Autophagy is how a cell breaks down its own worn-out parts and recycles the materials, and the biology behind it won a Nobel Prize. It runs constantly as housekeeping and speeds up when fuel is scarce, which is why fasting, energy restriction and exercise all plausibly engage it. Almost everything solid about autophagy and health comes from cells and animals.
Measuring it in a living person is hard, so the confident claim that a particular fast switches it on in you is an extrapolation, not a demonstrated human outcome. This page explains the machinery and marks where the evidence stops.
Findings & Outcomes
Autophagy sits underneath a large part of this section. When time-restricted eating, a ketogenic diet and exercise are sold on a longevity promise, the mechanism being invoked is usually this one. The version told in fasting circles has run ahead of what has been shown in people, so this page keeps the machinery and the human evidence apart.
The Cell's Recycling System
Autophagy, from the Greek for self-eating, is how a cell breaks down and reuses its own worn-out parts. A double membrane forms around a piece of cargo, a clump of damaged protein or a worn-out organelle, closes into a bubble called an autophagosome, and fuses with the lysosome, where enzymes break the contents down and return the raw materials to the cell. It is recycling and quality control in one process.
Autophagy has two speeds. A low level runs constantly as housekeeping, clearing the daily buildup of damaged components before it accumulates. Under starvation the process speeds up sharply, because a cell cut off from outside food can break down its own least essential parts to free amino acids and keep going. The core genes were first worked out in yeast by Yoshinori Ohsumi, who received the 2016 Nobel Prize in Physiology or Medicine for the work, and the same machinery is conserved from yeast to humans.
That autophagy exists and does this job is settled biology. The open questions here are not whether it happens, but how much any particular human behavior changes it, and whether that change does anything you would notice.
The Switches: mTOR And AMPK
Autophagy is regulated by how much fuel the cell has, through two enzymes that act in opposite directions on the same target. mTOR is the growth sensor: when nutrients and insulin are plentiful, mTOR is active and holds autophagy down, so the cell builds rather than recycles. AMPK is the energy sensor: when fuel runs low, AMPK is activated and turns autophagy up.
Both act on the same initiating enzyme, ULK1. When food is abundant, mTOR phosphorylates ULK1 at a site that keeps it inactive and blocks AMPK from engaging it. When energy drops, mTOR releases ULK1 and AMPK phosphorylates it at a different site to switch it on. This is why several different practices produce the same output: low insulin, low energy availability, and drugs that inhibit mTOR all converge on this one enzyme. It is also why the practices that plausibly raise autophagy are the ones that lower mTOR or raise AMPK: going without food, restricting energy, and exercising.
What Turns It On
Three triggers are solid in cells and animals.
Fasting. In mice, 24 hours without food produced a sharp rise in autophagy, including in neurons, a tissue once thought to be spared. This is a clean demonstration in a living animal that going without food drives the process. A mouse metabolism runs far faster than a person's, so these fasting durations do not transfer directly to people.
Energy restriction. In the worm C. elegans, eating less extends lifespan, and knocking down the essential autophagy genes abolishes that extension. In this model autophagy is not only switched on by eating less, it is part of why eating less lengthens life. The same theme runs through spermidine, a compound in foods like wheat germ and aged cheese that induces autophagy and extends lifespan in yeast, flies and worms, and through rapamycin, an mTOR inhibitor that extends lifespan in mice even when started in old age.
Exercise. Mice engineered to keep normal everyday autophagy but unable to ramp it up during exercise had reduced endurance and lost the blood-sugar benefit that exercise normally provides. Removing the exercise-induced pulse of autophagy removed one of exercise's metabolic payoffs, which ties the two together, in a mouse.
Every item on that list was measured in yeast, worms, flies, mice, or cells in a dish. That is the shape of the field.
What We Actually Know In Humans
Nearly all the evidence above is cellular or from animals. In people the picture is much thinner, for a technical reason: autophagy is hard to measure in a living human.
Autophagy is a rate, not a fixed amount. Its most-cited marker, LC3-II, rises both when autophagy speeds up and when it stalls at the final disposal step, so a single measurement cannot tell the two apart.
The field's consensus guidelines make this the central point. Autophagy is a flux, the rate at which cargo is captured and degraded. Distinguishing a true rise from a stalled pathway needs flux assays that are routine in cultured cells and largely impractical in living people. So a single blood marker cannot show that a fast "switched on autophagy" in you.
