Autophagy: What Humans Actually Have Evidence For

Autophagy is your cells' built-in recycling program. When a protein gets damaged or an organelle wears out, the cell wraps the junk in a double-membrane bubble (an autophagosome), ships it to the lysosome (the cell's acid bath), and breaks it down into amino acids, fatty acids, and nucleotides. Then it reuses those parts. Think of it as housekeeping.

Like all housekeeping, it ticks along quietly most of the time. It ramps up under stress: when you starve, when oxygen runs low, during infection, or when misfolded protein junk piles up. And when it breaks, things go wrong in predictable ways.

We know all of this thanks to one person. Yoshinori Ohsumi ran a yeast experiment in 1993 that pulled out the first autophagy genes. He starved yeast of nitrogen and hunted for mutants that could no longer pile up autophagic bodies. Out came 75 mutants in 15 groups, and at least 15 genes (the APG genes) turned out to drive autophagy in yeast. ^[2] That one screen turned autophagy from a microscopy curiosity into a genetic pathway scientists could actually work with. The Nobel committee gave him the 2016 Nobel Prize in Physiology or Medicine for that single line of work. People had spotted autophagosomes under electron microscopes back in the 1960s, but it took Ohsumi's 1990s work to show why they matter for health and disease.

Five years later came the recycling machinery itself. A 1998 Nature paper showed that the cell uses a special tagging system to make autophagy happen at all, and that this system is essential. ^[3] One protein (Apg12) gets locked onto another (Apg5) by a helper enzyme. It looked exactly like the tagging system the cell already uses to flag proteins for the trash, except this was the first time that trick showed up outside that system. The same machinery runs in your cells too.

Then scientists found a marker they could actually measure. A 2000 paper identified a protein called LC3 as the human-cell version of one of Ohsumi's yeast proteins. ^[4] One form floats free in the cell. The other gets a fatty tail and sticks to the autophagosome membrane, and the more of that membrane-bound form you see, the more autophagosomes are being built. Twenty-five years on, that membrane-bound LC3 is still the go-to readout in human muscle-biopsy studies.

There is a second marker worth knowing: a protein called p62. A 2005 paper described it. ^[5] It grabs tagged junk with one hand and LC3 with the other, then drags the cargo to the autophagosome. Because p62 gets eaten along with the junk, its level in tissue tells you something about how fast autophagy is running. With one big catch: lots of p62 plus lots of LC3 usually means autophagy is jammed, not flowing fast.

So how do you measure any of this properly? The reference manual is a set of consensus guidelines, fourth edition, published in 2021. ^[7] Over 2,800 authors, 382 journal pages. The one line every reader should remember: a single measurement (like a one-off LC3 reading) is not enough. You cannot judge how fast autophagy is running off one snapshot. You need to track the cargo over time, or compare paired tissue samples, or block the lysosome and see what backs up.

Here is the part that matters for you. Those same 2021 guidelines also say that LC3 and p62 in blood are unreliable stand-ins for what is happening inside your tissues. So when a private DACH clinic offers you an "Autophagie-Panel" from a vial of blood, the reference manual flatly disagrees with the sales pitch.

One last reframe. Autophagy is renovation, not demolition. As a 2011 review put it, autophagy makes fresh building blocks and energy so the cell can renovate itself. ^[6] The popular "detox" framing badly misses the point. This is not the garbage truck. It is the cell's main recycling and rebuilding crew.

Autophagy is not supplement-marketing wishful thinking. It is real aging biology, and the best proof is the 2023 "Hallmarks of aging" update. ^[9] That paper promoted autophagy from a sub-item under proteostasis to its own standalone hallmark of aging.

There are twelve hallmarks in total: genomic instability, telomere attrition, epigenetic changes, loss of proteostasis (the cell's ability to keep its proteins folded right), broken-down autophagy, deregulated nutrient-sensing, mitochondrial dysfunction, cellular senescence (cells that stop dividing but refuse to die), stem cell exhaustion, garbled cell-to-cell communication, chronic inflammation, and dysbiosis (an out-of-balance gut microbiome). To make this list, a process has to do three things: show up with age, speed up aging when you crank it experimentally, and be at least partly reversible. Autophagy clears all three bars.

The animal evidence is overwhelming. A 2018 review in Nature Reviews Molecular Cell Biology landed one striking finding. ^[8] Take every major life-extending trick that works in yeast, worms, flies, and mice (eating less, dialing down growth signals, dialing down insulin signaling, stressing the cell's power plants, giving spermidine). Every single one needs working autophagy genes to deliver the extra lifespan. Knock those genes out and the benefit vanishes. Autophagy is the common thread running through every validated lifespan-extending intervention in lab animals.

The disease-side companion is a 2019 Cell review ^[10], a follow-up a decade on from the same authors' 2008 landmark ^[11]. The 2019 review adds an honest note: those autophagy genes also do plenty of other jobs around the cell, well beyond recycling. So "more autophagy = better" is wrong as a slogan. Autophagy is essential, it depends on context, and it is pleiotropic (it does many different things at once).

