling near [God]") as purification. Jain practice includes graded fasts ranging from ekasana (one meal) to santhara (terminal fast as spiritual end-of-life practice). The Theravada Buddhist Vinaya prohibits monks from eating after noon, though the historical Buddha is recorded as rejecting extreme food deprivation in favor of the Middle Way.

Fasting is also documented as preparation for vision quests, healing ceremonies, and initiation rites across Indigenous North American traditions (Lakota hambléčeyapi), Amazonian traditions, and African shamanic lineages. These remain embedded in living ceremonial contexts and are categorically distinct from metabolic or therapeutic fasting.

Transmission & secularization

European natural-hygiene movements popularized supervised therapeutic fasting from the mid-1800s. Otto Buchinger, a German physician, fasted for 19 days in 1919 after developing severe rheumatoid arthritis; he attributed his remission to the fast and opened the first Buchinger fasting clinic in Wiesbaden in 1920. The Buchinger Wilhelmi clinics continue under his family's direction, employing physician-supervised 4–21-day fasting periods with ~250 kcal/day of vegetable broth and juices.^[7]

The contemporary intermittent-fasting movement emerged in the early 2000s through popular science and online communities. Martin Berkhan developed the 16:8 Leangains protocol — a 16-hour daily fast with an 8-hour eating window — around 2006–2007 in fitness contexts. Michael Mosley popularized the 5:2 protocol (five normal-eating days plus two 500-kcal restriction days per week) following a 2012 BBC Horizon documentary. Valter Longo of the USC Longevity Institute developed the Fasting Mimicking Diet (FMD), a five-day low-calorie protocol designed to trigger fasting-state physiology while permitting limited intake.^[8]

Contemporary context

Over 300 trials of fasting-related interventions are registered on ClinicalTrials.gov, spanning cardiovascular disease, type 2 diabetes, cancer adjunct therapy, neurodegeneration, and longevity biomarkers. Commercial commodification has proceeded in parallel: fasting apps (Zero, Fastic, BodyFast), coaching services, and supplement products have extracted the practice from both its religious community contexts and its clinical oversight frameworks. Eating-disorder specialists have noted that de-contextualizing fasting from communal and clinical supports may increase individual risk by removing protective structures.^[9]

Technique

The method is deliberate abstention from caloric intake — and in the strictest traditions, from water — for a defined window. What varies across the practice's range is duration, what passes the lips during the fast, and what social and medical context surrounds it.

The physiological cascade is predictable. Hepatic glycogen depletes within 12–24 hours; the liver shifts to gluconeogenesis and the body to ketone-body production; ketone (beta-hydroxybutyrate, acetoacetate) levels rise substantially by 24–72 hours.^[11] The most common departure from the method is unstructured "snacking on calorie-free flavored beverages" or chewing gum, which can interrupt the metabolic switch the practice depends on. The second is failing to plan the refeeding window — exiting a prolonged fast on dense carbohydrates rather than dilute liquids and slowly graded foods.

At the introductory tier, the eating window is compressed without total food restriction; this is metabolically modest and forms the basis of most clinical IF trials. At the established tier, the practitioner undertakes a complete 24–48 hour fast, typically with electrolyte supplementation; the metabolic switch to ketone oxidation is reliably engaged by hour 24–36. At the intensive tier — prolonged fasts of three or more days — medical supervision is the universal clinical standard. Full ketoadaptation, with fat providing more than half of energy via ketones, typically occurs between days three and five in individuals without prior ketogenic experience.^[12]

Setting & Preparation

For introductory and established tiers, no specific setting is required — fasting is integrated into ordinary daily activity. Prolonged fasts benefit from reduced physical and cognitive demands, a quiet environment, and ready access to electrolyte supplementation. The Buchinger clinical model adds residential supervision, daily physician monitoring, light physical activity (walking), and a structured refeeding protocol.^[7]

Equipment is minimal: a reliable timer or fasting app for window tracking; electrolyte supplements (sodium, potassium, magnesium) for fasts beyond ~24 hours; for prolonged fasts, access to clinical-grade blood pressure, glucose, and electrolyte monitoring. Religious fasts are typically conducted in the context of communal observance and ritual structure rather than with measurement tools.

