Longevity science guides, graded by evidence

How this longevity science archive is organized

Aging research moves fast, so every guide below carries its last-updated date, and none is older than 2 months. Most of that research happens in mice, worms and cell dishes long before anyone tests it in people, and the evidence table grades it all.

What is longevity science?

It is the study of why organisms age and whether aging can be slowed. The field spans cell biology, animal lifespan studies and human trials. The guides here keep those worlds apart, so a promising lab result never reads like a proven human benefit.

Every guide answers one of five kinds of question. Evidence reviews ask whether a compound or habit does what its marketing says. Explainers cover the biology underneath, from cell signaling to why damaged cells stop dividing.

• Dosing guides report the amounts actually used in published trials, not the amounts on a label.
• Safety guides separate side effects measured in people from those inferred from other populations.
• Comparisons set two options side by side on the same evidence standard.

The chart below counts the archive by kind. It is drawn from the same list as the cards, so it changes whenever a guide is added.

How the evidence is graded

The grades follow the GRADE approach, a framework for rating the quality of evidence published in the BMJ in 2008. We use three plain-language levels, because readers need to know what a claim rests on more than they need a score.

• Strong human evidence: several randomized trials in people with consistent results.
• Early or mixed evidence: some human data, or strong animal data with thin or conflicting human results.
• Limited evidence: mechanistic, preclinical or anecdotal support, with little or no human outcome data.

Why do so many longevity findings come from animal studies?

Human lifespan trials would take decades and very large groups of volunteers. Researchers test interventions in short-lived species first, then look for shorter-term markers in people. A result in mice is a reason to run a human trial, not proof of a human benefit.

• Even careful mouse work does not always hold up. The NIA Interventions Testing Program found fisetin did not significantly extend lifespan in either sex (Harrison et al. 2023, GeroScience), though only one dose and delivery method was tested.
• Supplements reach the shelf without any such test. The FDA does not approve dietary supplements before sale, and the manufacturer carries the duty to back a label claim (FDA structure/function guidance), although the agency can act after marketing.

Most compounds in the aging field sit in the middle or bottom grade. That is not a verdict that they fail; it means the human trials that would settle the question have not been run, or were too small and too short to measure healthy lifespan.

Can a single study settle a longevity question?

Rarely. A widely cited 2005 analysis in PLoS Medicine argued that small studies, flexible designs and crowded research fields make any single positive result provisional. That is why the guides weigh bodies of evidence rather than headlines.

The full rules, including how corrections are handled, are in the editorial standards.