The Longevity Reality Radar
46 approaches, rated by evidence.

How we rate longevity and cure claims

Every entry gets two separate judgements: a stage on the evidence ladder and an evidence grade. This page explains both, and the rules behind them.

The evidence ladder

The stage is decided by the most advanced credible program for an approach, checked against the trial registry rather than a press release. Counts are for all 46 entries on the radar.

  • Available now
    Approved or established. Usable today, with a clinician.    9
  • In human trials
    Phase 1 to 3. Being tested in people now. Most will not make it.    17
  • Early human signal
    First human or biomarker data. Promising, unproven.    5
  • Experimental
    Animal, lab or theoretical. Years away, if ever.    15

Evidence grades

The grade describes how strong the human evidence is for the claim, not our opinion of the idea.

  • ARandomised outcome data in people.3
  • BRandomised human data on surrogate markers, or consistent large cohorts.8
  • CSmall or uncontrolled human data, or strong animal data.21
  • DAnimal or lab only, single pilot, or case reports.14

Grades were last reviewed 14 September 2026 and are open to challenge.

Why stage and grade are separate

Stage says how far the most advanced program has got. Grade says how strong the evidence is for what is being claimed. The two can differ: a drug in Phase 3 for one disease can carry a C grade for an aging claim that rests on a small post-hoc biomarker analysis, as rentosertib does.

Randomised data on a marker, such as a blood level or an aging clock, can reach B at most. An A needs randomised trials in people that measure outcomes: disease, function or survival. Why markers are not outcomes.

The rules we follow

  1. Every load-bearing claim is checked against the primary paper, never against the framing of the post or press release that pointed us to it.
  2. Stages are conservative, and the gap is stated: Phase 3 for another disease is not Phase 3 for aging, and a randomised effect on a marker is not outcome evidence.
  3. Surrogate markers are labelled as surrogates.
  4. When a widely shared claim is wrong, the entry names it and corrects it with the counter-evidence, rather than leaving it out.
  5. Null and contradicting trials go into the same entry as the positive ones.
  6. Sellers are not sources. Posts written to sell a supplement, a protocol or an affiliate link are not featured, whatever they cite.
  7. Every move on the ladder and every correction is dated and public on the change log.
  8. No affiliate links, no sponsored entries, no industry money. Entries describe evidence; they never tell you what to take.

How sources are labelled

  • Primary source: a journal article, a registry record or a regulator's database.
  • Official or company statement: a press release, regulator announcement or institution news. Useful for dates and topline results, not for judging effect sizes.
  • Coverage, not a source: articles, posts and pages that pointed us to the research. They are linked for transparency and marked nofollow; the grade never rests on them.

Where an entry has no primary paper linked yet, its page says so.