Fasting environment requirements scale with duration. Introductory time-restricted eating is integrated into ordinary daily life with no setting requirement beyond a reliable timer or app. Established 24–48 hour fasts benefit from a quiet day with reduced physical demands, ready electrolytes (sodium, potassium, magnesium), and a planned refeeding meal of dilute, easily digested foods. Prolonged fasts (3+ days) require: a low-stimulation environment, reduced physical and cognitive demands, continuous access to water and electrolytes, a pre-arranged check-in person, and at the clinical standard, physician monitoring with baseline and periodic bloodwork (electrolytes, glucose, renal function). The Buchinger framework adds residential supervision and a structured 3–4-day stepped refeeding.^[7]

Preparation matters most for prolonged fasts. Reducing carbohydrate intake in the days before a multi-day fast can ease the metabolic switch and reduce headache severity in the first 24–48 hours. A planned refeeding protocol — what you will eat, in what order, over what hours — is part of the fast, not an afterthought; refeeding accounts for the majority of serious adverse events in the prolonged-fast literature.^[1]

The metabolic switch

The defining mechanism of fasting is the metabolic switch: the transition from glucose to ketone oxidation as primary fuel.^[11] Hepatic glycogen depletes within roughly 12–24 hours of zero caloric intake. The liver shifts to gluconeogenesis and to producing ketone bodies — beta-hydroxybutyrate (BHB) and acetoacetate — from fatty acids. BHB rises substantially by hours 24–72 and becomes the dominant energy source for the brain by days three to five in individuals without prior ketogenic experience.^[12]

Ketones are not merely an alternative fuel. BHB inhibits class I and II histone deacetylases (HDACs), upregulating antioxidant gene expression; it activates HCAR2 on microglia, reducing neuroinflammation; and it serves as a more oxygen-efficient substrate than glucose for neuronal ATP production.^[12] The transition itself — not just the ketotic state — appears to drive much of fasting's claimed neuroplasticity benefit.

Neuroplasticity, BDNF, and autophagy

Animal and limited human data converge on three mechanisms. First, the metabolic switch upregulates brain-derived neurotrophic factor (BDNF) expression in hippocampal neurons, promotes synaptic plasticity, and enhances adult neurogenesis in the dentate gyrus.^[12] These effects are well-established in rodent models; controlled human neuroimaging remains limited. Second, nutrient deprivation suppresses mTOR and activates AMPK, triggering cellular autophagy; in neurons, autophagy clears tau aggregates and amyloid-β precursors. Human autophagy during fasting is measured via proxy biomarkers (LC3-II, beclin-1, p62) rather than direct imaging.^[11] Third, a preclinical fMRI study in mice found that 12-hour fasting globally increased functional connectivity, most significantly between hippocampus, retrosplenial cortex, and sensory cortices, consolidating the brain from five functional modules to three — a network-integration signature under metabolic stress.^[16]

A quantitative EEG study found that fasting altered theta and alpha band activity during n-back working memory tasks, consistent with changed cognitive-load processing, though performance effects were task- and context-dependent.^[17]

Autonomic and neuroendocrine response

Long-term supervised fasting produces a biphasic autonomic response: early sympathetic activation in the first days transitioning to sustained parasympathetic dominance by later days.^[18] A 12-week IF regimen has been documented to raise heart-rate variability indices (RMSSD from 27.16 ± 10.5 to 32.92 ± 17.65 ms), indicating increased parasympathetic tone.^[18] Insulin-like growth factor 1 (IGF-1) is consistently reduced during extended fasting, with simultaneous IGFBP-1 increases.^[8] Ghrelin rises with fasting, but subjective hunger paradoxically diminishes after days three to four in prolonged fasts — the Buchinger cohort reported 93.2% of subjects experiencing no subjective hunger at this phase.^[13]