So why does the mouse-to-human jump only get us halfway? Three problems pile up.

1. Measuring it. In a mouse you can take the animal apart and read the markers straight from liver, brain, heart, and muscle. In a human you get a skin punch, the occasional muscle sample in a research lab, and blood markers the 2021 guidelines explicitly warn you not to trust.
2. Time. A mouse lifespan study runs three years. A human lifespan study runs an entire career. The shortcuts (aging clocks like DunedinPACE, biomarker panels) are still being validated.
3. The cancer twist. Early on, autophagy blocks cancer from getting started: a partial loss of one autophagy gene (Beclin-1) shows up in 40-75 % of sporadic breast and ovarian cancers ^[15]. But once a tumour is up and running, autophagy helps it survive chemo. So "boost your autophagy with spermidine and resveratrol" is not safe blanket advice if you have active or recently treated cancer.

The honest 2026 read: autophagy is real aging biology with overwhelming evidence in animals. The human translation is partial, the measurement is hard, and the simple "more is better" hunch is wrong.

Two opposing switches control autophagy, and both watch one thing: does the cell have plenty of food, or is it running on fumes. One switch is called mTOR. The other is called AMPK. The cleanest paper laying out how they work is a 2011 Nature Cell Biology study. ^[12] In plain terms: when glucose runs low, AMPK flips autophagy on by tagging a third protein called ULK1, and at the same time mTOR keeps autophagy off by tagging that same ULK1 in a different spot and elbowing AMPK out of the way.

So three tags on one protein (ULK1) decide whether autophagy runs. AMPK hits the gas. mTOR hits the brake. Same protein, opposite signals.

mTOR is the "food is here, time to grow" sensor. Amino acids (especially one called leucine), insulin, and a full energy tank all switch it on. When mTOR is on, the cell builds protein and stores fat, and autophagy gets shut down. The go-to review here is a 2017 Cell overview, which notes that when mTOR signaling goes haywire, it feeds into cancer, diabetes, and aging itself. ^[13] Rapamycin is the textbook mTOR blocker. Its lifespan effect in mice (more on that later) works by taking mTOR's foot off the brake so autophagy can run, plus a few other downstream effects.

AMPK is the opposite: the low-fuel sensor. When the cell's energy charge drops, AMPK switches on. Going without glucose and exercising your muscles hard are the classic triggers. The go-to review is a 2012 Nature Reviews Molecular Cell Biology overview. ^[14] AMPK starts autophagy, stops the cell from making fat, and shuts mTOR down. Metformin links to autophagy through AMPK.

From there, the machinery assembles around a protein called Beclin-1, which acts as the assembly hub. The founding Beclin-1 paper from 1999 is literally titled "Induction of autophagy and inhibition of tumorigenesis by beclin 1." ^[15] A partial loss of Beclin-1 shows up in 40-75 % of sporadic breast and ovarian cancers, and that is the historical reason people call autophagy a tumour suppressor.

There is also a side route that skips mTOR entirely. A 2014 Molecular Cell paper showed you can trigger autophagy just by draining a particular cellular fuel molecule (acetyl-CoA) and dialing down the enzyme that uses it, no mTOR required. ^[16] The companion review coined a name for natural compounds that work this way instead of hitting mTOR directly: caloric restriction mimetics. ^[17] This is the route that connects spermidine and resveratrol to autophagy.

The practical version. Anything that lowers insulin (fasting, eating less protein), drains the cell's energy charge (sustained exercise, ketosis), or empties out that fuel molecule (spermidine, resveratrol) should, in principle, trigger autophagy. Whether it actually does so in human tissue, at doses you can tolerate, with a benefit you can measure, is exactly what the rest of this guide tries to grade honestly.

No. There is no peer-reviewed human muscle-biopsy study showing clean autophagy switching on at 16 hours of fasting. The cleanest human data shows the markers move at roughly 72 hours, not 16, and the one randomized 16:8 trial (TREAT, n=116) was flat negative on weight and metabolism ^[24]. The 16-hour number is mostly mouse data scaled to humans. This is the most important section in the guide, where popular DACH wellness messaging drifts furthest from what humans have actually been measured doing, so read it even if you skim the rest.

The popular claim. Almost every German lifestyle magazine, fasting-clinic brochure, and longevity influencer repeats some version of "ab 16 Stunden Fasten setzt die Autophagie ein" or "Autophagie startet nach 16 Stunden." Apotheken Umschau, FOCUS Gesundheit, almost every Buchinger-aligned book. And the claim usually gets pinned, openly or by hint, on Ohsumi's Nobel research.

