Six Aging Clocks Agreed on One Drug, but They Cannot Prove It Slowed Aging

Six independently developed aging clocks looked at blood from a small drug trial and pointed in the same direction: patients who received rentosertib had protein patterns associated with a younger biological age. But the paper warns that the clocks cannot separate aging from changes caused by treating lung disease.

KEY RESULT
At week four, the 60-milligram once-daily group appeared 2.7 to 3.5 years younger on four chronological-age clocks, while the broadest cross-clock agreement appeared with 30 milligrams twice daily.

What the researchers measured

Rentosertib is an experimental inhibitor of a protein called TNIK. Artificial intelligence helped identify the target and design the molecule, which is being developed for idiopathic pulmonary fibrosis, a disease that progressively scars the lungs. The new Nature Biotechnology analysis did not run another clinical trial. It reused serum collected during an earlier randomized trial.

Of the original 71 participants, 42 had longitudinal protein data suitable for this analysis. Researchers measured 2,841 proteins at baseline and after two, four, and 12 weeks, then scored those profiles with six clocks trained by different groups and for different purposes. Treated arms generally moved toward a younger predicted biological age, while placebo changed little or moved slightly older.

Agreement matters more than the headline number

The strongest signal appeared at week four and then plateaued. The 30-milligram twice-daily regimen produced significant changes across the widest range of clocks. The dose with the largest lung-function improvement was different, and changes in forced vital capacity explained little of the variation in predicted age.

That separation is suggestive, but it does not prove an effect on aging. Pulmonary fibrosis changes inflammation and circulating proteins, so improving the disease could make an aging model report a younger score. The authors used pathway analyses to look for changes involving cellular senescence and metabolism, but described those findings as exploratory.

AI appears twice here: in designing the drug and interpreting thousands of proteins. NewTqnia has examined AI filtering weak drug ideas and how exercise changes blood proteins. Such models narrow hypotheses rather than prove clinical benefit.

What would settle the question

A convincing test would enroll a larger, diverse population, define aging outcomes in advance, follow participants longer, and connect clock changes to physical function and disease. Independent replication must also account for lung improvement.

The study’s strongest contribution is methodological: several protein-based clocks can agree inside a disease trial. Whether that agreement reflects slower human aging remains an open question.