Fasting's experiential character is biphasic and depends primarily on duration. The first 24–48 hours are dominated by metabolic adaptation: hunger waves, headache, irritability, and energy dips as glycogen depletes and ketone production begins. From day three onward, many practitioners report a phase of clarity and reduced hunger that contemplative traditions across Christianity, Sufism, and Buddhism describe as the threshold beyond purgation. The Buchinger observational cohort (n=1,422) documented that 93.2% of subjects reported no subjective hunger by the established-ketosis phase.^[13]

Effects at the introductory tier (12–16 hour TRE) are modest — reduced post-meal energy fluctuations, mild morning hunger, and for some, improved sleep continuity. Established 24–48 hour fasts add the full hunger arc and the onset of ketosis. Prolonged fasts (3+ days) intensify interoceptive sensitivity, can produce vivid dreaming and perceptual shifts, and in some practitioners across traditions are associated with mystical or visionary experiences. A 2025 review documented that, in vulnerable individuals, prolonged fasting can also precipitate dissociative states and — rarely — psychotic decompensation.^[19]

Adverse Effects

The eating-disorder evidence is the most consistent psychiatric signal. A five-year prospective study of 496 adolescent girls identified fasting as a significant predictor of onset of recurrent binge eating, subthreshold and threshold bulimia nervosa, and purging disorder.^[2] A 2022 cross-sectional study of Canadian adolescents and young adults engaging in IF found significantly elevated rates of objective binge episodes, self-induced vomiting, and laxative use relative to community norms.^[9]

Contraindications

Absolute contraindications

• Current or past eating disorder (anorexia nervosa, bulimia nervosa, binge-eating disorder, orthorexia)^[2]
• Type 1 diabetes and insulin-dependent type 2 diabetes (without close physician supervision and medication adjustment)
• Pregnancy and lactation
• Underweight status (BMI < 18.5)
• Pediatric and adolescent age (developing metabolism; elevated eating-disorder risk)^[9]
• Active malignancy (except under oncologist supervision in protocols designed for cancer-adjunct use)

Relative contraindications and elevated-risk populations

• History of psychosis or severe psychiatric illness, particularly for prolonged fasts (>72 hours)^[19]
• Current use of insulin, sulfonylureas, meglitinides, lithium, or hepatically-metabolized antiepileptics (requires medical review before any structured fasting regimen)
• Major cardiac arrhythmias
• Recent surgery or acute illness recovery
• Pre-existing electrolyte abnormalities or chronic kidney disease

Prolonged fasts (3+ days) are conducted under medical supervision in every established clinical framework; this is the universal standard, not an enhancement. Unsupervised prolonged fasting carries the highest preventable-risk profile in the practice's range.

Substance Interactions

The most clinically documented interaction is with hypoglycemic medications. Insulin, sulfonylureas, and meglitinides combined with fasting create compounding hypoglycemia risk; dose adjustment by a prescribing physician is required before any structured fasting regimen, and is the standard clinical guidance across IF protocols and diabetes management.

Oral NSAIDs and aspirin taken on an empty stomach substantially increase gastric irritation and ulceration risk; this is established GI pharmacology, not a fasting-specific novel finding. Lithium levels are sensitive to hydration and sodium balance, and fasting-induced dehydration with electrolyte shifts can drive serum lithium into the toxic range; individuals on lithium should not undertake extended fasting without close monitoring. The ketogenic state produced by extended fasting may alter the pharmacokinetics of hepatically-metabolized antiepileptics; effects are agent-specific and not uniformly characterized.

Alcohol consumed in a fasted state is absorbed more rapidly, potentiates hypoglycemia (alcohol inhibits gluconeogenesis), and increases hepatotoxicity risk. This interaction is well-established in clinical pharmacology.

