I read a piece in MIT Technology Review this week (October 6, 2026) that I wanted to share. It covers data that Novo Nordisk and Eli Lilly presented at the Aging Research & Drug Discovery conference, looking at whether GLP-1 medications change markers of biological aging — not just weight, not just blood sugar, but how old your tissues look on a molecular level. I want to walk through what was actually shown, because the headlines are getting ahead of the data, and the data itself is interesting enough without exaggeration.
What Novo Nordisk and Eli Lilly Actually Presented
Novo Nordisk’s analysis used proteomic clocks — models that estimate biological age from patterns of proteins in the blood. They pooled data from 10,052 people across five placebo-controlled semaglutide trials (SELECT, STEP-HFpEF, STEP-HFpEF DM, FLOW, and SOUL), with follow-up out to 156 weeks. Compared with placebo, semaglutide was associated with a biological age reduction of roughly two to three years across tissues on average, up to four years on a heart-specific protein clock, and reductions of roughly one to four years on clocks tied to the lung, kidney, brain, and pancreas. This is a preprint — it has not yet gone through peer review. What caught my attention most is that Novo’s own authors wrote that the effect was not completely explained by weight loss, blood sugar control, or inflammation. That is a claim worth sitting with: it suggests something beyond the obvious mechanisms might be happening.
Eli Lilly’s data came from a different angle and a much smaller study. Within a substudy of SURMOUNT-5, they used epigenetic clocks — which estimate age from patterns of DNA methylation — in 71 tirzepatide-treated participants over age 50, from baseline to week 72, a span of 1.38 years of actual chronological time. There was no placebo or control group in this substudy. All 15 clocks tested showed less aging than the 1.38-year benchmark would predict, and two of them reached statistical significance, with median gains of 1.1 and 0.7 years.
The Caveats I Want You to Hold Onto
Here is where I have to put my internal medicine hat back on, because the enthusiasm in the lay coverage is running ahead of what these two data sets actually support.
First, Lilly’s own VP of aging research, Kevin Duffin, was notably restrained about his company’s findings. He stated plainly that tirzepatide “works through blood sugar reduction and through weight loss, and the effects we’re seeing are mostly due to that.” In other words, the company that ran this study is not claiming an independent anti-aging mechanism for tirzepatide — they are attributing it largely to the known, well-established benefits of better glycemic control and lower weight. That is a meaningfully different claim than Novo’s, and it matters which one you remember.
Second, the Lilly substudy had no control arm. Without a placebo group, it is hard to separate a true drug effect from the many other things that happen to someone’s health and habits over 72 weeks of being in a structured trial.
Third, both of these were presented at a conference, not published in a peer-reviewed journal. That does not mean they are wrong. It means they have not yet been through the scrutiny that catches errors, and the numbers could shift.
Steve Horvath at UCLA, who is credited with inventing aging clocks, called the two-year effect “pretty strong” and said the data could indicate GLP-1s are acting as geroprotectors — agents that slow or possibly even reverse biological aging. But he also said, explicitly, that he is not yet taking these drugs himself. I find that honesty more useful than almost anything else in the article. The person who built the measurement tool is intrigued, not convinced.
There is also a mouse study worth mentioning for scale, not proof: semaglutide-treated mice lived an average of 834 days versus 742 days for placebo-treated mice. Mice are not people, and a clock reading is not the same thing as a longer, healthier life. These clocks are surrogate markers — useful signals, not endpoints. None of this tells us yet whether a lower biological age reading today means fewer heart attacks, less dementia, or more years lived twenty years from now. That is the honest state of the science.
Part of a Larger, Genuinely Exciting Pattern
What strikes me is that this fits into a broader trend I have been watching for a couple of years now: evidence accumulating that GLP-1 medications do things that are not fully explained by the pounds lost on the scale. We are not at the point of calling these drugs the first true longevity medicine — that is a bigger claim than the current evidence supports, and I am wary of anyone who says it with certainty right now. But we are at the point where serious researchers, at serious companies, using serious measurement tools, are asking the question in public and finding results that warrant a second look. That is genuinely exciting, and I think it is fair to say so, as long as we are honest about what is proven and what is promising.
What I Think the Real Opportunity Is
Here is the part I keep coming back to with my own patients, independent of whatever the aging clocks eventually show. When food noise quiets down — when someone is no longer spending enormous mental energy negotiating with hunger, cravings, and the next meal — something opens up. That mental bandwidth has to go somewhere. I think the real leverage in this moment is using that freed-up capacity deliberately, on the things we already know extend healthspan: consistent sleep, lower chronic stress, regular movement, the quality (not just quantity) of what we eat, the health of our relationships, and an honest look at alcohol or other substances. The medication may or may not be doing something independently interesting at the cellular level. What it is doing, reliably, for many people, is making room. What gets built in that room is still up to the person living in it, one ordinary day at a time.