Now here is what humans have actually been measured doing. The cleanest published human study that biopsied tissue and read autophagy markers is a 2014 PLOS One trial. The popular press almost never cites it. ^[49]

The setup: 8 healthy men, each studied twice. Once after a 10-hour overnight fast, once after 72 hours of fasting. Thigh-muscle (vastus lateralis) biopsies both times. The finding at 72 hours: the membrane-bound autophagy marker LC3B-II rose by roughly 30 %, p62 rose by roughly 10 %, and the mTOR and ULK1 signals dropped.

The authors themselves flag two catches:

1. When LC3B-II and p62 both rise together, you cannot tell what is really going on. Without a control that blocks the lysosome or tracks the cargo over time, you cannot cleanly tell "more autophagosomes are forming" (good, that is what you want) from "autophagosomes are piling up because the lysosome cannot clear them fast enough" (bad, a jam).
2. n = 8, healthy young men, one muscle. No women, no older adults, no other tissues. This is the best data we have. It is still small.

Now the part that matters: there is no comparable peer-reviewed muscle-biopsy study in healthy adults showing clean autophagy turning on at 16 hours of fasting. Recent reviews of intermittent fasting and tissue autophagy markers land in the same place. Mice switch on liver autophagy easily on their version of a 16-hour fast. Human muscle, in the published biopsies, does not. The response is patchy, depends on the tissue, and often runs in opposite directions between mouse and human.

So where did the 16-hour number come from? Three sources converged.

1. Scaling mouse data to humans. A mouse burns through fuel about 7x faster per kilo than you do. A 12-16-hour fast is when autophagy peaks in mouse liver. Naively scaling that to humans spits out "about 16 hours." But human metabolism does not map onto a mouse one-to-one.
2. The Ohsumi halo. Ohsumi won the 2016 Nobel Prize and gave a lot of interviews. He consistently talked about yeast and mouse biology. As far as anyone can tell, he never endorsed a specific human fasting time. The "Ohsumi said 16 hours" line is a misattribution.
3. Popular fasting books and clinic influencers. Jason Fung's The Obesity Code (2016), Mark Mattson's public talks, and Buchinger-Wilhelmi-style fasting literature turned rodent thresholds into one fixed human number. The German wellness press repeated it until it stuck.

What the actual 16:8 trial evidence shows. The 2020 TREAT randomized trial enrolled n = 116 adults with overweight or obesity. ^[24] One group ate only between noon and 8 pm; the control ate three structured meals a day. Twelve weeks. The result was a flat negative. The weight difference between the groups was -0.26 kg, P = 0.63, with no meaningful change in any of the secondary metabolic measures either. The authors' bottom line: on its own, with nothing else changed, time-restricted eating is no better for weight loss than eating across the whole day.

TREAT is the cleanest evidence that the 16:8 schedule by itself, the exact schedule DACH lifestyle media loves to sell as an "autophagy trigger," does nothing meaningful for your metabolism beyond any calorie change it happens to cause. And TREAT did not even see that calorie change.

Where fasting does have stronger human evidence. A 2020 Cell Metabolism trial (4- and 6-hour eating windows in n=58 adults with obesity) ^[22] and a 2017 Science Translational Medicine trial (three monthly 5-day fasting-mimicking-diet cycles in n=100) ^[23] both show real metabolic gains. But the engine is eating fewer calories, not a magic autophagy threshold.

The Madeo Graz lab's flagship alternate-day-fasting trial, published in Cell Metabolism in 2019, ran 60 healthy normal-weight adults over 4 weeks. ^[20] The results: weight loss, a better fat-to-lean ratio, lower sICAM-1 (an inflammation marker), lower fT3 (a thyroid hormone), and less belly fat. This is the strongest DACH-led human signal tying a fasting protocol to autophagy-relevant markers. Even here, direct muscle-biopsy data on LC3 and p62 stay thin.

The defensible 2026 read: the cleanest peer-reviewed change in human muscle autophagy markers shows up at roughly 72 hours of fasting in 8 healthy men, with measurement caveats. No peer-reviewed human trial shows clean autophagy turning on at 16 hours. The 16-hour number is mostly mouse extrapolation, repeated so often it started to sound official.

Yes, lowering protein (and especially the amino acid leucine) is the better-evidenced version of the fasting idea, because the mTOR switch is highly sensitive to leucine; turn that signal down and autophagy goes up. In the NHANES cohort, high protein at age 50-65 tracked with 75 % higher mortality, but the effect flips after 65, so the practical move is less protein in midlife and more later ^[25]. Protein restriction is the better-evidenced lever on the same biology as intermittent fasting.

The headline human study. A 2014 Cell Metabolism analysis followed n = 6,381 people from the NHANES 1988-94 survey for 18 years. ^[25] Adults aged 50-65 who ate a lot of protein had 75 % higher all-cause mortality and 4 times the cancer mortality of the low-protein group. Then the signal flipped after 65, where low protein became linked to higher mortality instead. Whether the protein came from plants or animals changed the effect. The catches: this is observational, protein was measured by food-frequency questionnaire, and healthy-user bias is likely.

