The LEV Scoreboard
The LEV Scoreboard · LEV Institute

Mouse lifespan results · NIA Interventions Testing Program

The U.S. government’s Interventions Testing Program gives candidate anti-aging compounds a rigorous, independent test in mice, and most of them fail.

Run by the National Institute on Aging, it tests every compound the same way at three independent labs, in genetically diverse mice, and publishes the failures alongside the successes. Here is every published result: what happened in each sex, and where the human evidence stands.

Not medical advice. This page reports mouse experiments and LEV’s reading of the human evidence. It doesn’t recommend any compound, and mouse doses are never converted into human doses.

    Every compound, one square each

    Select a square or a group to see the details.

    • Left half: males
      Right half: females
    • Lived longer
    • Lived shorter
    • No significant change

    All compounds

    1. Loading the data…

    How to read this

    The mouse result is a published fact. The human-evidence grade is LEV’s judgment. We keep them apart on purpose.

    What the ITP is

    The Interventions Testing Program is run by the U.S. National Institute on Aging. Researchers propose compounds; after expert review, the program tests the chosen ones in parallel at three independent labs (The Jackson Laboratory, the University of Michigan and the University of Texas Health Science Center at San Antonio) under the same protocol.

    The mice are UM-HET3: each one is a genetically unique cross of four inbred strains, so a result can’t hinge on one strain’s quirks. The compound is mixed into the food; doses are given in ppm (parts per million in the chow). The original design called for about 132 males and 108 females per compound across the three labs, enough to detect a 10% change in lifespan.

    Early tests started treatment at 4 months of age; many later tests start in middle or late life. In this dataset, start ages run from to months.

    How we read a result

    Primary test
    The ITP’s own pre-specified analysis: a log-rank test comparing treated and control mice of the same sex, stratified by lab, with the three labs pooled.
    Extended / shortened
    Significantly longer or shorter life (p < 0.05) in that test.
    Trend
    0.05 ≤ p < 0.10. Shown for completeness, never counted as a result.
    Median change
    Percent change in median lifespan versus same-sex controls, as reported in the paper. A dash means it wasn’t reported or we haven’t captured it. The small bars use one scale for every compound.
    Reanalyses
    Later statistical reanalyses, such as a 2024 reanalysis with the Gehan test and a 2026 reanalysis accounting for mice that share cages, are shown next to the result. They never change the verdict.

    Verdicts combine every test of a compound:

    Why mouse results don’t carry over to people

    • Mice aren’t small people. They live a few years, eat one controlled diet in a clean lab and die of different diseases than we do.
    • A dose in chow isn’t a pill. How much reaches the blood depends on the species, the food and the compound. We never convert mouse doses into human doses.
    • Lifespan is one number. These studies aren’t built to find the side effects that matter to people, or to show whether the extra time is healthy time.
    • Sex differences are real but unexplained. Several compounds helped only male mice. That doesn’t tell us how men and women would respond.
    • A failure isn’t final, and a success isn’t a prescription. Another dose, start age or formulation might behave differently. A success here means “worth testing in people,” not “worth taking.”

    Human-evidence grades

    Each grade names the claim we graded and the people it applies to, because a compound can be proven for one use and unproven for another. A drug approved for diabetes, for example, can still be “Speculative” for slowing aging.

      Grades carry a Draft badge until LEV’s editor approves them. Only approved grades appear on the public site.

      Independence

      LEV Institute sells nothing on this page: no products, no affiliate links, no sponsorships. Nobody can pay to be listed, graded or left out. We name compounds, never brands.

      Corrections

      Found an error? Send the compound, what’s wrong and your source to . We check every report against the primary paper (the paper wins over any summary, including ours) and log each correction below with its date.

      Updates

        Sources for this section

          Get notified when a grade changes

          We’ll email you when LEV changes a human-evidence grade or adds new ITP results.

          Download the data

          Every test row, with its source. The JSON holds the full record (sources, definitions, reanalyses); the CSVs open in any spreadsheet.