Measure it, then move it
Jan Gruber is an aging biologist at the National University of Singapore, in the same NUS longevity group as Brian Kennedy. He works on two of the field’s hardest problems at once: how to measure aging, and how to slow it.
On the measuring side, he builds aging clocks — and he is betting they are about to get much better. As the datasets grow, he expects mortality-based, multi-omics clocks to converge, so the different clocks finally give you the same answer. On the moving side, he screens geroprotector drugs in the nematode worm C. elegans, using a clever new readout — the “transcriptional noise” that rises with age — to test hundreds of interventions and score them for synergy before the slow work of lifespan studies.
His method is deliberately staged: reduce the noise in worms first, confirm lifespan and healthspan, then step up to flies and mice. It pairs naturally with the biology of his friend and collaborator Vincenzo Sorrentino — Jan’s screening platform, Vincenzo’s mitochondria and protein-folding questions.
He is also one of the most quietly optimistic people we met. He likes to imagine his own daughter’s generation looking back in disbelief: that there was once a world when people thought aging was completely immutable — a bit, he says, like the world before antibiotics.
The thing to know first is that he is a physicist. That is what he trained as, and it is why he is in this field at all: ageing clocks are classifiers built with machine learning, statistics and linear algebra, and for most of the history of the subject there were no data deep enough to build them on. Now there are. “As a physicist,” he says, “it’s easy to make a contribution in that context.”
He is exacting about what a clock is for. The clocks that will matter, he argues, are the ones that predict mortality — GrimAge, PhenoAge, and the clinical clocks built from blood proteins, lipids and signalling molecules, where the link to future death is already well evidenced. A clock that moves a marker with no bearing on whether you live is not measuring anything worth having. And what it gives you is not a date. It is a probability: your chance of getting through the next decade in good health, and whether an intervention changed it.
The first clock his lab ever built was a lipid clock in nematode worms, using mathematics that came originally from Peter Fedichev, to prove the method worked at all. Then they applied the same tools to a clinical dataset. That is the point he keeps making: the same mathematics runs the whole way up the ladder of biological complexity — worms, flies, mice, people — so you can develop a tool in an animal that lives three weeks and end up using it on a person who lives ninety years.
Ask him whether ageing can really be slowed and he answers with a number rather than a hope: we have been adding about three months of life expectancy a year, for nearly a century. And society has quietly tracked it. Actors in their eighties carry films now; a hundred years ago nobody would have staked a four-year production on one. What has not happened yet, he says, is the interesting part — all of that came from hygiene, nutrition, and treating disease. “I don’t think we have manipulated fundamental biology of aging yet. But that’s the change. That’s the program now.”
He would rather prevent than replace. Given the choice between printing a new kidney and understanding at forty what will fail at seventy, he takes the second every time — though he adds, with the timing of someone who has made the joke before, that if he had kidney failure he would quite like a new one. And he does not only prescribe it. He has taken metformin for a long time, keeps his cholesterol very low, and is openly interested in rapamycin — one of the very few compounds, he says, that works reliably every time he tries it, when many beautifully published ones do not.
