Autophagy, the cell's recycling system: what fasting actually does

Autophagy is the cell's recycling system: a set of pathways that wraps up worn-out proteins and damaged organelles, ships them to the lysosome, and breaks them down into raw materials the cell reuses. It is switched off when food is plentiful and switched on by energy stress, which is why fasting, exercise and caloric restriction all feature in the conversation. The honest position is that the molecular biology is Nobel-grade and robust, but most of the dramatic fasting-and-longevity claims come from yeast and mice. Direct human evidence is genuinely thin, and only in the last few years has anyone measured autophagy properly in living people.

What autophagy actually is

Every cell accumulates junk: misfolded proteins, protein aggregates, and mitochondria that have stopped pulling their weight. Autophagy is the controlled demolition crew. The best-studied form, macroautophagy, builds a fresh double membrane called a phagophore around a parcel of cytoplasm. That membrane closes into a sealed bubble, the autophagosome, which then fuses with a lysosome, the cell's acidic recycling vat. Enzymes inside the lysosome dismantle the cargo into amino acids, fatty acids and sugars that the cell feeds back into its own metabolism.¹

There are three broad routes to the lysosome. Macroautophagy (usually just called autophagy) uses those autophagosomes. Microautophagy lets the lysosome engulf small bits of cytoplasm directly. Chaperone-mediated autophagy is fussier: a chaperone protein recognises a specific tag on a target protein and threads it across the lysosomal membrane one molecule at a time.¹

Crucially, autophagy is not just bulk clearance. It can be selective, hunting down specific cargo. Mitophagy removes damaged mitochondria and depends on the PINK1 and parkin proteins, the same proteins implicated in Parkinson's disease. There are equivalent selective pathways for surplus lipid droplets, ribosomes and even invading bacteria. This is the genuinely important biology: failing autophagy is linked to neurodegeneration, certain cancers, infection and the broad decline of ageing.¹²

Ohsumi and the 2016 Nobel Prize

The phenomenon of cellular self-digestion was described in the 1960s, but nobody knew the machinery. That changed with Yoshinori Ohsumi, a Japanese cell biologist who took a deliberately unfashionable bet: study autophagy in baker's yeast, where he could watch it under a microscope and run genetics. In 1993 he published the discovery of a set of genes essential for autophagy in yeast, later named the ATG genes. He then showed that strikingly similar machinery operates in human cells.²

For this he was awarded the 2016 Nobel Prize in Physiology or Medicine, unshared, "for his discoveries of mechanisms for autophagy." The award citation is worth keeping in mind when reading wellness content: Ohsumi was recognised for working out how the recycling machinery is built and controlled, not for proving that skipping breakfast extends your life. Those are very different claims, and the gap between them is where most of the hype lives.²

How it is regulated: mTOR and AMPK

Autophagy is wired directly into how the cell senses food and energy, through two opposing master switches.

mTORC1 (mechanistic target of rapamycin complex 1) is the "you are fed, grow" signal. When amino acids, glucose and growth factors are abundant, mTORC1 is active, and it phosphorylates and silences ULK1, the kinase that kicks off autophagosome formation. Active mTORC1 therefore holds autophagy down.³

AMPK (AMP-activated protein kinase) is the "you are running low on fuel" signal. When cellular energy falls and AMP rises, AMPK switches on. It promotes autophagy two ways at once: it suppresses mTORC1, lifting the brake, and it also directly phosphorylates and activates ULK1, pressing the accelerator.³⁴

So the simple, broadly correct story is: high nutrients keep mTORC1 on and autophagy low; low energy turns AMPK on and autophagy up. That is exactly why fasting, exercise and caloric restriction are the usual suspects, since each tilts the balance towards AMPK. The caveat researchers now stress is that the real network is messier than that clean two-switch cartoon: the timing and tissue specifics vary, and the AMPK-mTOR relationship is not a simple seesaw in every context.⁴

What triggers it: fasting, exercise, compounds

Fasting and caloric restriction

Removing food lowers insulin, glucose, amino acids and growth-factor signalling, which quietens mTORC1 and rouses AMPK. In rodents this reliably raises autophagy markers, especially in the liver. Caloric restriction, eating fewer calories long term without malnutrition, is the most robust life-extending intervention across species, and increased autophagy is one proposed mechanism behind it.⁵ The leap that does not yet hold up is the precise human dosing: the popular "autophagy switches on at 16 hours" figure is an extrapolation, not a measured human threshold.