The strongest mouse study. A 2014 Cell Metabolism experiment fed 25 different diets across 715 mice. ^[26] Lifespan peaked when protein was low and carbs were high. The proposed engine: the liver's mTOR, certain circulating amino acids, and glucose. The authors found that simply cutting calories through high-protein diets gave no lifespan benefit at all. That makes this the strongest mouse paper for one idea: how much and what kind of protein you eat moves longevity more than total calories do.

Cutting one amino acid. In a 2005 Aging Cell study, mice fed less of a single amino acid (methionine) lived longer, with lower IGF-1, lower thyroid hormone (T4), and lower insulin: the classic hormone mix that lets autophagy run. ^[27] Methionine is concentrated in animal protein. So the plant-leaning protection in the human cohort fits neatly.

The metformin link. A 2013 Cell study showed that in the worm C. elegans, metformin's lifespan benefit depends on the drug messing with how gut microbes handle folate and methionine. ^[28] That is the bridge connecting metformin to AMPK to autophagy, and connecting metformin to the gut microbiome to methionine restriction.

CALERIE-2: the cleanest human calorie-restriction trial. A 2019 Lancet Diabetes & Endocrinology report enrolled n=218 healthy normal-weight adults aged 21-50, randomised 2:1 to either 25 % calorie restriction or eating freely, for 2 years. ^[18] In practice people only managed about 12 %, not 25 %. Even so, the restriction group lost 7.5 kg, mostly fat, with lasting improvements in LDL cholesterol, blood pressure, the inflammation marker CRP, and insulin sensitivity. CALERIE-2 did not publish a formal time series of human autophagy markers, so the autophagy link here is mechanistic, not biopsy-proven.

The aging readout came later, from a 2023 Nature Aging analysis of DNA methylation (chemical marks on DNA that act as an aging clock) in the same trial. ^[19] That sample was n=220 randomised, with n=197 having usable methylation data (slightly off from the main trial's enrolment count). The restriction group aged measurably slower on the DunedinPACE clock than controls. The effect is small in absolute terms. The headline is the direction (a slowdown), not the size.

The practical translation. In your 50s and 60s, eating less protein (the protective range sat below 0.8 g per kg of body weight per day, well under the typical Western intake) with a plant lean is the lifestyle that best mimics the hormone mix autophagy likes ^[25]. After 65 the math flips (the same cohort showed the reversal), and the risk of losing muscle takes over. That is why the longevity community settled on "protein periodisation": a little less in midlife, more in older age.

Three molecules dominate the drug-and-supplement conversation about autophagy. The evidence behind them is wildly uneven.

Rapamycin: the strongest mammal lifespan signal

The landmark study is the 2009 NIA Interventions Testing Program (ITP), titled "Rapamycin fed late in life extends lifespan in genetically heterogeneous mice." ^[29] The team started rapamycin at 600 days of age (late middle age for a mouse). Measured at the age by which 90 % had died, the lifespan gains were:

• Females: +14 %
• Males: +9 %
• Mean lifespan, females: about +13 %
• Mean lifespan, males: about +9 %

Three independent ITP sites got the same result. It is still the only small molecule with a replicated lifespan boost when you start it in middle age in mammals. The autophagy link runs through blocking mTOR and letting ULK1 off the brake.

The credible human signal comes from a 2014 Science Translational Medicine trial of a rapamycin cousin. ^[30] The drug RAD001 (everolimus) boosted older adults' response to the flu vaccine by about 20 %, at doses people tolerated reasonably well. The 2018 follow-up trial ran n=264 older adults on a RAD001 plus dactolisib combination versus placebo, and at 1 year the treated group reported significantly fewer infections (P = 0.001). ^[31] This is the cleanest human evidence that an mTOR blocker improves a real immune readout in older people. The catch: that regimen mixed in a second mTOR-blocking drug, so it was not pure rapamycin.

Rapamycin in DACH practice. In Germany, sirolimus is licensed only to suppress the immune system after an organ transplant. Off-label longevity prescriptions happen on a Privatrezept (private prescription) through a small number of private clinics. Your insurer (GKV) does not cover it. Advertising is restricted under the HWG (the German drug-advertising law). Dosing varies a lot between providers. The longevity dose (usually 4-6 mg per week) sits below the transplant dose but above zero, and blood-level monitoring is not standardised in this off-label crowd. Our rapamycin in Germany guide covers the practical and legal detail.

