Antibiotics, gut bacteria and fat in mice

Overview

Cox LM et al.. Cell 2014 · PMID 25126780 · DOI 10.1016/j.cell.2014.05.052

Low-dose penicillin early in life made mice fatter via their gut bacteria.

Does the study support the specific claim? Partly · Methodological quality: moderate

Original publication

Plain-language summary

Scientific in-depth review

1. Full reference

Authors

Cox LM, Yamanishi S, Sohn J, Alekseyenko AV, Leung JM, Cho I, Kim SG, Li H, Gao Z, Mahana D, Zárate Rodriguez JG, Rogers AB, Robine N, Loke P, Blaser MJ

Journal

Cell

Year

2014

Identifier

PMID 25126780 DOI 10.1016/j.cell.2014.05.052 · PMC4134513

Verification

verified – PubMed E-utilities efetch, 2026-10-25

Retracted

no

Corrections

none

Extent checked

full text

Conflicts of interest and funding

Funding: NIH grants R01 DK 090989 and 1UL1RR029893, Diane Belfer Program in Human Microbial Ecology, Knapp Family Foundation, Daniel Ziff Foundation. No competing-interests statement in the available full text.

2. Research question

Does early-life microbiota disruption cause lasting metabolic effects?

3. Study design and population

Study design

Animal study – Series of controlled mouse experiments (C57BL/6J): sub-therapeutic penicillin from birth vs weaning, with/without high-fat diet, early-limited (4/8 weeks) vs lifelong exposure, plus cecal microbiota transfer to germ-free Swiss-Webster mice and a sequential second transfer

Population

Laboratory mice (C57BL/6J, male and female; germ-free female Swiss-Webster recipients)

Country

USA

Period

Not stated (published 2014)

Sample size

Small groups per experiment, e.g. birth/weaning experiment 4-6 per sex and group; longitudinal experiment 9-10 per sex and group; transfer: 3 donors per group pooled, recipients 7 (control) and 8 (LDP); metabolic-cage and gene-expression subsets n=3-5

Inclusion and exclusion criteria

Not applicable (laboratory animals; IACUC-approved protocols)

4. Methods and endpoints

Exposure / intervention

Low-dose penicillin (LDP), 1 ug/g body weight, in drinking water, from birth (via dam) or from weaning; continuous lifelong or limited to first 4 or 8 weeks; normal chow (13.5% kcal fat) or high-fat diet (45% kcal fat)

Comparator

No antibiotic controls; recipients of control microbiota

Endpoints

Body composition (DEXA, MRI), growth rates, PYY/leptin/insulin, hepatic histology and gene expression, ileal immune gene expression and Th17 cells, 16S rRNA microbiota composition

Statistical methods

t-tests and Mann-Whitney U per time point, growth-rate differentials, ADONIS/UniFrac, LEfSe; mean +/- SEM; no correction for multiple testing; no sample-size calculation reported

5. Actual results

Results

LDP from birth: accelerated pre-weaning growth and increased total and fat mass in adult males; weaker effects with weaning start; females mainly increased total mass/bone density. Longitudinal: faster fat accumulation (rate differential 0.13%/week female, p=0.002; 0.15%/week male, p=0.02). With HFD, LDP about doubled fat mass in females by 30 weeks and raised insulin in males (combination only). Early-limited LDP (4 or 8 weeks) plus HFD: females had sustained increases in total, lean and fat mass; in males early changes were lost with ageing. LDP reduced hepatic steatosis in males on normal chow. Germ-free recipients of LDP microbiota gained total mass (0.078 g/day, p=0.01) and fat mass (0.058 g/day, p=0.0012) faster than recipients of control microbiota; in a second sequential transfer, the adiposity difference disappeared. Microbiota recovered after antibiotic cessation while phenotypes persisted.

