The gerotherapeutic drugs rapamycin, acarbose, and phenylbutyrate extend lifespan and enhance healthy aging in house crickets (paper Jan 26) - JH "I think the title is misleading"

chatGPT:

What the preprint did (summary)

This is a 2026-01-05 Research Square preprint (not yet peer reviewed) that tests whether three well-known “gerotherapeutics” — rapamycin, acarbose, and phenylbutyrate — can improve lifespan and late-life function in the house cricket (Acheta domesticus) under an intermittent, mid-life dosing schedule. (PMC)

Design

  • 274 sex-matched crickets.
  • Intermittent dosing for 2 weeks starting at mid-age (8 weeks), then behavioral testing at 10 weeks (geriatric stage).
  • Health/behavior assays:
    • Y-maze olfactory preference/discrimination
    • Open-field exploration (distance, speed, freezing, zone occupancy; automated tracking)
    • Treadmill running and jump performance
    • “Gaits” classified via velocity-based k-means clustering. (PMC)

Main findings

  • Olfaction (Y-maze): All three drugs preserved “food-odor preference” vs controls (largest effects reported for rapamycin). (PMC)
  • Open field: Drug effects were mixed and somewhat “behavioral-state”-like (e.g., rapamycin increased central engagement but also increased freezing; acarbose/phenylbutyrate increased peripheral freezing). (PMC)
  • Physical performance: Rapamycin and phenylbutyrate increased treadmill running duration relative to controls. (PMC)
  • Survival (post-treatment):
    • Rapamycin extended survival (reported HR ~0.42 overall; also significant in males).
    • Acarbose shortened survival in females (reported HR ~3.03 in sex-stratified analysis).
    • Phenylbutyrate showed no clear survival advantage in the text around the survival results. (PMC)

Overall, the authors argue this supports A. domesticus as a scalable geroscience screening model and that intermittent rapamycin is the most consistently beneficial of the three. (PMC)


What’s novel here

  1. Intermittent (“pulse”) dosing in a cricket geroscience model
    Prior work in many organisms uses continuous exposure; here the focus is whether a short mid-life course can yield downstream benefits (especially for rapamycin). (PMC)
  2. Multi-domain “healthspan” phenotyping in an insect
    They combine cognition-like behavior (olfactory choice), exploration/anxiety-like measures (open field), and induced endurance (treadmill), rather than just lifespan. (PMC)
  3. Quantitative gait-state classification using k-means thresholds
    Using velocity clustering to define walking vs running thresholds is a more explicit, data-driven approach than hand-picked cutoffs. (PMC)
  4. Clear sex-specific divergence, including a potential harm signal (acarbose in females)
    The female-specific survival penalty is a noteworthy (and testable) warning flag. (PMC)

Critique (what to be cautious about)

1) It’s a preprint + some internal references are “manuscript in progress”

The work hasn’t been peer reviewed, and at least some methodological context is referenced as not-yet-published. That doesn’t invalidate results, but it raises the bar for replication and clarity. (PMC)

2) Heavy reliance on historical/archived comparator cohorts

They compare treated animals not only to contemporaneous controls but also to “archived younger” and “historical geriatric” cohorts. That can be informative, but it can also introduce batch, husbandry, season, supplier, microbiome, or subtle environmental differences that masquerade as “rejuvenation.” (They mention using historical controls to reduce batch artifacts, but it remains a common risk.) (PMC)

3) Behavioral interpretation is ambiguous (especially “freezing”)

Increases in freezing could reflect:

  • reduced vigor / sedation,
  • increased anxiety-like avoidance,
  • altered risk assessment,
  • or context-specific motivational shifts.
    Because rapamycin shows both “central exploration” and increased freezing, it’s not straightforward to label that as purely improved healthspan without additional validation. (PMC)

4) Limited mechanistic anchoring

The paper is largely phenotypic: survival + behavior. They explicitly list future needs like histology (mushroom bodies, muscle), standardized geropathology, and more granular kinematics. Without tissue/cellular readouts, it’s harder to connect effects to conserved aging hallmarks versus nonspecific pharmacology. (PMC)