The best direct human signal so far is small. In 11 adults with overweight, eating within an early six-hour window instead of a twelve-hour one raised expression of one autophagy gene, LC3A, and lowered 24-hour average glucose by about 4 mg/dl, over four days. LC3A messenger RNA is a single upstream marker, not a measurement of autophagy running, so this is a hint of a signal in a small crossover trial, in eleven people, over four days. It is a long way from a health outcome driven by autophagy.
None of this means fasting does nothing, or that autophagy fails to respond in people. A lack of measurement is not a lack of effect. The mechanism is well worked out in simpler systems, but whether a sixteen-hour fast meaningfully changes autophagy in your body, and whether that would help you, has mostly not been shown either way.
Evidence
The Research & Studies
Everything here is based on the research we have collected and checked, sorted into groups and ordered with the strongest evidence first. Click any claim to open the studies behind it.
How it works
Cells wrap up their worn-out parts and recycle them in the lysosome
Autophagy, meaning self-eating, is how a cell clears out its own broken parts and reuses the raw materials. It runs at a low level constantly and speeds up when food is scarce.
This review lays out the process: a double membrane forms around cytoplasmic cargo, damaged proteins and organelles, closes into an autophagosome, and fuses with the lysosome where the contents are degraded and the building blocks returned to the cell. The genes that run it were first identified in yeast, work for which Yoshinori Ohsumi received the 2016 Nobel Prize in Physiology or Medicine, and are conserved across eukaryotes. Basal autophagy is constant quality control; starvation-induced autophagy is a survival response that liberates nutrients when external supply is cut off.
Who this may not transfer to:A mechanism established in cells across eukaryotes, including human cells; there is no participant group.
The study · 1
Mizushima, Autophagy: process and function · Genes Dev 2007
mTOR keeps autophagy off when fuel is plentiful; AMPK switches it on when fuel is low
Autophagy has an on-off switch tied to how much fuel the cell has. Plenty of food keeps it off through a sensor called mTOR; running low turns it on through a fuel gauge called AMPK.
Kim and colleagues showed that AMPK promotes autophagy by directly phosphorylating and activating ULK1, while under nutrient-rich conditions mTOR phosphorylates ULK1 at a different residue that prevents AMPK from engaging it. So the same initiating enzyme is held off by the growth sensor and switched on by the energy sensor, which is the molecular reason fasting, energy restriction and mTOR inhibitors all raise autophagy. This is biochemistry and cell-culture evidence, not a human outcome.
Who this may not transfer to:Biochemistry and cell-culture work, including human cell lines; not a measured human outcome.
The study · 1
Kim et al., AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1 · Nat Cell Biol 2011
No simple blood test shows autophagy running, because it is a flux markers read two ways
There is no simple blood test that shows autophagy is running in you. The usual markers can move for opposite reasons, and measuring it properly needs techniques that are hard to do in living people.
The consensus guidelines for monitoring autophagy set out that autophagy must be assessed as a flux, the rate at which cargo is delivered and degraded, not as the level of any one marker. LC3-II, the most-cited marker, accumulates both when autophagy is induced and when the final lysosomal degradation step is blocked, so a single measurement is ambiguous without a flux control. These controls are routine in cultured cells but largely impractical in living humans, which is why confident claims that a set fasting duration triggers autophagy in a person are extrapolations from cells and animals rather than direct measurements.
Who this may not transfer to:A statement about measurement drawn from consensus assay guidelines, not a human outcome.
The study · 1
Klionsky et al., Guidelines for the use and interpretation of assays for monitoring autophagy (4th edition) · Autophagy 2021
A 24-hour fast sharply raised autophagy in mouse neurons once thought spared
When mice were fasted for a day or two, autophagy switched on strongly even in brain cells, which had been assumed to be spared.
Alirezaei and colleagues found that a short fast produced profound upregulation of autophagy in the neurons of mice, measured by conversion of the LC3 marker and the appearance of autophagosomes, rising at 24 hours and further at 48 hours. The result mattered because neurons had been considered protected from starvation-induced autophagy. It is a clean demonstration that fasting drives autophagy in a living organism, in a specific tissue.
Who this may not transfer to:Measured in mice, whose metabolism runs far faster than a person's, so the fasting durations do not transfer directly.
