Evidence summaries on fasting and metabolic health

fasting.com

Cited guides to fasting basics, intermittent fasting, water-only fasting, and the fasting-mimicking diet.

How Long Until Autophagy From Fasting?

There is no agreed-upon number of fasting hours at which autophagy reliably begins in humans. Many of the fastest timing effects in the research come from studies in rodents, not people. Most human research relies on indirect markers, proteins such as LC3 and p62 and the activity of autophagy-related genes, measured in blood or muscle; the few studies that measured the rate of autophagy used blood cells. Results vary by study, by tissue, and by which marker is measured.

By the fasting.com team

Published

Last reviewed . How pages are reviewed

Medical caution

This is general information about fasting, not medical advice. Fasting can affect blood sugar, blood pressure, medication timing, and nutritional status. Talk to a clinician before starting or changing a fasting practice if you are pregnant, under 18, living with diabetes or another chronic condition, taking prescription medication, or have a history of an eating disorder.

The Short Answer: Nobody Knows an Exact Hour in Humans

No human study reviewed for this page establishes a specific number of fasting hours at which autophagy reliably begins. Many hour-based figures stated elsewhere online rest on studies in rodents, not people; see "Why Many Specific Hour Claims Come From Rodent Studies" below. What exists instead is a set of human studies measuring different markers, in different tissues, at different fasting lengths, with mixed results. For background on fasting itself and the different methods this covers, see what is fasting.

What Autophagy Is

Autophagy (from the Greek for "self-eating") is a basic cellular process: a cell breaks down and recycles its own damaged proteins and worn-out components. It runs at some background level all the time and can be upregulated under stress conditions, including food deprivation (Bagherniya et al., 2018) [14]. This page describes autophagy as a cellular process, not as a measured health outcome; no citation reviewed here shows that triggering more autophagy through fasting produces a specific, measured benefit in humans.

The 2016 Nobel Prize Was for Autophagy Mechanisms Found in Yeast, Not for Fasting

The 2016 Nobel Prize in Physiology or Medicine went to Yoshinori Ohsumi for discovering the mechanisms of autophagy, using baker's yeast as the experimental model. The prize recognized how autophagy works at a molecular level, not any finding about how long a person needs to fast to trigger it. This gets misread often enough that it is worth stating plainly: the prize citation reads "for his discoveries of mechanisms for autophagy," for work done in yeast; it says nothing about fasting duration in humans.

What Human Studies Actually Measure

Measuring autophagy inside a living person is difficult (Lauc et al., 2026) [12], so most human studies rely on indirect markers instead: proteins like LC3 and p62, and the activity of autophagy-related genes, measured in blood cells or a small muscle biopsy. A rise or fall in one of these markers is a signal consistent with more or less autophagy activity, not a direct, real-time readout of it. A few studies have measured autophagic flux, the rate at which autophagy is running, in blood cells drawn from participants (Pietrocola et al., 2017 [6]; Bensalem et al., 2025 [8]). Results differ by which marker is checked, which tissue is sampled, and when the sample is taken relative to a person's last meal, which is exactly what the studies below show.

What the Human Data Show

Blood Markers After a Time-Restricted Eating Window

In a 4-day randomized crossover study of 11 overweight adults, eating on a 6-hour early time-restricted schedule (8am to 2pm) was compared with a 12-hour control schedule (8am to 8pm). The morning blood draw, taken before breakfast and following the longer overnight fast in the early-eating arm, showed increased expression of the autophagy-related gene LC3A and the aging-related gene SIRT1 (Jamshed et al., 2019) [1]. In the evening, the same schedule increased expression of MTOR, the gene for a nutrient-sensing protein that regulates cell growth (Jamshed et al., 2019) [1]; the mTOR protein is generally known to hold autophagy back (general cell biology, not a finding of this study), so the two gene results point in different directions. The authors conclude that early time-restricted eating "may also increase autophagy" in humans (Jamshed et al., 2019) [1]. This was a small study (11 participants) measuring gene expression in whole blood, not autophagy activity.