No controlled studies have characterized the combination of fasting with classic psychedelics (psilocybin, LSD, DMT, mescaline). The evidence base is entirely anecdotal and community-reported. Fasting before ayahuasca ceremonies is part of the traditional Amazonian dieta; community reports across psychedelic contexts describe fasting as intensifying and accelerating onset, plausibly through faster gastric absorption, ketosis-related alterations in serotonergic and glutamatergic tone, and blood-glucose instability amplifying physiological stress responses. No peer-reviewed safety data exist; the combination is uncharacterized.

Habituation

Fasting itself does not produce pharmacological dependence. The compulsive-use risk profile is psychological and behavioral, and it concentrates at the intersection of fasting with disordered eating. The Stice five-year prospective study established fasting as a significant predictor of binge-eating and bulimic pathology onset in adolescent girls,^[2] and a 2024 cross-sectional study found that orthorexia fully mediated the association between fasting behavior and eating-disorder symptoms.^[3]

Practitioner-community signals of escalation include: extending fast windows beyond planned duration; concealing fasting from family or clinicians; tying mood or self-worth to weight outcomes during fasts; using fasting to compensate for eating episodes; and a steady drift toward longer or more frequent fasts in pursuit of intensifying clarity or weight effects. These are warning signs of eating-disorder transition, not features of sustainable fasting practice.

If Distress Arises

Mid-fast adverse events stratify by severity. Mild and expected: hunger waves (typically peaking around hour 24–36 and diminishing thereafter in prolonged fasts),^[13] headache (often related to caffeine withdrawal or sodium depletion), fatigue, irritability, light-headedness on standing (orthostatic hypotension from sodium loss). The primary response is electrolyte supplementation — a glass of water with salt, potassium, and magnesium — and rest. Persistent symptoms warrant breaking the fast.

Serious mid-fast symptoms requiring immediate cessation and medical attention include: chest pain or palpitations, syncope, severe confusion or disorientation, new neurological symptoms (numbness, weakness, vision changes), seizure, or new psychiatric symptoms (paranoid ideation, hallucinations, dissociation). Resume eating with a small amount of easily digested food and seek evaluation.

Critically: ending a fast is not a failure. The practitioner has agency throughout and can stop at any time. Forcing through serious symptoms in pursuit of duration is the failure mode that produces preventable harm in this practice.^[19]

There is no universally adopted formal stage model for fasting practice comparable to jhana maps for meditation. What the clinical, biochemical, and contemplative literatures converge on is a recognizable arc of metabolic and phenomenological phases that unfold across hours and days of sustained abstention.^[11]

Within the first 12 hours (post-absorptive phase), the liver supplies glucose from glycogen with gluconeogenesis supplementing as stores fall; ketone production is minimal. Between 12 and 24 hours, hepatic glycogen approaches depletion and ketone production begins — the onset of the metabolic switch. Hunger is typically elevated in this window. From 24 to 72 hours, BHB rises substantially; ghrelin may peak around hour 24–36 with subjective hunger paradoxically diminishing thereafter in many individuals. Between days three and five, full ketoadaptation establishes: fat oxidation becomes the dominant energy source, the Buchinger observational data show 93.2% of subjects reporting no subjective hunger at this phase,^[13] and autophagy proxy biomarkers are measurably elevated.^[12] Beyond day seven, autophagy markers remain elevated, but muscle protein catabolism risk increases and electrolyte management becomes clinically critical — this is the zone requiring medical supervision in every established framework.