Exercise

Exercise is a powerful, repeated energy stress, and in rodent muscle it clearly raises autophagy markers. Human muscle is more confusing. In people, a single bout of exercise can transiently lower the marker LC3B-II before it returns to baseline, and the bulk-clearance marker p62 often does not move at all. That does not necessarily mean autophagy fell, since these static markers are poor proxies for actual flux, but it does mean human exercise data cannot simply be copied from the mouse.⁷

Compounds: spermidine and resveratrol

A class of molecules called caloric restriction mimetics can induce autophagy without you eating less. Spermidine, a polyamine found in wheat germ, soybeans, aged cheese and mushrooms, induces autophagy in cells and animals, partly by inhibiting the acetyltransferase EP300. Resveratrol, from grape skins, activates the SIRT1 pathway and is a strong autophagy inducer in cell and rodent models.⁸ The pattern that should make you cautious is consistent: both look excellent in a dish and in mice, but human outcome trials have been underwhelming, which we cover below. If you are weighing up any of these, our stack builder and our wider health library take the same evidence-first line.

The honest bit: human evidence versus mouse hype

This is where the article earns its keep. Almost everything you read about fasting "flipping on autophagy" rests on yeast, worms, flies and mice. Translating it to humans hits three hard problems.

Problem one: tissue and species differences are real. A 2022 study by Chaudhary and colleagues, published in Nutrition, found that intermittent fasting raised autophagy markers in mouse liver but did not activate them in mouse muscle. In humans, fasting did not raise muscle autophagy markers either, and a 24-hour fast moved one marker (SQSTM1) while a 12-hour overnight fast lowered several. The picture is patchy and tissue-specific, not a clean whole-body switch.⁵

Problem two: we could barely measure it in people. Most human studies have relied on static snapshots of proteins or mRNA, which tell you how much machinery is present, not how fast it is running. As a 2021 review in Trends in Molecular Medicine put it bluntly, measuring autophagy in humans has been an underappreciated barrier to translation. Researchers have only recently developed blood-based assays that estimate autophagic flux, and even those are limited to circulating immune cells, not your brain or liver.⁶

Problem three: the few proper human trials are small and recent. In 2025, Espinoza and colleagues published the first randomised trial in GeroScience to directly measure autophagic flux during a dietary intervention: 30 healthy adults on a 5-day fasting-mimicking diet showed increased flux in blood immune cells alongside better metabolic markers.⁹ The same year, Bensalem and colleagues reported in The Journal of Physiology that six months of intermittent fasting plus time-restricted eating raised an autophagy marker in blood cells versus standard care, but the between-group difference reached significance only at six months and only in a post-hoc analysis (P = 0.04).¹⁰ These are encouraging, genuinely first-of-their-kind results, and also exactly the kind of small, blood-based, statistically fragile findings that should not be oversold.

The compound story tells the same cautionary tale. The SmartAge trial (Wirth, Schwarz and colleagues, 2022, JAMA Network Open) randomised 100 older adults with subjective cognitive decline to spermidine at 0.9 mg per day or placebo for 12 months. Despite strong mechanistic hopes, spermidine did not improve memory over placebo.¹¹ Resveratrol has a similarly long trail of impressive lab results and underwhelming human outcomes. And rapamycin, the drug that directly inhibits mTOR, extends lifespan in mice (the NIA Interventions Testing Program reported median lifespan increases of roughly 9% in males and 14% in females), yet there is still no human trial proving it slows ageing, and its use for that purpose is off-label.¹²

Evidence-graded takeaways

• Established: autophagy is a real, essential, lysosome-based recycling system, regulated by mTORC1 and AMPK, and its failure contributes to disease. This is Nobel-grade biology.²
• Reasonably supported: fasting, caloric restriction and exercise shift nutrient signalling in the direction that promotes autophagy, and raise autophagy markers strongly in animals.⁵
• Emerging and fragile: a handful of small 2025 human trials suggest fasting-style interventions can nudge autophagic flux upward in blood immune cells. Promising, not settled.⁹¹⁰
• Not supported: any specific claim that a set fasting window (16 hours, 24 hours) reliably triggers a defined, health-extending burst of autophagy in humans. The numbers come from rodents and extrapolation.
• Weak or negative: spermidine and resveratrol supplements for measurable human benefit; rapamycin as a proven anti-ageing therapy.¹¹¹²

Practically, the interventions that plausibly support autophagy, regular exercise, not constantly overeating, and a sensible eating pattern, are the same things that are good for you anyway, with or without the autophagy framing. That is reassuring: you are not missing a secret switch. For a structured way to think about adding interventions to your routine, our start here page walks through the logic, and you can browse the evidence behind specific compounds in our insights archive.

This article is educational and does not constitute medical advice, diagnosis, or a treatment recommendation. Medication uses described as “off-label” are not licensed for that purpose in the UK and should only be considered under qualified clinical supervision. Always speak to your GP, pharmacist, or a registered specialist before starting, stopping, or changing any treatment. If you have severe or alarm symptoms - unintentional weight loss, blood in your stool, difficulty swallowing, persistent vomiting, a fever, or severe pain - seek urgent medical care.