5) Power and multiple testing across many endpoints

They do multiple behavioral metrics and stratify by sex, and they acknowledge that some assay sample sizes may be underpowered for modest effects. This increases the chance that some “selective benefits” are fragile (and need preregistered replication). (PMC)

6) The acarbose “harm in females” needs careful follow-up

A female-specific hazard increase is important, but to interpret it you’d want:

  • replicated cohorts,
  • dose verification and intake confirmation,
  • and checks for tradeoffs (reproduction, infection, gut effects, etc.).
    Right now it’s a strong signal, but still provisional. (PMC)

If you want, I can also extract a compact “results matrix” (drug × domain × sex: benefit/neutral/harm) from the Results text so you can see the pattern at a glance.

Here’s a compact results matrix (direction vs contemporaneous controls), pulled from the Results text/figure captions. I’m using:

  • better / higher
  • worse / lower
  • 0 no clear change
  • ~ trend / mixed
  • Notes in parentheses give sex-specific effects when stated.

Results matrix (Drug × Outcome)

Outcome (10-week testing unless noted) Rapamycin Acarbose Phenylbutyrate
Olfaction (Y-maze vanilla preference) (strong; males, also females vs control) (overall; strongest in males) (overall; strongest in males)
Open-field total distance 0 (but males match juveniles in historical comparison) 0 (males lower vs control)
Open-field average speed 0 (but males retain juvenile-like speeds in historical comparison) 0 ~↓ (trend lower overall; males trend lower)
Walking:running distance ratio 0 / favorable vs historical ratio (more walking relative to running) 0
Walking speed 0 0 (driven by males)
Running speed 0 (not reduced; males match juveniles in historical comparison) (overall) (esp males overall; sex-stratified note highlights females reduced)
Exploration / spatial bias (central occupancy / engagement) (trend overall; males higher central occupancy/time ratio) ~ (some female trends; not consistent) 0 / ~ (no central occupancy change; some mixed trends)
Freezing (overall) ↑ freezing (both sexes) ~↑ (trend overall) ~↑ (trend overall)
Freezing (central zone) (only clear increase) 0 / mixed 0 / mixed
Freezing (peripheral zone) ~↑ (female trend) (mainly females) ~↑ (comparisons suggest more in periphery; pooled analyses emphasize males)
Treadmill running duration (endurance) (both sexes) 0 (not reported as improved) (both sexes)
Jump distance (power) 0 overall (but females better vs historical cohorts) 0 overall (but males better vs historical cohorts) 0
Weight change (during/after treatment) 0 (but females > males within group) (females gained more vs control females) 0 (but females > males within group)
Post-treatment survival (overall HR ~0.42; males benefit) (females harmed; HR ~3) 0 (no survival advantage vs control in text; females outlived males within PBA group)

Source lines for each block: olfaction Results/Fig 2; locomotion Results/Fig 4; exploration+freezing Results/Fig 5; treadmill+jump Results/Fig 6; weight Fig 7; survival Results/Fig 8. (PMC)


Quick takeaways from the matrix

  • Rapamycin is the only one that looks consistently pro-“healthspan” across olfaction + endurance + (some) exploration, but it also increases freezing. (PMC)
  • Acarbose shows olfaction benefit but a female survival harm signal, plus a shift toward more “walking vs running.” (PMC)
  • Phenylbutyrate improves endurance and olfaction, but shows reduced activity/walking speed in males and no clear survival benefit in the main text. (PMC)

If you want, I can convert this into a stricter matrix that only marks ↑/↓ when the text explicitly reports P < 0.05 vs controls (and otherwise marks 0/~).

