Overview
Daniel Belsky, who leads ARPA-H's FAST project with longevity pioneer Nir Barzilai, says he's optimistic, and it could happen even sooner.
Daniel Belsky has spent a dozen years trying to answer a deceptively simple question: How fast is a human being aging?
“You can see an animal age during your postdoc,” said Belsky, an associate professor of epidemiology at the Columbia University Mailman School of Public Health in the Robert N. Butler Columbia Aging Center. “But if I want to run a study of human aging, it's going to take me my entire career, and I'm going to die before I know what the answer is.”
That creates a peculiar bottleneck for longevity biotech, which is moving from studies in animals toward increasingly more ambitious human-led experiments.
Researchers can track thousands of biological markers in people receiving an experimental drug, yet the field still lacks a validated metric that shows whether a drug is actually slowing aging.
Now, the effort to solve that problem has significant backing. In February, the Advanced Research Projects Agency for Health awarded Belsky up to $10.5 million through its PROSPR program to fund FAST, an initiative he co-leads with Nir Barzilai, a professor of medicine and genetics at Albert Einstein College of Medicine and director of its Institute for Aging Research, and Mahdi Moqri, a Harvard professor and co-director of the Biomarkers of Aging Consortium.
FAST, short for Finding Aging Biomarkers by Searching Existing Trials, is mining completed placebo-controlled clinical trials for molecular signals that change across multiple interventions and can be tied to the biology of aging and meaningful clinical outcomes.
“We're asking two things: What is the best biomarker for aging and also changing within a year of treatment?” Barzilai said.
The Trouble With Today’s Aging Clocks
There’s already no shortage of ways to measure biological age. Researchers have built dozens of them, from epigenetic clocks such as GrimAge and PhenoAge to DunedinPACE, a measure of the pace of aging that Belsky co-developed. Several companies now sell tests built on them directly to consumers.
The problem is that researchers can get very different answers depending on which test they use.
Barzilai says he’s taken about 12 existing biological-age tests. The results have placed him anywhere from 20 years younger than his chronological age to four years older.
He wants something more useful: a test that could show whether an intervention has changed a healthy person’s biological aging, much as a cholesterol test can show whether an intervention has changed cholesterol.
The bigger challenge is proving that the change means something clinically.
“We don't have, what the Food and Drug Administration would call a surrogate endpoint for healthy aging,” Belsky said.
Cardiology has relied on one for decades. A company developing a cholesterol drug doesn't have to wait years to see whether it prevents heart attacks. It can show the drug lowers LDL cholesterol, which the FDA accepts as a stand-in. Aging has no equivalent, so a longevity trial would have to wait for the outcomes themselves.
Establishing a surrogate endpoint takes years of work and many steps. Among them, Belsky says, researchers need to connect three pieces of evidence: a change in the biomarker, a change in the clinical outcome, and a demonstration that the biomarker change helps explain the outcome.
“The critical missing link is studies that actually observe the process from intervention to an approved endpoint,” Belsky said. “So you need all three of those things. That's what FAST is designed to achieve.”
That missing link matters enormously for the companies trying to fund and develop drugs around aging biology, because a longevity company can have a drug, a mechanism, and human data, yet still struggle to answer one of the questions investors care about most: What exactly are we trying to prove?
This is one of the first questions Sergey Jakimov, a managing partner and co-founder of investment firm LongeVC, asks biotech companies that come to him claiming they’ve changed an aging biomarker:
“Which FDA process are you going through and, in that, what's your indication? Because aging is not an indication,” he said. “So what are you trying to fix within aging?”
The Answer May Already Be Sitting in Finished Trials
This is where FAST gets interesting.
Instead of starting with a new aging clock, the team is going back to completed clinical trials and looking at what happens across a range of interventions, including GLP-1s, SGLT2 inhibitors, metformin, rapamycin, caloric restriction, and exercise.
They’re looking across proteins, metabolites, and DNA methylation sites, searching for changes that keep showing up across different interventions, whether it’s a drug, a dietary change, or exercise. Belsky said those recurring signals may be more likely to reflect the biology of aging itself.
The scale of the project is enormous: roughly 11,000 proteins on one platform, 5,000 on another, about 1,000 metabolites, and close to 1 million DNA methylation sites.
“It’s a proving ground for surrogate endpoints that will open the doors to a whole new generation of geroscience clinical trials,” Belsky said.
Barzilai and Belsky said they expect to deliver first results to ARPA-H by August 2027. Those findings will then feed into work led by Michael Snyder at Stanford University to develop a test kit for future aging clinical trials, Belsky added.
Belsky is cautious about predicting what will emerge. Still, he thinks the field may be surprisingly close to a breakthrough in measuring and validating human aging.
“I am optimistic that we are within 5 years of this kind of a breakthrough, and it could come much sooner than that,” he said.
That does not mean longevity biotech suddenly gets an FDA-approved path to an aging drug. It means researchers may finally be approaching the point where they can rigorously test the idea.
Could Aging Become an FDA Indication?
If FAST succeeds, the implications could extend beyond finding a useful biomarker.
“What will be an interesting outcome though is: Does aging become a metric?” said Artem Trotsyuk, an operating partner at LongeVC.
The bigger question, Trotsyuk said, is what happens after that: “If it becomes a thing, does it become an indication to the FDA? That’ll be interesting.”
Jakimov sees another potential effect: capital. If researchers identify biomarkers that reliably respond to interventions and correlate with clinical outcomes, he said, “there will be an even greater spotlight on longevity biotech and more influx of capital,” helping companies and funds accelerate therapeutic development.
