Blood tests of body age ticked slower on a weight loss drug, small study finds

Everyone has two ages. One is counted in birthdays. The other is a guess at how worn the body seems, and some lab tests now try to read it from a single blood sample [1].

In a small trial, adults given a weight loss drug scored younger on several of these tests than adults given a dummy drug with no medicine in it. On one test, the chemical marks of aging were building up about 9 percent more slowly in the drug group [1].

The main caveat matters for everything that follows. The question about aging was asked only after the trial had ended, using blood that had been frozen along the way. The trial itself was built to study body fat, not aging [1].

The work was led by Michael Corley of the University of California San Diego and published in Nature Communications in May 2026. The PubMed record lists grants from the US National Institutes of Health [1][2].

The drug was semaglutide, best known for weight loss. The trial had 84 adults living with HIV, a virus that attacks the body's defences against germs. All of them also had fat building up in parts of the body where it does not usually gather. Chance decided, like a coin flip, that 45 got the drug and 39 got the dummy drug, for 32 weeks [1].

How can a blood test guess an age?

Each cell holds DNA, a long set of instructions for the body. Tiny chemical marks sit on the DNA and help decide which instructions are switched on. The pattern of those marks shifts as people get older. Computer programs called epigenetic clocks are trained to read the pattern and give back a number, a guess at the body's age [1].

One clock, called DunedinPACE, works a little differently. It does not say how old the body seems. It says how fast the marks are building up, more like a speedometer than a calendar. That is the clock that read about 9 percent slower in the drug group [1].

Other clocks pointed the same way. On a clock called PhenoAge, the paper reports a gap of 4.9 years per year in favour of the drug group, and 3.1 on one called PCGrimAge. Those are rates worked out from 32 weeks of data, not five years measured in anyone. The gaps held up when the team allowed for sex, body weight and swelling in the body [1].

What matters more than any single number is that every clock moved in the same direction. Even so, the authors spell out the limits in their own words:

The post hoc design, modest sample size, HIV-specific cohort, and 32-week follow-up limit generalizability.

In plain words, the question was asked after the fact, the group was small, everyone in it was living with HIV, and the trial was short. So the result cannot simply be applied to other people. A new look at saved samples is where ideas start, not where they get proven [1].

There is also a gap between a clock score and real health. A clock score is a number from a lab. It is not a record of illness, of how someone feels or of years lived. Researchers still do not know how closely these scores track anything a person would notice [1].

Some news stories said a weight loss drug slows aging. That is something this study did not test [4]. Elsewhere, a team at the University of Colorado Anschutz Cancer Center is asking whether drugs of this kind act on immune cells directly. That work is still only lab tests on cells, with no results in people yet [3].

The authors say what would come next:

Prospective trials are needed to determine whether GLP-1 receptor agonists can be repurposed as gerotherapeutics.

In plain words, a new trial would need to name body age as its goal before it starts, rather than looking back afterwards. Gerotherapeutics means medicines aimed at aging itself. GLP-1 receptor agonists are drugs that switch on the GLP-1 receptor, a kind of switch on a cell, and this drug is one of them [1].

This study does not show that anyone lived longer, felt better or aged more slowly in any way a person would notice. What it does show is narrower. In a small trial of people living with HIV, blood from the drug group scored younger on several aging clocks than blood from the dummy group [1].