Muscle Markers After an Overnight Fast

In 12 overweight, healthy young male volunteers, muscle sampled after an overnight fast had more LC3B-II, a marker of autophagosome content, in slow-twitch (type I) fibers than in fast-twitch (type II) fibers (Morales-Scholz et al., 2022) [2]. After a mixed meal, LC3B-II fell in whole muscle and in both fiber types within 90 minutes (Morales-Scholz et al., 2022) [2]. Other autophagy-related proteins (p62/SQSTM1, ULK1, ATG5, ATG12) showed no difference between fiber types. The study had no measurement before the fast and tested no other fasting durations, so it shows that eating quickly lowers this marker, not when fasting raises it.

Muscle Markers in the First Hours of a Fast

In 10 healthy males, muscle biopsies taken before, 4 hours into, and at the end of two supervised 8-hour fasts (one with 2 hours of added arm exercise) showed no significant increase in autophagy and mitophagy marker proteins (LC3BI, LC3BII, BNIP3); if anything, these were non-significantly reduced, and the LC3II:I ratio was unchanged (Islam et al., 2021) [3]. The study's main question was about pathways that build new mitochondria, not autophagy; its authors note, citing earlier work, that these pathways switch on quickly with fasting in rodent muscle but were not seen to do so in earlier human studies (Islam et al., 2021) [3]. Its autophagy result is the 4-hour and 8-hour finding above: no significant increase.

Muscle Markers Across a 36-Hour Fast

In untrained and endurance-trained adults who fasted for 36 hours, with thigh-muscle biopsies taken 2, 12, 24 and 36 hours after a standardized meal, fasting reduced the autophagy-related proteins LC3I, LC3II and p62 in the untrained participants only, and the two groups differed in several autophagy-signalling markers during the fast (Dethlefsen et al., 2018) [4]. The authors' own summary is that muscle autophagy was "only modestly affected" by 36 hours of fasting (Dethlefsen et al., 2018) [4].

Muscle Markers After 72 Hours

In 8 healthy men studied once in the postabsorptive state and once after 72 hours of fasting, the muscle autophagy marker LC3B-II rose by about 30%, but p62, a protein normally broken down during autophagy, also rose by about 10%; the authors state that this makes interpretation of autophagy activity "problematic" (Vendelbo et al., 2014) [5].

Blood Cells During Fasts of Several Days

In human volunteers who fasted for up to 4 days, an increase in autophagic flux (the rate at which autophagy runs) in white blood cells could be detected only when the cells were cultured with and without a drug that blocks the final breakdown step, and among the white-cell types studied only neutrophils showed signs of activated autophagy (Pietrocola et al., 2017) [6]. After a 72-hour water-only fast, gene and protein profiling of white blood cells found higher levels of autophagy-related genes and proteins, which the authors interpret as enhanced autophagy; this is expression data, not a flux measurement (Qian et al., 2021) [7].

Autophagic Flux Over 6 Months of Intermittent Fasting

In a randomised trial of 121 adults with obesity assigned to standard care, calorie restriction, or intermittent fasting combined with time-restricted eating for 6 months, researchers measured autophagic flux in blood immune cells. No between-group difference was significant at 2 months. At 6 months the change from baseline differed significantly between the intermittent-fasting group and standard care (P = 0.04, a post hoc analysis), an effect the authors say may be driven partly by flux tending to fall in the standard-care group; flux did not rise significantly from baseline within the intermittent-fasting group, and calorie restriction did not differ from standard care (Bensalem et al., 2025) [8]. The authors conclude that autophagy "may be" increased by intermittent nutrient restriction in humans and that further studies are required.

Intermittent Fasting and Muscle Markers Over 8 Weeks

In 50 women with an average BMI of about 32 who followed 8 weeks of intermittent fasting (24-hour fasts on 3 non-consecutive days a week, eating either 70% or 100% of energy needs), thigh-muscle samples taken after 12-hour and 24-hour fasts did not show higher autophagy markers overall: a 24-hour fast raised p62 (SQSTM1), and several autophagy-related gene levels fell in the 70% group, which the authors attribute likely to weight loss (Chaudhary et al., 2022) [9]. In the same paper, autophagy markers rose in mouse liver but not mouse muscle.

