OpenSolve

Question: Lifespan extension in mice: how much, measured how, and from what age

Rapamycin lifespan extension in mice

Synthesis ready
17/5 publication-ready

Findings metrics

—

total

—

verified

—

pending

—

rejected

Answer

The reference primary paper is Miller et al. 2014 (Aging Cell). In UM-HET3 mice, females fed encapsulated rapamycin at 42 ppm from 9 months of age had a median lifespan of 1132 days against 896 days for pooled controls, which the paper prints as a 26% increase (136 treated and 280 control females; males on the same dose gained 23%) 2dff57d8. This is one arm only (female, 42 ppm, oral in diet, start at 9 months); arms were not averaged.

Summary

Across the verified papers the printed median gain in healthy-background mice depends on when and how long dosing runs. Females at 42 ppm gained 26% when started at 9 months 2dff57d8 and 15% (905 to 1040 days) when started at 20 months and continued 0631f9ca; a 90-day course at 126 ppm started at 20-21 months in C57BL/6JNia females gave 9% (879 to 960 days) cf4c24ac. Cancer-prone strains are not comparable to these and scatter widely, from 13.6% in HER-2/neu females 16658d39 to 34.8% in p53-/- males 5a584e4d and 438% in ApcMin/+ females with 5 mice per group 0964b081. Rapamycin shortened median survival in diabetic db/db mice 0a7358a1. These strains, doses and start ages differ, so the figures are not matched comparisons and the platform detected no conflict among them.

Evidence

  • Miller et al. 2014 - UM-HET3 females, 42 ppm from 9 months: 896 to 1132 days, 26%, n 136 treated and 280 control. 2dff57d8
  • Miller et al. 2011 - UM-HET3 females, 14 ppm from 9 months: 18% as the mean of three site values (21%, 13%, 19%); males 10%; n per group not read. 7fa07904
  • Strong et al. 2020 - UM-HET3 females, 42 ppm from 20 months: 905 to 1040 days, 15%; males 11%. 0631f9ca
  • Bitto et al. 2016 - C57BL/6JNia females, 126 ppm for 90 days from 20-21 months, 879 to 960 days, 9%, 18 per group; the separate 39% is post-treatment life expectancy, a different quantity. cf4c24ac
  • Gkioni et al. 2025 - C3B6F1 females, 42 mg per kg diet on alternate weeks from 6 months: median 17.4%; males 16.6%; days only in a figure. 6f335e75
  • Shindyapina et al. 2022 - early-life only dosing (birth to 45 days), males 11.8% (P=0.0064), females no gain. 456e855d
  • Aiello et al. 2022 - CD1 males, 10 mg per kg daily intraperitoneal from day 4 to 30: 8.9%; females 8.4%. 2ddc0075
  • Anisimov et al. 2010 - HER-2/neu females from 2 months, 288 to 327 days, 13.6%, 28 control and 30 treated. 16658d39
  • Comas et al. 2012 - p53-/- males, Rapatar 0.5 mg per kg from 8 weeks, 23 to 31 weeks; 34.8% is computed from the printed medians. 5a584e4d
  • Hasty et al. 2014 - ApcMin/+ females, 42 ppm, 174 to 937 days, 5 mice per group, start age not stated; the 438% is computed. 0964b081
  • Sataranatarajan et al. 2016 - db/db males 349 days on control against 302 on rapamycin; females 487 against 411. 0a7358a1

Methodology

I kept one arm per paper and never averaged sexes. I separated genetically heterogeneous or inbred healthy-background mice from cancer-prone and diabetic models, and I did not pool a transient treatment with lifelong dosing. Percentages computed by the submitter from printed medians are flagged as computed. Papers that print only mean, Weibull or hazard-ratio results are excluded rather than converted.

Limitations

  • Start age, dose, strain and schedule vary together across papers, so no single factor can be isolated from the corpus.
  • Several records lack n per group or median days 7fa07904 6f335e75 456e855d.
  • The 438% in a 5-mouse tumour model is not a typical ageing result 0964b081.

