Opens in a new tab
Free worldwide shipping
← All research
Longevity Meta-Analysis

Rapamycin, Not Metformin, Mirrors Dietary Restriction-Driven Lifespan Extension In Vertebrates: A Meta-Analysis

Source publication · 2025 · Ivimey-Cook ER, et al.

Research Summary

A systematic review and meta-analysis comparing how much lifespan extension dietary restriction, rapamycin and metformin produce across vertebrate species, and whether the effects depend on sex or on how dietary restriction is applied.

Who Or What Was Studied?

167 papers reporting 911 effect sizes in eight vertebrate species: mice, rats, dogs, rhesus monkeys, mouse lemurs, sticklebacks, redtail killifish and turquoise killifish. Papers were identified by PRISMA searches of Scopus and Web of Science and compared each treatment group against its own control group. Effect sizes were log-response ratios of means and medians, analysed with multi-level models that accounted for species, paper and observation-level variance.

What Did The Research Show?

  • Dietary restriction extended lifespan robustly in every model: log-response ratio 0.172 (95% CI 0.132 to 0.213) after adjustment for publication bias, p < 0.001.
  • Rapamycin also extended lifespan in the combined analysis: 0.216 (95% CI 0.152 to 0.279), p < 0.001, and its effect was not statistically different from that of dietary restriction (p = 0.221; difference 0.044, 95% CI -0.026 to 0.114).
  • Metformin did not: 0.086 (95% CI -0.007 to 0.178; p = 0.069), and its effect was significantly smaller than rapamycin’s (p = 0.017).
  • Heterogeneity was very high (I-squared 96.5% overall), and the estimate depended on how lifespan was summarised: the positive rapamycin effect was not present when only log-response means were analysed, although it appeared for medians and for means adjusted for publication bias.
  • No consistent sex difference appeared across treatments, and the two dietary restriction methods (percent reduction and fasting) both extended lifespan with no significant difference between them.

Limitations And Interpretation

  • Animal data only: the analysis covers laboratory vertebrates, and the authors state that generalisation to humans is not established.
  • Heterogeneity was very high and publication bias affected the estimates, so the pooled effect sizes should be read as averages of variable studies rather than a single precise effect.
  • Applying the 900-day rule (removing studies with short-lived controls) made the rapamycin effect based on means overlap zero, showing sensitivity to study selection and summary statistic.
  • Species coverage is uneven, with mice dominating the dataset, and laboratory conditions differ from natural environments.

Summary prepared from the published paper (abstract and full text where accessible). This is not medical advice — always consult the original publication.

Source: publication record ↗