6,442 compounds went into the screen. Fourteen came out flagged as pro-longevity. Fourteen more came out flagged for the opposite.
The lab was Albert-László Barabási’s at Northeastern, working with Harvard, published in Nature Aging this year. The method maps 2,358 longevity-associated genes onto the human interactome and measures how near each compound sits to that map, with a transcription-based score called pAGE running alongside. Three hundred and seventy compounds landed near at least one hallmark of aging.
Nobody swallowed anything.
The candidate that travelled furthest through the coverage is oxymetazoline, the active ingredient in Afrin and Sinex, also sold as redness eye drops and as a rosacea cream. The screen placed it near cell-to-cell communication. Northeastern’s own write-up notes it has no currently known impact on longevity.
Grade that honestly before it reaches a supplement label. This is a hypothesis generator: one rung above a hunch, several rungs below anything that should move your week. The authors put it plainly, calling the work a roadmap toward interventions that can be tested in cells, animals and eventually humans.
The word carrying that sentence is eventually.
The word carrying that sentence is eventually.
The more useful half of the result is the half nobody will market. The same screen flagged fourteen approved drugs that may push aging the wrong way, and that finding has no sponsor, no launch and no reason to be funded through to a trial.
You are unlikely to read a second story about those fourteen.
Meanwhile the money keeps flowing to the expensive end of the shelf. China’s anti-aging cosmetic sector ran to CNY 82 billion in 2021 and is projected at RMB 153.2 billion by 2026, and more than 46 percent of Chinese consumers were already budgeting over RMB 1,000 a year on anti-aging skincare alone.
Against that, a computational screen is close to free.
So price what is actually on offer here. Watching the list costs you nothing, and gives you a set of names to recognise if one of them reaches a registered trial. Buying on the strength of the screen costs you money today against evidence that does not exist yet, on a molecule whose approved use is a blocked nose.1
The interventions carrying real human outcome data have not changed this year, and none of them is for sale.
Load-bearing exercise. Weight and glucose in range. Sleep.
People who would notice if you went quiet for a week.
What to watch next is narrow and dated: whether any of the fourteen enters a registered human trial with an aging endpoint, and which one goes first. Barabási’s framework was built to be falsifiable, which is the most valuable property in it and the one least likely to survive translation into a product page.
A screen that can be proved wrong is worth more than a bottle that cannot.
A screen that can be proved wrong is worth more than a bottle that cannot.
Footnotes
-
The paper counts 2,358 longevity-associated genes; Northeastern’s own account describes narrowing 1,250 genes tied to the hallmarks. Both numbers are in the public record and describe different steps. That is a fair preview of how carefully any of this will be reported by the time it reaches a label. ↩