In November 2006, Nature published a paper showing that resveratrol, a compound found in red wine, improved the health and survival of mice fed a high-calorie diet[1]. Within eighteen months, a company founded on the finding had been bought by GlaxoSmithKline for around $720 million[2]. Within four more years, the lead compound had been shelved[3] and the mechanism everyone had assumed was doing the work had been directly challenged in print[4].
That sequence is the most useful thing in longevity research, and almost nobody tells it. The field has a volume problem: the interventions with the best evidence are the dullest to write about, and the ones with the thinnest evidence produce the most content. Sorting one from the other is most of the work, and the resveratrol story is a free lesson in how.
What actually happened
The 2006 result was real and well conducted. Mice on a high-fat diet given resveratrol lived longer than mice on the same diet without it, and looked metabolically healthier by several measures[1]. The proposed explanation was that resveratrol activated SIRT1, a sirtuin protein already linked to lifespan in yeast. A biotech company, Sirtris, was built to turn that mechanism into drugs, and in 2008 GSK acquired it[2].
Then two things went wrong, in the ordinary way science goes wrong. In 2010, a paper in the Journal of Biological Chemistry reported that resveratrol and Sirtris's synthetic compounds were not direct activators of SIRT1 at all, and that the earlier result depended on an artefact of the assay used to measure it[4]. Separately, GSK halted development of SRT501, its resveratrol formulation, after kidney problems appeared in a trial in myeloma patients[3].
Neither event proved resveratrol useless. Both established that the story was more complicated than the version that had already reached supplement shelves, and the supplement shelves did not update. You can still buy resveratrol today sold on the strength of a 2006 mouse paper, marketed with the phrase “activates sirtuins”, nineteen years after the acquisition and fifteen after the mechanism was contested.
The compound did not fail. The story about the compound failed, and only the story was ever being sold.
What the story teaches
Three things, each of which transfers to almost every longevity claim you will encounter.
A mouse result is a hypothesis about humans, not a finding about them. The 2006 paper was strong evidence about mice on a high-calorie diet. It was never evidence about a person taking a capsule, and the gap between those two statements is where most of this industry lives.
A mechanism is not an outcome. “Activates SIRT1” sounds like a benefit and is not one; it is a description of a chemical event several steps removed from anything that happens to you. When the mechanism turned out not to hold, nothing about the marketing changed, because the marketing had never depended on it being true.
Money is not evidence. A $720 million acquisition is a statement about how a pharmaceutical company priced a risk in 2008, not about whether the compound works. Funding, press coverage and founder credentials are all easy to mistake for validation, and none of them are.
What has held up
The unglamorous list, and it is short. Not smoking: the British Doctors Study followed 34,000 physicians for fifty years and found lifelong smokers lost around ten years of life, while those who stopped at 30 avoided almost all of that loss and those who stopped at 50 avoided about half[5]. Treating high blood pressure: the SPRINT trial randomised more than 9,000 people to a target systolic pressure below 120 rather than below 140, and stopped early because the intensive group had substantially lower all-cause mortality[6]. Cardiorespiratory fitness: in a cohort of over 120,000 people who underwent treadmill testing, higher fitness tracked with lower mortality with no observed upper limit to the benefit[7].
Add adequate sleep and maintaining muscle mass into later life, both supported by large observational literatures rather than single decisive trials. That is close to the whole of the settled ground, and every item on it was established before longevity became a market.
The honest middle
Some interesting things are neither proven nor discredited. They are in trials, and should be described that way rather than sold. Metformin, a cheap and long-used diabetes drug, is the subject of a proposed trial designed to test whether it delays age-related disease in people who do not have diabetes[8]. Rapamycin and related mTOR inhibitors have shown effects on immune function in older adults in early human work[9]. Both are worth following. Neither is a protocol, and anyone selling you either as one has skipped a step you should not skip.
Most supplements sit in a different category again: tested, and found wanting. A Cochrane review of antioxidant supplements in healthy participants found no reduction in mortality, with signals of increased mortality for beta-carotene and vitamin E[10]. That is a stronger result than “no evidence”, and it is worth knowing before buying anything sold as an antioxidant.
How to read the next claim
Four questions do most of the filtering. What was measured, a biomarker or something that actually happened to a person? What was the comparison group? How long was the follow-up, given that the thing being claimed takes decades to appear? And who benefits if you believe it?
Resveratrol would have failed the first question in 2006, when the honest description was that it worked in mice. It failed the fourth by 2010, when the mechanism was contested and the capsules kept shipping anyway. Nothing about that required expertise to spot, only the willingness to ask what the claim was resting on.
This is not medical advice, and it does not replace a clinician who knows your history. It is a reading strategy, which in a field this loud is worth more than another list of compounds.