Ramadan Fasting and Blood Markers

Ramadan fasting runs from dawn to dusk and excludes water during daylight, so it differs from the other fasts on this page. In 50 healthy adults (24 fasting, 26 not fasting), the autophagy-initiation gene Beclin-1 was higher in the fasting group at the end of Ramadan, while LC3β and p62 levels in blood immune cells were lower (Dastghaib et al., 2025) [10]. In 51 adults with overweight or obesity, measured before and at the end of four weeks of Ramadan fasting with no non-fasting comparison group, whole-blood expression of LAMP2, LC3B and ATG5 rose about 4.2-, 1.9- and 1.4-fold, with no significant change in ATG4D (Bou Malhab et al., 2025) [11]. Neither study measured autophagic flux.

A 72-Hour Fast Where Autophagy Wasn't Directly Measured

A 2026 pilot study put 5 healthy adults through a 72-hour water-only fast and tracked a broad panel of metabolic, hormonal, and glycomic markers, finding a biphasic pattern that partly reversed after 11 days of refeeding (Lauc et al., 2026) [12]. The authors describe the overall pattern as "consistent with an environment conducive to autophagy," while stating plainly that autophagy itself was not measured directly, and call for future studies to add direct autophagy markers (Lauc et al., 2026) [12]. Even a recent, purpose-built prolonged-fasting human study, with 5 participants labeled a pilot by its own authors, did not directly measure autophagy: it inferred a permissive environment for it, nothing more.

Why Many Specific Hour Claims Come From Rodent Studies

Some of the "fasting for X hours triggers autophagy" framing seen online traces back to animal research, not human studies. In one frequently cited example, mice subjected to short-term food restriction showed a dramatic increase in neuronal autophagy, confirmed in Purkinje cells, measured with a method the researchers validated against known liver autophagy induction (Alirezaei et al., 2010) [13]. This is mouse tissue, not a human timing study, and should never be read as a human finding or a specific hour count. A broader 2018 literature review concludes that fasting and calorie restriction both have a role in upregulating autophagy, describing the evidence as "overwhelmingly" supportive of autophagy induction across a wide variety of tissues and organs in response to food deprivation (Bagherniya et al., 2018) [14]; the review's own abstract gives no specific human hour threshold and does not break out what share of its underlying evidence is human versus animal, so it supports the general cross-species framing, not any specific number. A related 2020 review focused on mitophagy (autophagy that targets mitochondria) reports that most studies it found showed fasting or calorie restriction inducing mitophagy markers, and describes the evidence as "overwhelmingly" pointing that way, while also calling the evidence "still limited," noting heterogeneous methods, and noting some studies found no fasted-versus-fed difference or found mitophagy in the fed state in muscle after exercise (Mehrabani et al., 2020) [15]. For more on how fasting research generally should be read, including why animal findings don't automatically generalize to humans, see fasting research facts.

Does Fasting Type Matter?

Whether the method of fasting, a water-only fast like the one covered under prolonged fasting, a daily time-restricted eating window, Ramadan-style dawn-to-dusk fasting, or an every-other-day pattern like alternate-day fasting, changes how or when autophagy markers show up in humans is an open question. None of the citations on this page directly compare fasting types against each other for autophagy timing; each study above tested one protocol in isolation. Treat any claim that one fasting method is specifically "better for autophagy" than another as unsupported by the current human evidence.

What This Doesn't Mean

Nothing on this page claims that fasting for a certain number of hours, or fasting longer, cures or prevents any disease, slows aging, or produces a specific measured health outcome in humans. Autophagy is a real, well-established cellular process, and it is robustly induced by fasting in animal models (Bagherniya et al., 2018) [14], but no citation reviewed here connects a specific human fasting duration to a specific measured autophagy level and a specific health outcome. Fasting-mimicking diet trials have also measured autophagy-related markers, as part of structured, low-calorie diet protocols rather than ordinary fasting; those results are specific to those protocols, not a general human autophagy-timing finding, so they are covered on the fasting-mimicking diet research page rather than repeated here.