The study · 1
Alirezaei et al., Short-term fasting induces profound neuronal autophagy · Autophagy 2010
Eating less extends worm lifespan only when the autophagy genes work
In worms, eating less makes them live longer, but only if their autophagy genes are working. Switch autophagy off and the life extension from dietary restriction disappears.
Hansen and colleagues showed that dietary restriction in C. elegans requires functional autophagy genes to extend lifespan: knocking down core autophagy genes prevented the lifespan extension that dietary restriction otherwise produced, and blunted the longevity of related low-nutrient-signalling mutants. It is among the cleanest evidence that autophagy is part of why eating less lengthens life, in an invertebrate model.
Who this may not transfer to:Established in the worm C. elegans; there is no human equivalent and it could not be tested the same way.
The study · 1
Hansen et al., A role for autophagy in the extension of lifespan by dietary restriction in C. elegans · PLoS Genet 2008
Spermidine switched on autophagy and extended lifespan in yeast, flies and worms
Spermidine, a compound in foods like wheat germ and aged cheese, switched on autophagy and lengthened life in yeast, flies and worms, and its life-extending effect needed autophagy to be working.
Eisenberg and colleagues reported that spermidine triggered autophagy and extended lifespan in yeast, fruit flies and nematodes, and suppressed markers of oxidative stress and necrosis in cultured human immune cells. When autophagy genes were deleted, the lifespan benefit disappeared. Spermidine has since been studied as a caloric-restriction mimetic, a compound that reproduces some effects of eating less without the food restriction.
Who this may not transfer to:Model organisms plus cultured human immune cells; not a demonstrated human lifespan effect.
The study · 1
Eisenberg et al., Induction of autophagy by spermidine promotes longevity · Nat Cell Biol 2009
Mice that could not ramp up autophagy lost exercise's blood-sugar benefit
Mice that could do everyday autophagy but could not ramp it up during exercise did not get the usual blood-sugar benefit from exercise, suggesting the autophagy boost is part of why exercise helps.
He and colleagues made a knock-in mouse whose autophagy machinery could not be stimulated by exercise or starvation while basal autophagy stayed normal. These mice showed decreased endurance during acute exercise and did not gain exercise's usual protection against high-fat-diet-induced glucose intolerance. Removing the exercise-induced pulse of autophagy removed one metabolic benefit of exercise, tying the two together in a mouse.
Who this may not transfer to:A mouse-genetics result; exercise-induced autophagy has not been measured this way in people.
The study · 1
He et al., Exercise-induced BCL2-regulated autophagy is required for muscle glucose homeostasis · Nature 2012
Longevity And Mortality
Rapamycin extended mouse lifespan by 9 to 14 percent even when started in old age
A drug that blocks the mTOR sensor, one of the ways autophagy gets switched on, made mice live longer even when given in old age.
In the NIA Interventions Testing Program, rapamycin fed to genetically heterogeneous mice starting at 600 days of age (roughly age 60 in human terms) extended both median and maximal lifespan. On the basis of age at 90 percent mortality, the increase was 9 percent for males and 14 percent for females, and the effect held across three independent test sites. mTOR inhibition raises autophagy among its effects, and the result has been repeated across labs, making it one of the more robust pharmacological lifespan extensions in mammals. Rapamycin also has immunosuppressive and metabolic side effects.
Who this may not transfer to:Measured in mice; rapamycin has immune and metabolic side effects and no human lifespan effect has been shown.
The study · 1
Harrison et al., Rapamycin fed late in life extends lifespan in genetically heterogeneous mice · Nature 2009
Heart And Vascular
People eating more spermidine had lower blood pressure and less heart disease, an association
People who ate more spermidine-rich food had lower blood pressure and less heart disease. This is an association in a survey, not proof that spermidine caused it.
Eisenberg and colleagues paired animal work, spermidine fed to mice reduced cardiac ageing and extended lifespan, with a human epidemiological analysis in which higher dietary spermidine intake correlated with reduced blood pressure and a lower incidence of cardiovascular disease. The human portion is observational: it shows a correlation, and dietary spermidine intake travels with a plant-rich diet, so it cannot on its own establish that spermidine is the active ingredient.
Who this may not transfer to:A mixed-sex community cohort; the mechanism arm was in mice, so the human data point to spermidine as a correlation and cannot show it is the cause on their own.
