Research Guides Glossary Longevity Atlas Intervention Atlas Blog About Support Contact Join the research →

A study gets published. A press release follows. Within a day, the headline says “Scientists discover the key to reversing aging.” The actual paper said something far more modest — and almost nobody reads it.

This guide gives you a fast filter: enough to tell a real result from a marketing result in about twenty minutes. You don’t need a statistics degree. You need four questions.

Question 1: What did they actually study?

The most important sentence in any paper is the one describing what was measured and in whom.

  • In mice, cells, or people? Mouse results are real science and genuinely informative — but a mouse is not a person. If the paper is in mice and the headline implies it applies to you, that’s not the paper’s fault; that’s the media’s framing.
  • How many subjects? A study of 12 people can produce a real, honest finding. It just can’t produce a general one. Small samples are for generating hypotheses, not settling them.
  • How long? A six-week intervention tells you about six weeks. Longevity is a decades-scale question; short studies can’t answer it alone.

Question 2: What was the outcome?

The outcome is what they actually measured. It matters enormously:

  • A hard outcome — disease, death, a biomarker with known meaning — is what you want to see.
  • A soft or surrogate outcome — a questionnaire score, a “wellness score,” a lab value with unclear significance — can be statistically impressive and clinically meaningless.

A study showing a statistically significant change in a self-reported energy questionnaire is not a study showing less disease. Read the abstract for the exact outcome. If you can’t tell what it was, that’s a problem with the study’s reporting.

Question 3: What does “statistically significant” actually mean?

It means something very specific: the result is unlikely to be pure chance, given the study’s assumptions. It does not mean:

  • the effect is large,
  • the effect matters,
  • the effect applies to you.

A precise-sounding p-value on a tiny effect in a narrow population is a true statement that changes nothing. Think of significance as an entry ticket to the conversation — not the verdict.

Question 4: Who paid for it, and what’s the field consensus?

Science is a body of work. A single study — even a good one — is one data point. Before you change your behavior:

  • Check whether the finding has been replicated by independent groups.
  • Check who funded it. Funding doesn’t invalidate research, but a seller-funded study deserves extra scrutiny, especially in the longevity supplement space.
  • Check the broader picture: if one study is being quoted, ask what the other ten found. Cherry-picking a single favorable result is the oldest trick there is.

The fast filter, in practice

When you see a longevity claim, run it through the four questions in order:

  1. Species? Humans > mice > cells.
  2. Outcome? Real and meaningful > surrogate > questionnaire.
  3. Effect? Large and precise > small but real > statistically significant and nothing else.
  4. Consensus? Replicated by independent groups > single study > seller-funded single study.

If the claim fails early, you can stop reading. If it passes, you can dig deeper — and our Intervention Hierarchy scores the popular interventions against exactly this kind of evidence.

Bottom line: You don’t need to be a statistician to read a study. You need to ask what was measured, in whom, and by whom — and to treat any single study as a conversation, not a conclusion.

Not medical advice. This guide is about evaluating research claims; it is not guidance on making health decisions. Always consult qualified healthcare professionals.

Not medical advice. This guide describes current scientific understanding. It cannot tell you anything about your own biology — always consult qualified healthcare professionals for health decisions.
Support the research

Keep these guides free

Every guide on this site is free, forever — no paywalls, no premium editions, no products. If this one helped you, a one-time supporter badge keeps the evidence-based research and writing going.

← All guidesJoin the research →

Every guide is written for one reader: you, with no biology degree — and every claim is traced back to the evidence it actually has.

Editorial principle · Souverain Labs