I had a further question (chatGPT response)

From Figure 8 (Kaplan–Meier medians), the paper reports these median survival (days) and the implied lifespan change vs control:

Overall (both sexes pooled)

  • Control: 104.0
  • Rapamycin: 136.0 → +32.0 days (+30.8%)
  • Acarbose: 91.0 → −13.0 days (−12.5%)
  • Phenylbutyrate: 99.5 → −4.5 days (−4.3%)

Females

  • Control: 112.0
  • Rapamycin: 136.0 → +24.0 days (+21.4%)
  • Acarbose: 93.5 → −18.5 days (−16.5%)
  • Phenylbutyrate: 116.5 → +4.5 days (+4.0%)

Males

  • Control: 97.0
  • Rapamycin: 132.5 → +35.5 days (+36.6%)
  • Acarbose: 87.5 → −9.5 days (−9.8%)
  • Phenylbutyrate: 83.0 → −14.0 days (−14.4%)

So the clear “lifespan extension” signal in their own survival table is rapamycin (largest in males, substantial in females, and +30.8% overall).

Another response:

Here’s what’s been published since 2020 in model species on the lifespan effect of acarbose (ACA), focusing on primary data where possible.

Mice (Mus musculus)

1) Genetically heterogeneous UM-HET3 (NIA Interventions Testing Program context)

Most of the core lifespan demonstrations for ACA in UM-HET3 are pre-2020, but post-2020 papers continue to quantify dose/sex/timing and provide contextualized effect sizes:

  • A controlled comparison paper on early vs midlife starts summarizes prior lifespan results, including:

    • ACA 1000 ppm from ~8 months: ~16–17% male lifespan extension; ~5% female (UM-HET3).
    • ACA started at ~16 months: smaller effects, ~6% male, ~2% female. (PMC)
      (This paper itself is mainly about function/cardiac structure, but the lifespan numbers are explicitly stated as prior findings.)

2) Combination treatment: Rapamycin + Acarbose (UM-HET3; ITP multi-site)

  • Strong et al., 2022 (Aging Cell): Rapamycin + acarbose started at 9 months produced a large lifespan increase:

    • Median lifespan: +28% (females) and +34% (males) (pooled across sites), and it was the only treatment that significantly improved survival in both sexes at all three sites. (PMC)
      Important nuance: this is a combo effect; it doesn’t isolate ACA alone, but it’s one of the strongest modern datasets involving ACA in an aging paradigm.

3) Mitochondrial disease model (Leigh syndrome): Ndufs4−/− mice

  • Bitto et al., 2023 (also appeared as a 2022 bioRxiv preprint): ACA improved survival / extended lifespan in Ndufs4−/− mice, and the authors report additive benefits with rapamycin plus microbiome/SCFA changes. (PubMed)
    (This is a disease model, not normative aging, but it’s a clear post-2020 demonstration of lifespan extension with ACA.)

Insects

4) House cricket (Acheta domesticus)

  • Liao et al., 2026 preprint: ACA did not extend lifespan overall and showed a sex-specific harm signal:

    • Post-treatment lifespan reduced in females (HR ~2.9–3.0 reported). (PubMed)

5) Fruit fly (Drosophila melanogaster)

  • Lüersen et al., 2025: ACA’s lifespan effect depended on diet composition; on certain disaccharide-based high-sugar diets, ACA was associated with shortened lifespan (and delayed development / reduced activity), while effects differed on monosaccharide diets. (PubMed)

What’s missing / notable gaps since 2020

  • I didn’t find strong post-2020 primary lifespan papers where ACA clearly extends lifespan in C. elegans or zebrafish (most zebrafish hits were diabetes/metabolic models rather than lifespan). (ScienceDirect)

Bottom line across model species (since 2020)

  • Best supported pro-longevity evidence remains in mice, especially male-biased effects in UM-HET3 (often cited in post-2020 work) and strong benefits in some contexts like Ndufs4−/− and rapamycin+ACA combos. (PMC)
  • In insects, recent evidence is not pro-longevity and can be harmful depending on sex (crickets) or diet (flies). (PubMed)

If you want, I can turn this into a species × dose × start-age × sex × effect-size table (with the exact reported medians/HRs wherever the paper provides them).

Mixed messaging - comment not from AI

Clearly Acarbose is not a nobrainer whereas Rapamycin appears to be.