The phenomenological arc tracks the metabolic one. Days 1–3 are the purgation phase across Christian, Sufi, and Buddhist literatures — discomfort before clarity, with headache, fatigue, and irritability common and the highest psychological dropout rate in unstructured fasting. Days 3–7 are repeatedly described, across both contemplative and clinical settings, as a clarity phase: heightened mental clarity, elevated mood, reduced hunger. Beyond day seven, practitioners across traditions and clinical case reports describe heightened interoceptive sensitivity, vivid dreaming, perceptual shifts, and occasionally mystical or visionary experiences; these are not uniform across individuals, and adverse psychological states are documented in vulnerable individuals at this stage.^[19]

The intermittent-fasting evidence base is large relative to most practices in this domain, comprising hundreds of randomized controlled trials with substantial heterogeneity across protocols (time-restricted eating, alternate-day fasting, 5:2, fasting-mimicking diet, prolonged water fasting), populations, and outcomes. A 2024 umbrella review searching PubMed, Embase, Web of Science, and Cochrane through January 2024 synthesized systematic reviews and meta-analyses of RCTs on IF and human health outcomes.^[3] A PRISMA-guided meta-analysis of 15 RCTs (n=758) in overweight/obese adults documented significant reductions in weight (–3.59 kg), BMI (–1.39 kg/m²), LDL (–56.22 mg/dL), and systolic blood pressure (–5.54 mmHg), with minimal HDL change and non-significant glucose changes.

The most-cited counterweight is the TREAT trial: a pre-registered RCT (n=116) of 8-hour TRE vs. unrestricted eating for 12 weeks at UCSF, which found a non-significant weight reduction (–1.17% vs. –0.75%) and no significant cardiometabolic benefits on secondary outcomes.^[10] This is the most widely cited null result in the TRE literature and an important moderating datapoint against uncritical claims for 16:8 protocols.

For prolonged fasting, the Buchinger therapeutic fasting observational cohort (n=1,422 across 4–21-day supervised fasts) documented mild and transient adverse events and improvement in 84.4% of subjects with pre-existing health complaints.^[13] A 5-day water-only fasting clinical trial in normal-weight individuals showed significant cardiometabolic improvements without serious adverse events,^[14] and a study of prolonged water-only fasting (median ~17 days) and refeeding documented sustained reductions in body weight, abdominal circumference, blood pressure, LDL, total cholesterol, and hsCRP at 6-week follow-up.^[15] The FMD RCT (n=100) found that three monthly 5-day cycles reduced IGF-1, fasting glucose, triglycerides, cholesterol, C-reactive protein, and blood pressure.^[8]

Limitations

The IF evidence picture is rated moderate quality overall by GRADE-informed reviews, limited by protocol heterogeneity, variable follow-up periods, and the impossibility of participant blinding.^[3]

Several specific gaps shape what can and cannot be claimed. Most TRE and IF trials run 8–24 weeks; long-term cardiometabolic and mortality outcomes are inferred from short-term biomarker changes rather than measured directly. The most rigorous TRE trial (TREAT, Lowe et al. 2020) found no benefit over control, raising the question of whether positive findings in less-rigorous trials reflect caloric reduction (a confounder ubiquitous in unblinded eating trials) rather than the metabolic switch itself.^[10] The FMD evidence base remains substantially developer-led, with independent replication ongoing.^[8] The largest prolonged-fasting safety dataset is observational, not controlled;^[13] a 2024 narrative review of prolonged water fasting explicitly identifies safety as under-characterized in rigorous controlled settings.^[21]

Human neuroimaging of fasting is sparse; the strongest neuroplasticity claims (BDNF upregulation, adult neurogenesis, autophagy induction) rest on rodent models and human proxy biomarkers.^[12] Cognitive and neurological benefits claimed for IF in popular literature remain preliminary at the level of controlled human evidence. Adverse-effect literature has historically under-represented eating-disorder risk; the Stice 2008 and Linardon 2022 datasets are foundational but not yet integrated into most IF clinical guidelines.^[2][9] The psychiatric-consequences literature for prolonged fasting is dominated by case reports and a single 2025 review;^[19] controlled prospective data on psychiatric adverse events at the intensive tier do not yet exist.