Safety and Cautions

This is general information, not medical advice. The evidence on autophagy timing in humans is early, indirect in most studies, and inconsistent across studies, whatever fast length is being considered. Talk to a clinician before starting or extending any fast if you are pregnant, under 18, have a history of an eating disorder, or manage diabetes or blood pressure with medication.

What the evidence actually supports

No human study reviewed for this page establishes a specific number of fasting hours at which autophagy reliably begins. The human data that exist are mostly indirect (gene expression or protein markers rather than a measurement of autophagy activity, with a few exceptions in blood cells), and they vary by tissue, by marker, and even by fiber type within muscle. One small study found increased expression of the autophagy-related gene LC3A, alongside increased expression in the evening of MTOR, the gene for a nutrient-sensing protein that generally holds autophagy back (general cell biology, not a finding of this study), in the same 4-day time-restricted-eating protocol; the study's own authors nonetheless conclude that early time-restricted eating "may also increase autophagy" in humans (Jamshed et al., 2019) [1]. Another found an autophagosome marker in muscle higher after an overnight fast than after a meal, in both fiber types, with no earlier or later fasting timepoints tested (Morales-Scholz et al., 2022) [2]. A separate study found no significant increase in muscle autophagy or mitophagy markers across two supervised 8-hour fasts in 10 healthy males, one combined with added exercise (Islam et al., 2021) [3]; the study's main focus was mitochondrial-building pathways, not autophagy.

Several other human studies extend this picture, and do not agree with each other either. A 36-hour fast reduced several autophagy marker proteins in muscle, but only in untrained participants; the researchers describe the overall effect as "only modestly affected" (Dethlefsen et al., 2018) [4]. In 8 healthy men, a 72-hour fast raised one autophagy marker (LC3B-II) in muscle but also raised p62, a protein normally cleared during autophagy, which the study's authors say makes the result "problematic" to interpret (Vendelbo et al., 2014) [5]. In white blood cells from people fasting up to 4 days, a detectable rise in autophagic flux required culturing cells with and without a drug that blocks the final breakdown step, and only one white-cell type (neutrophils) showed clear activation (Pietrocola et al., 2017) [6]; after a 72-hour water-only fast, a separate study found higher levels of autophagy-related genes and proteins in white blood cells, an expression finding rather than a flux measurement (Qian et al., 2021) [7]. A randomized trial that tracked autophagic flux over 6 months found no significant within-group change over that period in an intermittent-fasting group, though it differed significantly from a standard-care comparison group at the 6-month mark in a post hoc analysis (Bensalem et al., 2025) [8]. An 8-week trial of repeated 24-hour fasts did not raise autophagy markers in muscle overall: one marker (p62) rose after a 24-hour fast, but several autophagy-related genes fell, which the authors attribute, as likely, to weight loss (Chaudhary et al., 2022) [9]. Studies of Ramadan fasting (dawn-to-dusk, which also excludes water) found mixed blood-marker results: one found an autophagy-initiation gene up and two other autophagy markers down (Dastghaib et al., 2025) [10], and another found three autophagy-related genes up with no non-fasting comparison group (Bou Malhab et al., 2025) [11]; neither measured autophagic flux. A dedicated 2026 pilot study of a 72-hour human fast did not measure autophagy directly at all, inferring only a metabolic environment described as "conducive to" it (Lauc et al., 2026) [12]. Some of the specific "hours until autophagy" framing seen in general circulation traces back to animal studies, most visibly mouse neuronal autophagy research (Alirezaei et al., 2010) [13], not human data.

