Gut & MicrobiomeResearch PaperOpen Access

Wiping out the gut microbiome late in life eased frailty in aged male mice

In aged male mice, antibiotics or germ-free rearing shifted frailty toward young values, but motor coordination and recognition memory did not improve.

Thursday, November 19, 2026 0 views
Published in Gut Microbes Rep
A laboratory mouse with graying fur on a researcher's gloved hand beside cages in an animal research facility

Summary

Frailty measures how vulnerable an older body is, beyond its calendar age. Researchers compared young (10-week) and aged (72-week) male mice. Aged mice had higher frailty scores, worse motor coordination and a restructured gut microbiome. The team then depleted the microbiome in aged mice in two ways: about a month of broad-spectrum antibiotics late in life, or germ-free rearing from birth. Both moved frailty scores toward young values. Motor coordination and recognition memory did not shift with either approach. The frailty benefit coincided with calmer liver inflammation signatures and attenuation of aging-associated blood-lipid patterns, including the reduced polyunsaturated-to-monounsaturated fat ratio. This is mouse work in males only, with groups of about 10. It points to gut-liver signaling as a mechanism worth pursuing, not a treatment, and it does not show that people should take antibiotics.

Detailed Summary

Frailty tracks biological aging better than birthdays do, and in older people it correlates with gut microbiome composition. Whether the microbiome causes frailty has been unclear. This study tested that question in male C57BL/6J mice. It reports mechanism (microbiome, lipid and liver signatures) and function (frailty and behavior). Lifespan was not measured.

The team compared young (10-week) and aged (72-week) specific-pathogen-free mice, n = 10 per group. Aged mice had higher mean frailty scores (Wilcoxon p < 0.0001), heavier ceca (p < 0.05) and shorter rotarod fall latency (p < 0.001). Five of 19 frailty items rose significantly: fur color loss, kyphosis, coat deterioration, alopecia and whisker loss. Novel object recognition did not differ between young and aged mice. Open-field speed showed a nonsignificant decrease (p = 0.08).

The aged microbiome was restructured rather than depleted. Community composition differed (PERMANOVA p = 0.001), while richness, Shannon diversity and the Firmicutes/Bacteroidetes ratio did not. Microbiome distance tracked physio-behavioral decline (Mantel r = 0.52, p < 0.0001). Several taxa rose with age and mostly tracked with higher frailty: Lactobacillus intestinalis, Ligilactobacillus murinus, Alistipes finegoldii, Bacteroides acidifaciens and Parabacteroides goldsteinii. Dubosiella newyorkensis and B. thetaiotaomicron fell with age and tracked with lower frailty. Lower abundance of microbial coenzyme A and phosphopantothenate biosynthesis genes accompanied higher frailty.

To test whether frailty depends on the microbiome, the authors used two depletion models in aged mice. One was late-life broad-spectrum antibiotics (0.5 g/L over a roughly 30-day course). The other was germ-free rearing from birth. Per the abstract, both moved frailty toward young values, but motor coordination and recognition memory did not shift. The authors interpret this as selective microbiome regulation of aging phenotypes.

Per the abstract, reduced frailty coincided with attenuated hepatic inflammatory signatures, including lower expression of acute-phase reactants, and altered fatty acid and bile acid metabolism. Depletion also attenuated aging-associated serum lipid patterns, including the reduced polyunsaturated-to-monounsaturated fatty acid ratio that is associated with poorer metabolic health. Multi-omics integration implicated lipoteichoic acid, microbial phytoestrogen metabolites and loss of immunomodulatory microbes in hepatic senescence, impaired bioenergetics and inflammation. The supplied text ends during the young-versus-aged microbiome results. The depletion, lipidomic, liver and hippocampal findings above therefore come from the abstract, and their detailed statistics could not be checked.

Caveats are substantial. All mice were male, groups were small, and lifelong germ-free development differs from late-life antibiotic exposure. The multi-omic links are correlative hypotheses, and nothing was tested in humans.

Key Findings

  • Aged (72-week) vs young male mice, n = 10 each: higher frailty score (p < 0.0001) and shorter rotarod fall latency (p < 0.001); novel object recognition did not differ
  • The aged microbiome was restructured (PERMANOVA p = 0.001) without changes in richness, Shannon index or Firmicutes/Bacteroidetes ratio; microbiome distance tracked phenotype decline (Mantel r = 0.52)
  • Late-life antibiotics (about 30 days) and germ-free rearing both shifted frailty scores toward young values in aged male mice
  • Null on function: per the abstract, neither depletion approach shifted motor coordination or recognition memory toward young values
  • Per the abstract, depletion attenuated aging-associated serum lipid patterns, including the reduced polyunsaturated-to-monounsaturated fatty acid ratio
  • Per the abstract, reduced frailty coincided with lower hepatic acute-phase reactant expression and altered fatty acid and bile acid metabolism; lipoteichoic acid and microbial phytoestrogen metabolites were implicated
  • Two Lactobacillus species rose with age and tracked with frailty, contrary to their health-promoting reputation in colitis; no lifespan data were collected

Methodology

Male C57BL/6J mice were compared young (10 weeks) vs aged (72 weeks), n = 10 per group, alongside aged antibiotic-treated (0.5 g/L over a roughly 30-day course) and aged germ-free groups. Phenotyping included a 19-item mouse clinical frailty score, open field, novel object recognition and rotarod. Fecal shotgun metagenomics, serum lipidomics, and liver and hippocampal transcriptomics followed. Statistics in the supplied text include Wilcoxon tests, PERMANOVA, a Mantel test and MaAsLin3 models. Lifespan was not assessed.

Study Limitations

Only male mice were studied, and groups were small (n = 10 per group in the young vs aged comparison). Germ-free rearing from birth is not equivalent to late-life antibiotic depletion. The multi-omic links are correlative hypotheses. No lifespan or human data were reported. The supplied text ends during the young-versus-aged microbiome results, so the depletion-arm, lipidomic, liver and hippocampal statistics, and any conflict-of-interest statement, could not be verified.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: