The Longevity Industry and Its Favourite Molecules

The reception is in a converted townhouse, with pale oak, a single orchid and a card menu propped on the counter. It lists consultations, blood panels, an infrared room and, near the bottom, a short column of peptides, each with a line of copy beneath it. The copy is scrupulous: support cellular renewal, optimise mitochondrial function, promote healthy ageing. Nothing on the card says any of these things will make anyone live longer, and nothing on it needs to. The word longevity is on the door.

What is the evidence for the peptides the longevity industry sells? For the claim the industry is named after, extending human lifespan or the healthy years within it, there is no controlled human evidence for any of them. For most, the evidence consists of cell cultures and rodents, sometimes from a single institution. One molecule that entered the industry's vocabulary has been approved as a medicine, but for a rare inherited disease rather than for ageing. This piece is a critique of a way of talking, not a catalogue of compounds. It asks why the language is so careful while the implied claims are not, and where each favourite molecule's evidence actually stops.

An industry that grew out of one small study

Modern anti-ageing medicine can be traced, more precisely than most industries, to a single journal issue. In 1990 a group led by Daniel Rudman gave a small number of men over sixty, whose levels of a growth-related marker were low, six months of human growth hormone and compared them with untreated men. The treated men gained lean body mass, lost fat and showed thicker skin 2. It was a short, small, careful study with a modest conclusion. Within a few years it had become a founding citation for a clinical movement promising to reverse ageing.

The longer-term evidence did not follow the movement. A systematic review published in 2007 gathered the randomised trials of growth hormone in healthy older people. It found small changes in body composition, a little more lean tissue and a little less fat, but no improvement in strength or aerobic capacity. It also found clearly higher rates of adverse events, among them soft-tissue swelling, joint pain, carpal tunnel syndrome and breast enlargement in men, with a tendency towards impaired glucose handling 3. The reviewers concluded that the hormone could not be recommended as an anti-ageing therapy. By then the industry no longer depended on it. It had learned to sell the idea, and the molecules could be swapped.

In 2002, fifty-one scientists who study ageing signed a position statement aimed squarely at the industry. No intervention then on the market, they wrote, had been shown to slow, stop or reverse human ageing, some of the products might be harmful, and those selling them often misrepresented the science on which they rested 1. Nearly a quarter of a century later that sentence has not needed revising. What has changed is the list of molecules it applies to.

Why ageing is not an indication

There is a structural reason the evidence stays thin, and it is worth understanding before blaming anyone. Medicines are approved for indications, meaning specific diseases or conditions, and regulators do not treat ageing itself as one. A company cannot run a trial showing its product treats ageing, because there is no agreed endpoint that would count, and it cannot wait for participants to die, because lifespan trials in humans would take decades and very large populations.

Researchers who take the problem seriously have had to design around it. The proposed trial of metformin in ageing, set out in 2016, was built to persuade regulators that a composite endpoint could stand in for ageing itself: delaying the onset of a cluster of age-related diseases such as cardiovascular disease, cancer and dementia 4. Even that, for one of the cheapest and best-understood drugs in medicine, required years of negotiation and a large, costly trial. The longevity peptides have nothing comparable.

The consequence for the industry is subtle. Because no product can be approved for ageing, none can be advertised as treating it. The whole commercial field therefore operates in a zone where approval is impossible by definition, and its claims are shaped to stop just short of the medical. That is why the menu says support and optimise: those are the words available to someone forbidden from saying treat.

The hallmark as a sales pitch

The most intellectually respectable piece of the industry's vocabulary comes from a genuine scientific framework. In 2013, and in an expanded version in 2023, a group of researchers proposed a set of hallmarks of ageing, processes that change with age and seem to contribute to it. They include telomere shortening, mitochondrial dysfunction, loss of protein maintenance, cellular senescence and chronic low-grade inflammation, among others 5. The framework was meant to organise research, and it does that well.

It also, unintentionally, gave marketing a template. Take a peptide, find a cell-culture or rodent paper in which it nudges one of the hallmarks, and a sentence writes itself: this molecule targets a hallmark of ageing. The sentence is often technically true and almost entirely uninformative, because touching a mechanism in a dish is a very long way from changing the course of ageing in a person. The framework is a map of processes that change with age 5. It does not certify that nudging any one of them, in any organism, changes how a person ages. The industry reads it as a shopping list.

The favourite molecules, and where their evidence stops

The same handful of names recurs on the menus, and each carries a story. Epitalon, a four-residue peptide developed at an institute in St Petersburg, is linked to telomeres through a 2004 paper reporting that it induced telomerase activity and telomere elongation in cultured human fetal fibroblasts 6. MOTS-c, a short peptide encoded in mitochondrial DNA and described in 2015, improved metabolic measures in mice fed a high-fat diet 7. A copper-binding tripeptide found in plasma has a long cell-culture and cosmetics literature. Growth hormone secretagogues inherit the story of the 1990 study, and with it the 2007 review's verdict 23.

The last row is the instructive one. Elamipretide, a peptide that binds a lipid in the inner mitochondrial membrane, received accelerated approval in the United States in September 2025 for Barth syndrome, an ultra-rare inherited mitochondrial disease that affects mainly boys and young men 8. That was a real regulatory achievement, years in the making. It is also exactly the opposite of what the longevity menu implies. The molecule was approved because it was tested in a defined disease with defined endpoints, and approval for one condition says nothing about healthy people growing older. Approval came for the rare disease, not for the general promise.

Careful words, careless claims

Go back to the menu card. Each line on it is written to survive scrutiny, and read in isolation almost every line would. Support is not a claim of effect. Optimise has no measurable endpoint. Healthy ageing is something everyone wants and nobody can disprove. The discipline is the same one cosmetic copy observes, for the same reason: the precise sentence would make a medicinal claim, and a medicinal claim would require a licence.

But a claim does not have to be written to be made. The room makes it: the name on the door, the blood panels that precede the peptides, the price, and the fact that the peptides appear on a longevity menu rather than a dermatology one. Each sentence is careful and the whole is not. A customer is invited to assemble the promise from compliant parts, and the parts are designed so that nobody has to put them together on paper.

What a real answer would need

None of this means ageing research is a sham, or that no peptide will ever matter to it. The field is serious and growing, some of its molecules are genuinely interesting, and at least one has already become a medicine by the slow, honest route. The critique is narrower. It concerns an industry that sells the conclusion of research that has not been done.

What would a real answer require? A defined population, a validated endpoint or an agreed composite of age-related disease, a randomised comparison, independent replication, and years. That is where a piece about an industry has to stop, and where trial design begins: choosing endpoints, sizing samples and defining what counts as delay. It is a long, expensive discipline that nobody on the townhouse's pale oak counter is proposing to pay for, which may be the most telling fact on the card.