Where the studies disagree

The Jamshed trial's results contain a tension worth stating plainly rather than resolving on this page's own authority: LC3A (autophagy-related) rose while MTOR, the gene for a protein that generally holds autophagy back (general cell biology, not a finding of this study), also rose, at different times of day, in the same small study; the authors themselves conclude the schedule "may also increase autophagy" (Jamshed et al., 2019) [1]. In the first hours of a fast, the Islam trial found no significant rise in muscle autophagy markers (Islam et al., 2021) [3]. Across a longer, 36-hour fast, autophagy marker proteins in muscle fell rather than rose, and only in untrained participants (Dethlefsen et al., 2018) [4]; at 72 hours, one muscle marker rose but so did a protein that autophagy is supposed to clear, which its own authors say complicates the result (Vendelbo et al., 2014) [5]. An 8-week trial of repeated 24-hour fasts found one muscle marker rise alongside a fall in several autophagy-related genes, which its authors attribute, as likely, to weight loss (Chaudhary et al., 2022) [9]. The flux studies in this set, the closest thing here to a direct measurement, disagree on how easily a fasting signal shows up at all: one found a rise detectable only under specific laboratory conditions and only in one white-blood-cell type (Pietrocola et al., 2017) [6], and another, a 6-month randomized trial, found no significant within-group rise, though it differed from standard care at 6 months in a post hoc analysis (Bensalem et al., 2025) [8]. The review papers in this set describe the mechanistic evidence for fasting-induced autophagy or mitophagy as "overwhelmingly" supportive, largely across species (Bagherniya et al., 2018 [14]; Mehrabani et al., 2020 [15]); the mitophagy review also calls that evidence "still limited" and methodologically heterogeneous (Mehrabani et al., 2020) [15]. The honest synthesis: autophagy is a real, well-established cellular process, robustly induced by fasting in animal models, and plausibly active in humans at some point during a fast, but the human evidence, which disagrees on direction, duration, and tissue, does not currently support stating a specific hour count, and any source that states one is going beyond what these studies show.

Authors and funding of the cited studies

  • Lauc et al., 2026 [12]: the paper's funding statement reads, "Funding provided by St. Catherine Specialty Hospital, Genos, the International Center for Applied Biological Research, and the International Center for Applied Biological Sciences." Two of those funders, Genos Glycoscience Research Laboratory and St. Catherine Specialty Hospital, are also listed among the authors' own institutional affiliations. The paper's separate competing-interest statement declares "no support from any organization for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work."
  • Bensalem et al., 2025 [8]: a UK patent, GB2603664B, "Methods and products for assessing lysosomal system flux" (assignee: South Australian Health and Medical Research Institute; published 2024), lists Julien Bensalem and Timothy Sargeant, the paper's first and last authors, as inventors (Google Patents record). PubMed lists the South Australian Health and Medical Research Institute (SAHMRI), the patent's assignee, as an affiliation of all 12 authors of the study. The paper's own competing-interests statement was not available to us.

Disclosures are listed where the full text of the paper was available to us; a study not listed here is not a statement that its authors have no ties.