Absolute risks

not applicable (animal study)

Authors' conclusion

Early-life LDP transiently perturbs the microbiota and is sufficient for sustained effects on body composition; LDP amplifies diet-induced obesity; the altered microbiota, not antibiotics per se, is causal; phenotype transmissibility is lost with successive transfers. Authors state further studies are needed to determine whether these changes are important in humans.

6. Strengths

• Controlled experiments with multiple designs testing timing, diet interaction and causality
• Microbiota transfer to germ-free mice isolates microbiota effect
• Longitudinal microbiota profiling
• Authors report inconsistent and sex-dependent findings

7. Weaknesses

• Mice; very small groups (3-10) with many t-tests and no multiplicity correction
• Sub-therapeutic continuous dosing (growth-promoter model) unlike short human therapeutic courses
• Sex-inconsistent effects (male effects lost with ageing in early-limited experiment)
• Transfer uses different mouse strain; donors only 3 per group, pooled
• Phenotype not maintained in second transfer
• Blinding and randomisation of animals not reported

8. Risk of bias

Orienting tool

No standard tool fits (animal experiments); structured validity judgement (SYRCLE-type items)

Type of assessment

structured assessment based on the full text by a single reviewer (no formal double review)

Full formal assessment

no

Overall risk of bias

some concerns

Rationale

Single-reviewer structured judgement based on the full text; no standard RoB tool applies to animal experiments in this framework; not formal (needs two independent reviewers). Overall changed from not_assessable to some_concerns for internal validity; external validity to human therapeutic courses is low.

Full-text review (25 October 2026): Added dose (1 ug/g body weight), group sizes (3-10), sex-dependent and non-persistent results (male effect lost with ageing; lost in second transfer), funding; structured animal-study validity domains; bias overall changed from not_assessable to some_concerns; methodological quality downgraded to moderate and precision to low.

9. Assessment of the statistical interpretation

Small-group experiments with multiple uncorrected comparisons; effects are directionally consistent for increased mass but differ by sex, tissue and experiment; the germ-free transfer result rests on 3 pooled donors and 7-8 recipients.

10. Relevance to the specific claim

C-ABX-05, C-RES-05.

11. Independent contrary findings

12. Transferability to Germany and Europe

13. Overall methodological judgement

Well-designed mechanistic mouse work showing that continuous low-dose penicillin in early life can alter microbiota and increase adiposity in mice, with microbiota transfer supporting a causal role in mice. It models agricultural growth promoters, not human therapeutic courses, and effects were sex-dependent and not maintained after sequential transfer.

14. Open questions

Whether short therapeutic human antibiotic courses produce comparable, lasting microbiota-mediated adiposity in humans.

15. Linked fact checks

• Antibiotics make people (especially children) gain weight by altering gut bacteria, as they do in fattened poultry; early-childhood antibiotic courses raise obesity risk. – Studies supporting the claim
• Antibiotics were (and are) used in poultry production to make chickens grow faster. – Studies supporting the claim

Use by the Tayebat archive

Note: according to the archive, the studies were compiled by the archive operators. Whether Dr. Elawady himself cited this study is not established.

What the archive says about the study

Low-dose penicillin in a critical developmental window altered the murine microbiota and caused lasting metabolic and adiposity changes - transferable to germ-free hosts; AR headline: 'mechanistic evidence of antibiotics' role in obesity'.

A. What was actually studied?

Mouse microbiota and metabolism.

B. What did the authors find?

LDP (1 ug/g, continuous or first 4-8 weeks) increased mass/fat in mice, more in males with birth exposure and in females with HFD; microbiota transfer reproduced faster weight gain in germ-free mice; effect lost on a second transfer.

C. What conclusions do the authors themselves draw?

Microbiota causal in mice.

D. What does the archive derive from it?

Mechanistic proof for human obesity.

E. Is this derivation scientifically justified?

Correct as mouse mechanism (and for agricultural growth promotion, C-RES-05); transfer to human therapeutic courses and human obesity is not established, as the authors themselves state.

Adequacy of the interpretation (by the archive)

partly correct but over-interpreted