Sixty-four milligrams per kilogram, by oral gavage, three times a week, in mice already 23 to 25 months old. Rotarod latency came back to where four-month-old animals sit. Frailty scores held flat while the untreated animals kept getting worse.

The compound is palbociclib. It has been on pharmacy shelves since 3 February 2015.

Rajesh and 25 co-authors report that cyclin D1 accumulates in senescent cells that have stopped dividing, and that it drives the inflammatory secretions those cells produce. The route runs through DNA damage: chromatin fragments leak into the cytoplasm, cGAS-STING fires, interferon genes switch on.

Knock the gene out in the liver of a 17-month-old mouse and the damage marker falls within three weeks. Give the whole animal an approved CDK4/6 inhibitor instead and the same thing happens.

Two months of dosing, three times a week, starting at 18 months.

Now the part the abstract does not price.

At the approved oncology schedule, 125 milligrams daily for 21 days on and 7 off, PALOMA-1 recorded grade 3 neutropenia in 57 percent of patients and grade 4 in another 5. That trial ran in 165 women with metastatic breast cancer, where losing your white cells is a trade you make on purpose.

Evidence grade on the ageing claim: preclinical, single species, liver-weighted. Your liver is not a 24-month-old mouse liver, and nobody has run this in a healthy human.

The molecule’s real asset is the eleven years of published harm behind it.

The molecule’s real asset is the eleven years of published harm behind it, which the mouse result does nothing to add to.

That is a genuinely different asset class from what the ageing market usually sells, and a second paper landed in the same week to make the point twice. On 20 August, a group reported that blocking RANKL, the target of the monoclonal antibodies already sold for osteoporosis, preserved bone, improved grip strength and endurance, and extended survival in progeroid mice.

Two approved mechanisms. Two ageing results. One week.

Set that against what is actually being sold in Singapore this month. A clinic here runs NAD+ infusions at 100 milligrams over an hour or 200 over two, in courses of three to ten consecutive days, on a page that claims ageing care and immunity and presents no human trial.

A biological age blood test on the same island costs S$108 at one clinic, and the same page puts comparable tests at longevity clinics at S$1,500 to S$4,000 and up.

You are being sold the newest evidence at the highest price and the oldest evidence at the lowest.

You are being sold the newest evidence at the highest price and the oldest evidence at the lowest.

What that buys you is a sharper question for any clinic taking your money. Every intervention on the market is backed by a study, so that fact sorts nothing.

Ask which species the study ran in, at what dose relative to the one you are being offered, and what the published harms are at that dose.

A molecule with eleven years of prescribing data can answer all three in a paragraph. A drip cannot answer the third at all, because nobody has looked hard enough to know.

The next thing to watch is a trial registration, not a paper. The step that would move palbociclib from interesting to usable is a dose-finding study in healthy older adults at an intermittent schedule far below the oncology one, with neutrophil counts as the primary safety endpoint.

Until that registration appears, the honest description of what happened this month is that two drugs your parents may already be taking for something else did something interesting in a mouse.