References

  1. Jamshed H, Beyl RA, Della Manna DL, Yang ES, Ravussin E, Peterson CM (2019). Early Time-Restricted Feeding Improves 24-Hour Glucose Levels and Affects Markers of the Circadian Clock, Aging, and Autophagy in Humans. Nutrients. PMID 31151228 doi:10.3390/nu11061234 Finding: limited evidence
  2. Morales-Scholz MG, Wette SG, Stokie JR, Tepper BT, Swinton C, Hamilton DL, Dwyer KM, Murphy RM, Howlett KF, Shaw CS (2022). Muscle fiber type-specific autophagy responses following an overnight fast and mixed meal ingestion in human skeletal muscle. American Journal of Physiology-Endocrinology and Metabolism. PMID 35793481 doi:10.1152/ajpendo.00015.2022 Finding: limited evidence
  3. Islam H, Amato A, Bonafiglia JT, Rahman FA, Preobrazenski N, Ma A, Simpson CA, Quadrilatero J, Gurd BJ (2021). Increasing whole-body energetic stress does not augment fasting-induced changes in human skeletal muscle. Pflügers Archiv - European Journal of Physiology. PMID 33420549 doi:10.1007/s00424-020-02499-7 Finding: no effect found
  4. Dethlefsen MM, Bertholdt L, Gudiksen A, Stankiewicz T, Bangsbo J, van Hall G, Plomgaard P, Pilegaard H (2018). Training state and skeletal muscle autophagy in response to 36 h of fasting. Journal of Applied Physiology. PMID 30161009 doi:10.1152/japplphysiol.01146.2017 Finding: contradicts
  5. Vendelbo MH, Møller AB, Christensen B, Nellemann B, Clasen BF, Nair KS, Jørgensen JO, Jessen N, Møller N (2014). Fasting increases human skeletal muscle net phenylalanine release and this is associated with decreased mTOR signaling. PLoS One. PMID 25020061 doi:10.1371/journal.pone.0102031 Finding: limited evidence
  6. Pietrocola F, Demont Y, Castoldi F, Enot D, Durand S, Semeraro M, Baracco EE, Pol J, Bravo-San Pedro JM, Bordenave C, Levesque S, Humeau J, Chery A, Métivier D, Madeo F, Maiuri MC, Kroemer G (2017). Metabolic effects of fasting on human and mouse blood in vivo. Autophagy. PMID 28059587 doi:10.1080/15548627.2016.1271513 Finding: limited evidence
  7. Qian J, Fang Y, Yuan N, Gao X, Lv Y, Zhao C, Zhang S, Li Q, Li L, Xu L, Wei W, Wang J (2021). Innate immune remodeling by short-term intensive fasting. Aging Cell. PMID 34705313 doi:10.1111/acel.13507 Finding: supports
  8. Bensalem J, Teong XT, Hattersley KJ, Hein LK, Fourrier C, Dang LVP, Singh S, Liu K, Wittert GA, Hutchison AT, Heilbronn LK, Sargeant TJ (2025). Intermittent time-restricted eating may increase autophagic flux in humans: an exploratory analysis. Journal of Physiology. PMID 40345145 doi:10.1113/JP287938 Finding: limited evidence
  9. Chaudhary R, Liu B, Bensalem J, Sargeant TJ, Page AJ, Wittert GA, Hutchison AT, Heilbronn LK (2022). Intermittent fasting activates markers of autophagy in mouse liver, but not muscle from mouse or humans. Nutrition. PMID 35660501 doi:10.1016/j.nut.2022.111662 Finding: contradicts
  10. Dastghaib S, Siri M, Rahmani-Kukia N, Heydari ST, Pasalar M, Zamani M, Mokaram P, Bagheri-Lankarani K (2025). Effect of 30-day Ramadan fasting on autophagy pathway and metabolic health outcome in healthy individuals. Molecular Biology Research Communications. PMID 40028479 doi:10.22099/mbrc.2024.50105.1978 Finding: limited evidence
  11. Bou Malhab LJ, Madkour MI, Abdelrahim DN, Eldohaji L, Saber-Ayad M, Eid N, Abdel-Rahman WM, Faris ME (2025). Dawn-to-dusk intermittent fasting is associated with overexpression of autophagy genes: A prospective study on overweight and obese cohort. Clinical Nutrition ESPEN. PMID 39542136 doi:10.1016/j.clnesp.2024.11.002 Finding: limited evidence
  12. Lauc G, Brlek P, Bulić L, Šimunić Briški N, Šimunović J, Duvnjak Orešković I, Butumović L, Marjanović D, Klarić D, Petrović A, Tesla Frcko J, Primorac D (2026). Systemic metabolic, hormonal, and glycomic remodeling during a 72-hour fast in healthy adults: a pilot study. Croatian Medical Journal. PMID 42286908 doi:10.3325/cmj.2026.67.226 Finding: limited evidence
  13. Alirezaei M, Kemball CC, Flynn CT, Wood MR, Whitton JL, Kiosses WB (2010). Short-term fasting induces profound neuronal autophagy. Autophagy. PMID 20534972 doi:10.4161/auto.6.6.12376 Finding: limited evidence
  14. Bagherniya M, Butler AE, Barreto GE, Sahebkar A (2018). The effect of fasting or calorie restriction on autophagy induction: A review of the literature. Ageing Research Reviews. PMID 30172870 doi:10.1016/j.arr.2018.08.004 Finding: supports
  15. Mehrabani S, Bagherniya M, Askari G, Read MI, Sahebkar A (2020). The effect of fasting or calorie restriction on mitophagy induction: a literature review. Journal of Cachexia, Sarcopenia and Muscle. PMID 32856431 doi:10.1002/jcsm.12611 Finding: limited evidence