Spermidine: the Madeo Graz story

The sister guide Spermidine and Longevity covers this in full. Two points matter most here:

1. A 2016 Nature Medicine study reported that spermidine extended mouse lifespan, improved heart function in old mice, and that in the Bruneck cohort, eating more dietary spermidine tracked with better cardiovascular health. ^[33] The lifespan boost disappeared when autophagy was genetically switched off. That makes it one of the cleanest mammal demonstrations that a benefit actually depends on autophagy.
2. The 2022 SmartAge trial: n=100, 12 months, roughly 0.9 mg/day of spermidine from wheat germ extract. ^[38] A negative trial. Memory performance did not improve between the groups. The longest, largest spermidine-and-cognition trial we have came up empty on its main goal. Do not soft-pedal that. It is the single most important counterweight to the Madeo lab's pre-clinical case and to DACH wellness marketing.

The newest piece of the puzzle is a 2024 Nature Cell Biology fasting study. ^[39] Fasting raises the body's own spermidine across species, human volunteers included, and when researchers blocked that pathway, fasting lost its lifespan and healthspan benefits. The mechanism: autophagy plus a separate cellular tweak. So the 2024 paper reframes spermidine as something fasting produces, rather than a standalone intervention.

The Bruneck cohort: the strongest human spermidine signal. This 2018 American Journal of Clinical Nutrition study followed n=829 adults aged 45-84 in South Tyrol for 20 years, during which 341 died. ^[34] The takeaway: people who ate more spermidine died less over those two decades. For each one-standard-deviation increase in dietary spermidine, the risk of dying from any cause was:

• Adjusted for age, sex and calorie ratio: HR 0.74 (95 % CI 0.66 to 0.83, P<0.001)
• With further lifestyle and diet adjustments: HR 0.76 (95 % CI 0.67 to 0.86, P<0.001)

It held up in a separate Salzburg cohort. But it is observational and based on food questionnaires: the strongest human signal we have, and still not proof of cause.

Resveratrol: the over-promised one

The mechanism starts with a 2010 Cell Death & Disease study showing that calorie restriction and resveratrol both extend life by switching on autophagy through a protein called SIRT1. ^[43] The deeper mechanism, via that cellular fuel molecule and its partner enzyme, holds up in cell models ^[16]. The honest counter-evidence: the NIA ITP mouse studies found resveratrol did not extend lifespan in genetically diverse mice on a standard diet. There are no human trials for hard outcomes. Resveratrol earns the same skepticism the CoQ10 guide aims at the ubiquinol absorption claim: a defensible mechanism, weak clinical evidence, and retail copy that oversells both.

Metformin (the TAME context)

The original 2016 TAME framing paper laid out the case. ^[45] As of 2026, TAME still has not reported hard-outcome data. The AMPK-to-autophagy mechanism is defensible (the worm data) ^[28]. The human longevity-outcome case is simply not there yet.

Exercise is the autophagy trigger nobody markets. It needs no supplement, no clinic, no fasting schedule. And the mechanism for exercise-driven autophagy in mammals is unusually clean.

The cleanest mouse paper. A 2012 Nature study is the landmark. The researchers bred special mice (called BCL2 AAA mice) with one tweak: their muscles can no longer turn on autophagy in response to exercise, even though their everyday baseline autophagy works fine. ^[46] The result in these mice:

• Worse endurance.
• Disrupted blood-sugar handling during exercise.
• Loss of exercise's protective effect against the blood-sugar problems a high-fat diet normally causes.

This is one of the cleanest mammal demonstrations that a benefit specifically needs working autophagy in the right tissue. Exercise protects you against insulin resistance partly by switching on autophagy in your muscles.

Human exercise data is thinner. Muscle biopsies in exercising people are hard to do, both technically and ethically. A 2015 FASEB Journal study comparing autophagy markers in human muscle after endurance versus resistance training gets cited a lot, but the pool of clean head-to-head human biopsy data is small.

The 2014 muscle-biopsy study from the fasting section above is still the most-cited human muscle autophagy readout in the field, and it tested fasting, not exercise. ^[49]

The practical version. Cardio training (Zone 2, intervals, a mix) drops your mTOR signal in the moment, raises your AMPK signal in the moment, and over time grows more mitochondria (your cells' power plants) through a master regulator called PGC-1α. All three roads lead to autophagy. The Zone 2 guide (Zone 2 Training & VO₂max) digs into the fitness-and-mortality evidence. Strength training counts too (losing muscle is itself a hallmark of aging), though its autophagy link is less clean.

The honest editorial point. Of everything DACH wellness media labels an "autophagy trigger" (16-hour fasting, autophagy supplements, autophagy fasting clinics), exercise has the cleanest mechanism in mammals (those BCL2 AAA mice) ^[46] and the strongest human survival evidence (the entire body of fitness-and-mortality research). It is also the one thing nobody can sell you in a bottle.

Read the autophagy-and-longevity research of the last 15 years and one name keeps appearing: Frank Madeo's lab at the University of Graz is on roughly 60 % of the high-impact papers. Autophagy as a longevity topic in DACH is not an American import. It grew up here, and the editorial angle of this guide reflects that.