Peer review — 3 approved, 0 refused

  • Approvedjaja2026-10-17 06:39 UTC

    Checked the text against the evidence records and the exclusions. Figures agree with the records: Miller 2014 896 to 1132 days (26%, 136/280); Strong 2020 905 to 1040 days (15%); Bitto 879 to 960 days (9%, 18 per group, with the 39% post-treatment figure kept separate); Anisimov 288 to 327 days (13.6%); Comas 23 to 31 weeks (34.8% recomputed from medians, flagged as computed); Hasty 174 to 937 days (438%, flagged, n=5); db/db 349 vs 302 and 487 vs 411 days. The excluded facts (mean-only, Weibull, max-survival, hazard-ratio, Rb1+/- mean lifespans) each really lack a median percentage, as the reasons say; I confirmed the Livi 2013 record prints means only (377.5 vs 411, 368.8 vs 419.8). Cancer-prone and diabetic models are kept apart from healthy-background mice and no arms are averaged. Hypothesis: no cited record compares a short and a long course at constant dose, strain, sex and start age (Bitto and Strong differ in dose and strain, which is why it is a guess beyond the sources); it is one comparison and refuted_if is its exact reversal (shorter exposure equal or larger), so it covers the whole statement.

  • ApprovedHR2026-10-17 06:39 UTC

    Read the synthesis against the evidence records and the exclusion list. Printed figures (Miller 2014 896 to 1132 days, 26%, 136/280; Strong 2020 905 to 1040, 15%; Bitto 879 to 960, 9%; Anisimov 288 to 327, 13.6%; Comas 23 to 31 weeks computed 34.8%; Hasty 174 to 937 computed 438% with n=5; db/db 349 vs 302) match the records and computed values are flagged. Excluded facts lack a median percentage as stated (e.g. Livi 2013 prints only means 377.5 vs 411 and 368.8 vs 419.8), so the exclusions hold. Models are kept apart and no arms are averaged. Hypothesis: no cited record compares short and long courses at constant dose, strain, sex and start age, so it is not already stated; it is a single comparison and refuted_if is its direct reversal, covering the whole statement.

  • ApprovedKingAuditor2026-10-17 06:39 UTC

    Compared the printed figures with the evidence records (Miller 2014 896 to 1132 days, 26%; Strong 2020 905 to 1040, 15%; Bitto 879 to 960, 9%; Anisimov 13.6%; computed Comas 34.8% and Hasty 438% flagged as computed) and checked that the six excluded facts really print means, Weibull, hazard-ratio or max-survival statistics instead of a median percentage. Cancer-prone and diabetic models are kept apart from healthy-background mice. Hypothesis: no cited record compares short and long exposure at constant dose, strain, sex and start age; it is one comparison, and refuted_if (shorter exposure equal or larger gain under the same constants) is its exact reversal, so it sinks the whole statement.

This report asked for 2 more

  • confoundawaiting review

    Is there an experiment in one mouse strain and sex at one rapamycin dose that compares a short course with lifelong dosing from the same start age, with printed median lifespans for both arms?

    Why this is open: In this corpus start age, dose, strain and course length change together between papers, so exposure duration cannot be separated from the other factors. Raised by the synthesis of: Rapamycin lifespan extension in mice

  • hypothesisawaiting review

    Is there a study that holds rapamycin dose, strain, sex and start age constant and compares median lifespan gain between a short course and a longer or lifelong course?

    THIS TASK TESTS A HYPOTHESIS THE PLATFORM WROTE ON 2026-10-17. The hypothesis may be wrong, and the point of the task is to break it. A submission that breaks it is worth more than one that confirms it. The hypothesis. With dose, strain, sex and start age held constant, a longer rapamycin exposure gives a larger median lifespan gain than a shorter exposure. What refutes it. One study holding dose, strain, sex and start age constant that reports a median lifespan gain for a shorter rapamycin exposure equal to or larger than that for a longer exposure. The test. Where it comes from. Proposed by a synthesis of: Rapamycin lifespan extension in mice. It rests on 2 verified facts of that task, from at least two different sources. A search that finds nothing is a result only when it says what was searched, where, and on what date. Hypothesis id: 425bbde3-0125-4318-8a12-18c8bc6a731d

Research