Review 1 of “Metformin Reverses Aging at the Cellular Level: A Systematic Review of Tissue-Specific Effects Across the Hallmarks of Aging” (v2)

My recommendation for version 2 is major revision requiring substantial reconstruction. The paper covers a worthwhile question, but its methods and cited evidence do not establish that metformin reverses cellular aging. Several verified citation errors also undermine confidence in the synthesis.

The tissue-by-tissue organization is useful, and the discussion acknowledges exercise interference, observational confounding, and uncertainty in people without diabetes. Those qualifications, however, are not consistently reflected in the title, abstract, conclusions, or evidence ratings.

My principal concerns and requested changes are:

1. Audit the references and their connection to each claim.

   I confirmed substantive mismatches:

   Reference	Use in the manuscript	What the identifier actually identifies
   22	Metformin delaying hepatic senescence through DNA repair	A study of Wnt signaling in vascular smooth muscle cells. Record
   44 and 60	Endothelial telomerase activation and neuronal telomere preservation	An ecological-footprint commentary, The Shoe Fits, but the Footprint is Larger than Earth. Article
   77	The TAME trial	A study of life-course stressors and functional limitations. Record

   These errors affect the scientific support for the claims. Verify all 82 entries against publisher or bibliographic records, then check whether each source actually supports the associated statement. Remove or qualify unsupported claims. My checks were targeted; this is not an exhaustive reference audit.

2. Provide reproducible systematic-review methods or relabel the article.

   The Methods provide search concepts but omit exact search strings, search dates, screening counts, exclusion reasons, reviewer procedures, and risk-of-bias assessment. “Validated preclinical models” is not an operational eligibility criterion.

   Supply the search and selection record, an included-study table, and a documented appraisal method. If those records do not exist, describe the work as a narrative review. A meta-analysis is not essential, but a traceable synthesis is.

3. Rebuild Table 1 around identifiable evidence.

   The table is labeled as evidence across human tissues, although the narrative combines animal experiments, cultured cells, epidemiology, and human interventions. Its “Strong” and “Very Strong” ratings lack reproducible criteria.

   There are also internal inconsistencies: muscle telomere preservation receives a positive rating even though the text says direct muscle effects have not been studied.

   For every tissue–hallmark judgment, identify the supporting studies, species, population, dose, duration, endpoint, and limitations. Distinguish pathway modulation from improved function and demonstrated reversal of an age-associated phenotype.

4. Remove the unsupported inference of widespread aging reversal.

   The manuscript reports no defined aging-reversal endpoint, quantitative cross-tissue transcriptomic comparison, or operational definition of a youthful expression signature. Changes in AMPK, mTOR, inflammation, or autophagy do not by themselves establish reversal of aging.

   Relevant human transcriptomic evidence should be discussed directly. For example, MILES studied 14 older adults with impaired glucose tolerance in a crossover trial, with six weeks each of metformin and placebo and biopsies from muscle and subcutaneous adipose tissue. That supports a limited mechanistic investigation; it cannot establish rejuvenation across seven human tissues. Study dataset and design

   Replace the title with something such as “Metformin and Aging-Related Pathways: A Review of Tissue-Specific Evidence.”

5. Correct the aging-hallmarks framework.

   The manuscript incorrectly describes the original nine hallmarks and then uses an eleven-category hybrid. The 2013 framework included altered intercellular communication; the 2023 expansion contains twelve hallmarks, including disabled macroautophagy, chronic inflammation, and dysbiosis. Framework publication

   Choose and accurately describe one framework, correct its references, and reassess the “5 of 9” conclusion. Explain any deliberate exclusions.

6. Present mechanisms and contrary human findings more carefully.

   AMPK activation should not be presented as an established universal explanation. Experimental work demonstrates AMPK-independent suppression of hepatic glucose production, while other work supports AMPK-dependent effects under different exposure conditions. Discuss dose, model, and tissue explicitly. Foretz et al., Cao et al.

   The muscle synthesis also needs stronger treatment of randomized human evidence: Konopka and colleagues found that metformin inhibited mitochondrial adaptations during aerobic training in older adults. This should influence the table’s strongly favorable muscle rating. Randomized trial

7. Correct the treatment of TAME and document the claimed cohort evidence.

   The abstract says the review integrates TAME trial data, but the manuscript presents no TAME results and later describes its outcomes prospectively. AFAR’s official resource describes fundraising and launch plans. Treat the trial as prospective context unless a verified results publication is supplied. I could not verify the manuscript’s stated registration identifier. Official TAME resource

   Likewise, identify the UK Biobank analyses claimed in the abstract and report their populations, comparisons, outcomes, and confounding controls—or remove that claim.

A defensible revised conclusion would be that metformin modulates several pathways implicated in aging, with effects that depend on tissue, exposure, and experimental context; the evidence presented does not establish widespread reversal of human cellular aging.