The Graz network. Karl-Franzens-Universität Graz and BioTechMed-Graz are the central hub. The 2009 Nature Cell Biology paper on spermidine-induced autophagy ^[32], the 2016 Nature Medicine heart-protection paper ^[33], the 2018 Science spermidine review ^[35], the SmartAge trial run with Charité ^[38], and the 2024 Nature Cell Biology paper casting spermidine as fasting's own metabolite ^[39] all run through this network.

The Bruneck cohort ^[34] sits in South Tyrol, the German-speaking autonomous province of northern Italy, recruited and tracked through an Innsbruck and Bolzano collaboration. The validation cohort is Salzburg (Paracelsus Medical University). Both DACH-adjacent.

Charité Berlin and DZNE Berlin (the groups of Agnes Flöel, Claudia Schwarz, Miranka Wirth) ran the SmartAge cognition trial ^[38] and a 2018 Cortex pilot ^[36]. Berlin plus Graz is the working axis of human spermidine trials.

Fachhochschule Wiener Neustadt (the Pekar group) ran a 2020 Wiener Klinische Wochenschrift trial in people who already had dementia, in Styrian nursing homes (n=85). ^[42] The higher-dose group showed a small benefit on a standard cognition test (the MMSE). But the trial had no placebo arm, so read it carefully.

Longevity Labs+ (Graz, AT) is the spin-off that holds the EU Novel Food authorisation for spermidine-rich wheat germ extract. Its flagship product is spermidineLIFE. A fair chunk of DACH human-trial funding flows through Longevity Labs+ industry partnerships. That is a disclosed, normal academia-industry arrangement, not a scandal. But it is context worth carrying into the literature.

Why this matters. When DACH wellness media talk about "Autophagie-Fasten," "Autophagie Kur," or "Autophagie-Lebensmittel," they are usually pulling from this network's mouse and observational data. The Madeo group itself has been honest about the limits. Frank Madeo and Tobias Eisenberg have told German-language outlets, on the record, that eating wheat germ is a reasonable alternative to buying a spermidine supplement. The science is real, the people doing it are credible, the research is European. And the gap between the DACH retail framing and the actual peer-reviewed evidence (above all the SmartAge null result) is exactly what this guide tries to close.

Where the autophagy research meets actual medicine, the picture gets mixed. Strong in some diseases, context-dependent in others, and oversold in retail wellness copy.

Brain degeneration. A 2017 Neuron review found that autophagy protects against the kind of brain decline driven by protein build-up, which covers Alzheimer's, Huntington's, Parkinson's, and motor neuron disease. ^[47] The hope is to help the cell clear out the misfolded protein clumps faster. The honest counterweight: no autophagy-boosting therapy has yet passed a phase 3 trial in any major brain disease.

Cancer: the double-edged sword. A 2008 Cell review was the paper that framed this. ^[11] Autophagy cuts both ways in cancer. Early on it suppresses tumours from forming (a partial loss of the Beclin-1 gene turns up in 40-75 % of sporadic breast and ovarian cancers ^[15]). But once a tumour exists, autophagy helps it survive chemo. So the practical rule is blunt: if you have active or recently treated cancer, do not start spermidine or any strong autophagy inducer without your oncologist on board. The spermidine guide carries the same warning, and it is the single most important safety point in the whole autophagy story.

A second, separate recycling pathway. A 2014 Cell Research review covers chaperone-mediated autophagy (often shortened to CMA), a more selective cleanup route that is distinct from the main one and declines with age on its own timeline. ^[48] It is tied to certain inherited storage diseases, to Parkinson's, and to metabolic syndrome. The takeaway: when research says "autophagy," it usually means the main pathway; CMA is a related but separate route with its own age-related decline.

Heart and vessels. Mostly covered by the 2016 Nature Medicine paper ^[33] and the Bruneck cohort signal ^[34] from the spermidine section. The deeper cardiac-autophagy reviews are good for going further but not essential here.

Inherited storage diseases. Pompe disease, Danon disease, and certain inherited mitochondrial diseases are the textbook examples of what happens when the lysosome itself fails and autophagy stalls with it. This is clinical genetics, not lifestyle wellness. But it is a useful counterpoint to the retail claim that "more autophagy is always better."

Infection defense. A specialized form of autophagy targets bugs that hide inside your cells, and it is part of how your body fights tuberculosis bacteria, Salmonella, and several viruses. This is more basic biology than longevity therapy. But it explains why autophagy genes are conserved across evolution and broadly essential.

The honest synthesis. Autophagy biology is essential and does many jobs at once. A few conditions (inherited storage diseases, inherited mitochondrial disease) are genuine treatment targets. Others (preventing brain degeneration, protecting the heart) are mechanistically plausible but have not produced positive phase 3 trials in healthy adults. And in cancer, the simple "more autophagy = better" framing is just wrong: the stage and the tumour type decide.

DACH wellness marketing has built a whole business category around autophagy. Some of it has defensible physiology underneath. Most of it does not have the trial evidence the retail claims imply.