The study · 1
Eisenberg et al., Cardioprotection and lifespan extension by the natural polyamine spermidine · Nat Med 2016
Blood Sugar
A 6-hour eating window nudged up one autophagy gene and cut 24-hour glucose 4 mg/dl in 11 adults
In a small crossover trial, squeezing meals into a 6-hour morning window nudged up one autophagy gene and modestly lowered average blood sugar compared with a normal eating window.
Jamshed and colleagues ran a 4-day randomized crossover in 11 adults with overweight comparing early time-restricted feeding (8am to 2pm) with a 12-hour window (8am to 8pm). Early time-restricted feeding increased morning expression of the autophagy gene LC3A (p<0.04) and the ageing gene SIRT1, decreased mean 24-hour glucose by 4 plus or minus 1 mg/dl and reduced glycaemic swings, and shifted circadian clock gene expression. LC3A messenger RNA is a single upstream marker of autophagy capacity, not a measurement of autophagic flux, so this shows a hint of an autophagy signal rather than proof that autophagy increased.
Who this may not transfer to:Eleven adults with overweight, 7 men and 4 women, over four days, measured through a single gene-expression marker instead of autophagy running.
The study · 1
Jamshed et al., Early Time-Restricted Feeding Improves 24-Hour Glucose Levels and Affects Markers of the Circadian Clock, Aging, and Autophagy in Humans · Nutrients 2019
Two notes on reading these findings. Many are graded animal or mechanistic, because that is
where the work was done and you cannot randomize a person to a different autophagy system. And the
grades are deliberately uneven: the mechanism findings rest on firm ground, while the two human
findings are graded emerging and marked as observational or single-marker, which is exactly what they
are.
Go Deeper
This page describes the machinery. The practices that plausibly act on it:
- Time-restricted eating and the ketogenic diet, which lower insulin and mTOR for longer stretches of the day.
- Hormesis, the general principle that a controlled stress drives an adaptation, and mitochondria, the other cellular system those same practices act on.
Common Questions
What is autophagy in simple terms?
It is the cell's own recycling and clean-up system. A membrane wraps around damaged proteins and worn-out parts, delivers them to the lysosome to be broken down, and the raw materials are reused. A low level runs all the time as housekeeping, and it speeds up when the cell is short of fuel, which is why fasting and energy restriction are linked to it.
Does a sixteen-hour fast really trigger autophagy?
In animals, fasting clearly drives autophagy. In humans it is hard to say, because autophagy is a rate that is difficult to measure in a living person, and the popular sixteen-hour figure comes from extrapolating animal work rather than from measuring people. The best human study so far nudged a single autophagy gene upward in eleven adults over four days, which is a hint, not a demonstration that a set fasting length switches the process on in you.
Why is autophagy so hard to measure in people?
Because it is a flux, a rate of capture and disposal, not a fixed amount. The usual marker can rise either because autophagy sped up or because it jammed at the final step, so a single snapshot is ambiguous. Telling the two apart needs techniques that work in cultured cells but are largely impractical in a living human, so there is no simple blood test that shows autophagy is running.
If most of the evidence is from animals, is fasting pointless?
No. A lack of measurement is not the same as a lack of effect. The mechanism is well worked out in simpler systems, and fasting and time-restricted eating have their own human evidence for things like glucose and weight, covered on their own pages. What has not been shown is the specific step in the popular story: that a given fast produces a measured autophagy increase in you, and that this is what delivers a health benefit.
Do spermidine or rapamycin supplements boost autophagy for longevity?
Both induce autophagy and extend lifespan in animals, and higher dietary spermidine intake tracks with better heart outcomes in people, though that is an association tangled up with an overall plant-rich diet. Rapamycin extends lifespan in mice but has immune and metabolic side effects, so off-label use for longevity is risky and not backed by human evidence. Neither has been shown to extend human life.
The Chinese Medicine Reading
Explore Related
Other pages this one connects to, by the evidence they share, the outcomes they touch, and the ground they cover.
All 10 sources on this page independently checked and cross-referenced.
Thomas Dehli, Founder & Editor, Sacred Lotus
Sacred Lotus has published Chinese medicine reference material since 2001. Integrative pages are held to the same standard as the herb and formula library: cite the source, grade the claim at its real strength, and say where the research has not looked. This page is educational and it is not medical advice. Last reviewed and updated August 10, 2026.
Evidence strength
How confidently the research supports a claim. Strength describes the evidence, not our endorsement.