A university in Singapore announced today that one of its professors had, in a study published yesterday in PLOS ONE, reduced his estimated biological age from 47 to about 32. The study had one participant, the author, over two years. The protocol: a four-hour daily eating window, periodic 48-hour water-only fasts, daily strength training plus high-intensity intervals, a Mediterranean-style diet with supplements, a bedtime moved nearly three hours earlier, and continuous monitoring with three wearables, blood panels and microbiome sequencing. The headline will travel as "professor reverses aging by fifteen years." The paper is more honest than the headline, and the honesty is the useful part.
What the paper says about itself. The "biological age" came from a custom AI tool the authors built, which they describe as having no external validation and no pre-specified performance criteria. The resilience measure at the center of the study is, in their words, a proof-of-concept rather than a validated biomarker. There is one subject, so nothing can be separated from anything else. The author discloses that he co-founded one company, advises three others and is a co-inventor on pending patents covering the tools used. None of this is hidden; it is all in the paper. And look at what actually moved: resting heart rate from 65 to 46, sleep from about five hours a night to nearly eight, and, after the first 48-hour fast, a doubling of homocysteine that resolved on refeeding. The measurable changes came from sleeping three more hours and lifting weights. The number everyone will quote came from the unvalidated tool.
Ben’s takeAn n of 1 is an anecdote with a spreadsheet, and the spreadsheet is what makes it dangerous: it looks like data. This one is unusually well labeled — the limitations are stated, the conflicts are disclosed, the tool is called unvalidated by the people who built it — and it will still be stripped of all of that by the time it reaches a product page. The pattern is the longevity industry in miniature: a story of self-transformation, an impressive number, a method for sale. The person is not the problem. The genre is.
Loree’s takeStrip the protocol to what the paper's own measurements support and you get: he went to bed at nine, slept eight hours and trained every day. Those are fundamentals, and they would do most of what this study reports for most people without a 48-hour fast or a four-hour eating window, both of which carry risks the paper itself recorded. If a patient brought me this, I would keep the bedtime and lose the fasting. The biological-age number I would not discuss at all, because the authors say it means nothing yet.
Why it matters to a recovering optimizer. Self-experiments are seductive because the subject is the hero and the numbers are his. This one is a rare case where the hero wrote the limitations section himself. Read it before the headline, and notice what it proves: an intense protocol, honestly measured, produced changes that sleep and movement explain. That is the finding. The rest is a tool that has not been tested.
When one person's results come with a number, find the limitations section. If the authors call the tool unvalidated, so should you.
This post reports on evidence and expert disagreement. It is not medical advice; decisions about treatment belong with you and your clinician.