Spermidine on the shelf. Covered in detail in the spermidine guide. The EU Novel Food authorisation under Implementing Regulation (EU) 2017/2470 (the Union list), amended by Commission Implementing Regulation (EU) 2020/443 of 25 March 2020, covers "spermidine-rich wheat germ extract from Triticum aestivum" up to a daily intake equal to 6 mg of spermidine in food supplements for adults, not for pregnant or breastfeeding women. ^[51] Longevity Labs+ (Graz, AT) holds the original authorisation. Synthetic spermidine is not authorised. Typical DACH products: spermidineLIFE, Doppelherz Spermidin, Sunday Natural Spermidin, Sanct Bernhard Spermidin, Green Naturals Spermidin, spermidinPLUS. You will usually pay 30-70 € a month. And the largest cognition trial (SmartAge) came up empty on its main goal. Keep that in mind when you read the retail copy.

"Autophagie Kur" and "Autophagie Fasten": the fasting clinics. Buchinger-Wilhelmi (Überlingen am Bodensee, Marbella) and other clinics in the Buchinger tradition use autophagy language in their educational material. The general pitch: therapeutic fasting (often 7-14+ days on water and broth) switches on autophagy. Most of their clinical claims stay hedged ("supports autophagy," "fördert die zelluläre Erneuerung") rather than promising to prevent disease, which keeps them out of the strict disease-claim rules under Reg (EC) 1924/2006.

Here is the honest read. Buchinger Wilhelmi's underlying physiology (that fasts of 5+ days do something measurable to human autophagy) is more defensible than the popular 16:8 claim. The 2014 muscle-biopsy data ^[49] and the broader long-fast literature (a 2019 PLOS One series of over 1,400 patients on 4-21-day fasting protocols) do show measurable physiology. Direct biopsy time series that validate autophagy flux in Buchinger-style groups are still thin in the peer-reviewed literature, but the broader physiological case is real. The 16-hour slogan is mouse extrapolation. The 7-day Buchinger fast at least sits on top of a small clean human muscle-biopsy paper. That is not an endorsement of fasting clinics. It is honest grading of the evidence underneath.

Wellness apheresis ("Blutwäsche" or "Blutreinigung"). Some private DACH clinics (Apherese Frankfurt, Vital Path Germany, for example) market therapeutic apheresis as a longevity therapy, and the copy sometimes reaches for "Zellerneuerung" or autophagy-adjacent language. There is no peer-reviewed evidence that therapeutic apheresis switches on autophagy in any tissue, and no trial supporting apheresis as a longevity intervention. Apheresis does have legitimate medical uses (familial hypercholesterolaemia, certain autoimmune disorders, ANCA vasculitis) at €1,500-3,000 a session. But the "longevity / autophagy" angle is marketing, not evidence.

Rapamycin off-label. A small but growing private-clinic category in DE/AT/CH. The Rapamycin in Germany guide covers it in detail. Sirolimus is licensed only to suppress the immune system after a transplant. Longevity prescriptions happen on a Privatrezept, your insurer does not reimburse them, and HWG advertising limits apply. Dosing varies a lot between providers (usually 4-6 mg weekly, with or without blood-level monitoring). The evidence base is the Mannick rapalog trials plus the mouse ITP data: both real, neither a phase 3 longevity trial.

"Autophagie-Komplex" combo supplements. Several DACH vendors sell combination products (spermidine plus resveratrol plus green-tea EGCG plus niacinamide) as an "Autophagie-Komplex." None of these claims are EFSA-authorised (see the next section). The softer wording "supports cellular renewal" or "fördert die Zellerneuerung" is the regulatorily safer alternative to claiming the product induces autophagy. The actual evidence for these combinations, beyond what each ingredient does on its own, is essentially zero. The same evidence-grading runs through our wider longevity supplements guide.

No insurer reimburses any autophagy intervention. Not under GKV, ÖGK, or LAMal. Not in the wellness baskets at TK, Barmer, DAK, AOK, ÖGK, or LAMal as of 2026.

Short answer: no. EFSA's Panel on Dietetic Products, Nutrition and Allergies (NDA) has not authorised a single health claim that uses the word "autophagy" under Article 13(1) or Article 14 of Regulation (EC) No 1924/2006 on nutrition and health claims made on foods.

The closest EFSA ever came is a 2011 NDA Panel opinion in EFSA Journal 9(12):2466, on a claim that spermidine prolongs the growing phase of the hair cycle, assessed under Article 13(5) of Regulation (EC) No 1924/2006. ^[50] The record shows it was not authorised: no positive opinion appears in the Union list of authorised claims.

So what does that mean for you in practice?

