Your body eats its own damaged cells. Being overfed switches it off.

Autophagy is real, and worth caring about. The part the supplement aisle skips is that the main switch is behind you at the dinner table, not in a bottle.

A decade ago, autophagy was a word you only met in a cell-biology lecture. The 2016 Nobel Prize in Medicine went to Yoshinori Ohsumi for working out how it happens. Since then it has become a wellness buzzword and a product category: capsules and protocols that promise to boost your autophagy, switch on cellular cleanup, or sell you autophagy in a bottle. The biology is genuinely interesting. Most of what is sold on the back of it is not what the science says.

Autophagy, literally self-eating, is the process your cells use to break down and recycle their own worn-out parts: damaged proteins, spent organelles, the molecular debris that builds up with age and use. It is quality control and recycling in one. When it runs well, cells stay cleaner and work better. When it falters, the debris accumulates, which is part of why it matters for ageing and the diseases that come with it.

Here is the part the boost it marketing quietly inverts. Autophagy is, at its core, a response to scarcity. When nutrients are abundant, a cellular sensor called mTOR is switched on, growth is prioritised, and autophagy is turned down. When nutrients are scarce, mTOR quietens and the cell falls back on recycling what it already has. In plain terms: eating, especially eating a lot and often, is the off-switch. Going without food for a while, and to some extent exercising, is the on-switch.

Which is where the honesty has to start, because there is a real problem underneath all of this: autophagy is very hard to measure in a living person. Most of what we know comes from yeast, worms, mice, and cells in a dish, where you can watch it directly. In people you are mostly reading indirect markers or inferring from what works in animals. So when anyone, a supplement brand or a fasting app, tells you they have switched your autophagy on, treat the confidence as the tell. Nobody is actually measuring it.

With that caveat, the levers with the strongest case are behavioural and free. Not eating around the clock, whether that means a longer overnight gap or a proper period of fasting, is the most direct way to take your foot off the mTOR pedal. Exercise pushes in the same direction. And for anyone carrying excess weight, losing it lifts the chronic nutrient overload that keeps the off-switch pressed down. None of these is a clean, measured dose of autophagy, but they act on the actual switch, which is more than the alternatives can say.

So what about the compounds? Two have real human data, and both are more interesting and more modest than the marketing. The first is spermidine, found in foods like wheat germ, aged cheese, and soy, which triggers autophagy in animals. In people, the headline finding is an association: in a long-running Austrian study, those whose diets were richest in spermidine had lower death rates over twenty years, a gap roughly equivalent to being about 5.7 years younger (Kiechl 2018). That is a real and striking signal. It is also dietary intake, not a supplement, and an association: people who eat more spermidine-rich food differ in many other ways. When spermidine was actually put in a capsule and tested, a small pilot in older adults with memory complaints looked promising (Wirth 2018), but the larger, longer trial built to confirm it found no benefit to memory over a year (Schwarz 2022). Promising pilot, negative confirmation, is one of the most common and least reported stories in this field.

The second is urolithin A, a compound your gut bacteria make from the ellagitannins in pomegranates, walnuts, and berries, which triggers mitophagy, the specific recycling of worn-out mitochondria. It has been through actual randomised trials, which is more than most longevity supplements can claim. The honest reading of them is mixed: in older and middle-aged adults, urolithin A was safe and improved some measures of muscle endurance and strength and lowered some inflammation markers, but in both trials it missed its main pre-specified endpoint (D'Amico 2022; Singh 2022). And both trials were run by the company that sells it. A real signal, a modest size, and a conflict of interest to keep in view.

Put it together. Autophagy is a real process worth caring about. The switch that controls it responds mainly to how much and how often you eat, and to whether you move, which is why the strongest levers are behavioural and cost nothing. The supplements with genuine human data, spermidine and urolithin A, are downstream adjuncts: interesting, mostly modest, sometimes failing their own confirmatory trials, and loudest where a company has something to sell. There is no autophagy in a capsule, because the capsule is not where the switch is.

If cellular housekeeping is what you are after, the unglamorous version is also the honest one: leave longer gaps between meals, move your body, keep your weight in a healthy range, and eat the actual foods, wheat germ, legumes, aged cheese, nuts, that carry these compounds in the first place. A supplement might turn out to add something at the margin, but not the job the marketing implies.

The cell's recycling system was built to run on going without, not on buying more. The most honest thing the science says about switching on your autophagy is that the main switch has always been how, and how often, you eat.