For most of medicine’s history, a person’s age was one number: the years since birth. Over the past two decades, researchers have learned to measure ageing inside the body, and what they have found is uneven. They have named twelve processes that wear a body down, including damage to DNA, chemical tags drifting out of place on the genes, energy-making parts of cells failing, and worn-out cells that stay alive and inflame their neighbours. From proteins in a blood sample, they can now estimate the age of one organ at a time, because some of those proteins come mainly from a single organ. In a study of nearly 5,700 adults, almost one in five had a single organ running well ahead of the rest, and the fast organ pointed to the illness that person was more likely to develop: a quickly ageing heart to heart failure, a brain to Alzheimer’s.
The frontier drugs
Once a process can be named and measured, drug developers can aim a medicine at it. Four medicines at the front of the field each act on one. Rapamycin, a transplant drug, blocks a switch cells use to decide, according to how much food is available, whether to grow or conserve. Semaglutide, a GLP-1 drug sold for diabetes and weight loss, lowers inflammation and shrinks deep abdominal fat. Rentosertib quietens worn-out cells. NMN, a supplement, raises a molecule cells use to turn food into energy, which falls as people age. Three of the four were developed for other illnesses, and none has yet been shown in a large clinical trial to slow ageing.
The newest result shows why that is hard to establish. In September 2026, researchers reported a trial of rentosertib, a drug built for a scarring lung disease, in 42 patients with that disease. In the group on the highest daily dose, predicted age fell by about three years within four weeks. That figure came from ageing clocks, computer models built from the blood of thousands of people that read a new sample and return an age, and what a clock’s number means depends on what it was taught to do.
Two kinds of clock
Some clocks were taught to recognise age: shown a blood sample, they estimate how old the donor is, so a lower score means only that the blood looks like a younger person’s. Others were taught to recognise risk of illness and early death. Researchers followed thousands of people for years, recorded who fell ill or died early, and trained the model on the blood patterns that came before, so a lower score means the blood looks like that of people who stayed well for longer. In the rentosertib trial, the four age clocks said the patients looked about three years younger, while the two risk clocks showed no real change. Their blood looked younger without looking any less at risk. The trial’s authors also say they cannot separate slower ageing from a lung that was being treated. A Yale review of 16 clocks across 51 studies of treatments meant to slow ageing, published in August 2026, found that risk clocks respond dependably to real change in the body, and age clocks move in no clear direction.
{{/usCountry}}Some clocks were taught to recognise age: shown a blood sample, they estimate how old the donor is, so a lower score means only that the blood looks like a younger person’s. Others were taught to recognise risk of illness and early death. Researchers followed thousands of people for years, recorded who fell ill or died early, and trained the model on the blood patterns that came before, so a lower score means the blood looks like that of people who stayed well for longer. In the rentosertib trial, the four age clocks said the patients looked about three years younger, while the two risk clocks showed no real change. Their blood looked younger without looking any less at risk. The trial’s authors also say they cannot separate slower ageing from a lung that was being treated. A Yale review of 16 clocks across 51 studies of treatments meant to slow ageing, published in August 2026, found that risk clocks respond dependably to real change in the body, and age clocks move in no clear direction.
{{/usCountry}}“The clocks are misleading unless they are deployed in the context of a specific physiological condition. If all clocks show different time, then the clocks are wrong, not the time,” says Deepak Kumar Saini, a professor at the Indian Institute of Science who coordinates BHARAT, an Indian study of how people age. A reading, in his view, means something only when it is tied to a specific condition in the body being measured.
What the evidence supports
The interventions with the most evidence behind them are the oldest and cheapest. In healthy older adults, four things have slowed the clocks in trials that ran for two to three years: a shingles vaccination, omega-3 and vitamin D, exercise, and a daily multivitamin. Each moved them by months. Some of the newer drugs carry risk. Rapamycin suppresses the immune system to protect a transplanted organ; in a healthy person taking it off label, for a purpose it was never approved for, that leaves infections free to take hold. The risk is stark in India, where tuberculosis and other infections circulate far more widely than in Western countries where the idea of taking rapamycin to slow ageing took hold. Quietening a healthy adult’s immune system here raises the chance of infection.
Watching the fast organ
For now, the most useful product of this research is the map itself: knowing which organ in a body is ageing fastest. An organ ageing faster than the rest of the body is a reason to watch that organ more closely, with the checks doctors already use for its diseases, whether or not any drug ever slows the process underneath. His group’s aim, Saini says, is “to identify features which can identify the organ which is undergoing accelerated degeneration, for which appropriate clinical intervention can be deployed”. Whether slowing a clock adds years to a life is a question only long trials can answer, since they have to follow people for years to see who stays well, and several are under way.