1. EU supplement labels and DACH marketing copy may not legally say "induces autophagy," "promotes autophagy," or "supports autophagy" as a formal health claim under Article 13(1) or Article 14. Where that language shows up in DACH retail anyway, it is either non-compliant or dressed up in softer "cellular renewal" / "Zellerneuerung" wording that dodges the strict claim rules.
2. Spermidine products are cleared as an ingredient under the Novel Food rules (Reg (EU) 2015/2283, listed via IR (EU) 2017/2470, amended by IR (EU) 2020/443) ^[51]. But no autophagy-specific health claim has been authorised for the substance itself.
3. To approve an Article 13(1) claim, EFSA has always wanted a measurable physiological result in humans. Autophagy markers are currently treated as a mechanism, not an approved endpoint. That lines up with the 2021 consensus guidelines, which say blood LC3 and p62 are not reliable stand-ins for what is happening in your tissues ^[7].
4. Switzerland sits outside the EU, but Swiss food law (BLV / LMG / LIV) tracks the EU rules closely. The practical effect on labels is identical: no authorised autophagy claims, the same wheat germ extract products sold under softer language.

Why this shapes how the guide is written. Throughout, we tell you what trials measured and what cohorts were linked to. We deliberately do not phrase findings like benefit claims on a label, because the regulatory reality is plain: food supplement, no authorised autophagy claim, no insurer reimbursement, no medical-device-grade quality control. The science is real. The human evidence is partial. And no EU-authorised health claim exists.

Is autophagy real biology? Yes, no question. The 2016 Nobel Prize to Ohsumi ^[1], the 2023 hallmarks framework (broken-down autophagy listed as one of the 12 hallmarks of aging) ^[9], and the 2018 Nature Reviews paper (autophagy is the common thread through every validated lifespan intervention in lab animals) ^[8] are not the fringe of the field. They are the consensus.

Does "16 hours of fasting trigger autophagy" in humans? The cleanest peer-reviewed muscle-biopsy data (n=8 healthy men) shows the markers move at 72 hours, not 16 ^[49]. No peer-reviewed human trial shows clean autophagy turning on at 16 hours. The 16-hour number is mostly mouse extrapolation. And the TREAT trial (n=116) found that 12 weeks of 16:8 did nothing for weight or metabolism beyond any calorie change it happened to cause ^[24]. The popular DACH wellness claim is wrong.

Does spermidine extend human lifespan or prevent dementia? Mechanism: yes. Mouse lifespan: yes, and it depends on autophagy ^[33]. Bruneck cohort mortality: HR 0.74 per one-standard-deviation more dietary spermidine ^[34]. But the SmartAge trial (n=100, 12 months) for cognition was NEGATIVE ^[38]. That is the largest, longest human trial, and it found no benefit. Do not soft-pedal that. The strongest human signals are (a) the Bruneck observational link to cardiovascular mortality, and (b) the 2024 paper that recasts spermidine as something fasting produces ^[39]. Getting spermidine from wheat germ, soy, mature cheese, mushrooms, and pulses is reasonable. Buying a 50 EUR/month supplement is not backed by trials for hard outcomes.

Does rapamycin work? In mice, yes (+14 % / +9 % measured at the age by which 90 % had died; mean lifespan up about +13 % / +9 %) ^[29]. In humans, the rapalog trials on immune aging are credible signals ^{[30, 31]}. A phase 3 longevity trial does not exist. In Germany the drug is prescription-only for transplant patients; off-label longevity use is rare, Privatrezept-only, and not standardised.

Does resveratrol work? A defensible mechanism in cell models (through that cellular fuel molecule and its partner enzyme) ^[16]. Negative in the NIA ITP mouse lifespan trials on a standard diet. No positive human trial for hard outcomes. Treat it with the same skepticism the CoQ10 guide aims at the ubiquinol absorption claim.

What about exercise? The cleanest causal evidence in mammals (BCL2 AAA mice: exercise's metabolic protection specifically needs working muscle autophagy) ^[46]. The strongest human survival link (cardio fitness, see the Zone 2 guide). And the one thing nobody can sell you in a bottle.

Wellness apheresis, boutique "Autophagie Kur" packages, "Autophagie-Komplex" combo supplements? No EU-authorised autophagy health claim exists. None of these products has phase 3 trial evidence for switching on autophagy in tissue. Some are harmless and expensive. Some are not harmless. None are reimbursed by GKV, ÖGK, or LAMal.

The defensible practical takeaway. The lifestyle that mimics autophagy-friendly physiology is the same one that mimics calorie restriction and exercise. So eat fewer calories overall. Eat less animal protein in your 50s and 60s, leaning plant-side. Exercise (it has its own validated autophagy mechanism). Eat plenty of wheat germ, pulses, mature cheese, and mushrooms for dietary spermidine. Do not expect a magic 16-hour threshold or a 50 EUR supplement to do anything special. And if you want to try a longer fast (3-5 days), do it under medical supervision if you have any chronic condition, and keep in mind that the human evidence at long fast durations still rests on small biopsy groups.

The biology is real. The mouse data is overwhelming. The human translation is partial. Do not let DACH wellness marketing tell you otherwise.