### Consensus consistency validation — 9 October 2026

Cutoff: 2026-10-09. Provider artifacts: `consensus-browser-response.md` (verbatim, citation markers preserved), `browser-thread.json`, `consensus-search-raw.json`. Consensus thread: [Metformin effects on human aging](https://consensus.app/search/metformin-effects-on-human-aging/IHRUTYBhR96iwVt7569lEw/), Pro. The browser response correctly surfaced METFORAGING and distinguishes epigenetic biomarkers from healthspan in places. Its cited reviews and lifespan meta-analysis are discovery sources; primary claims below were checked against original reports. Separately, the protocol/preprint search surfaced an important non-peer-reviewed older-prediabetes frailty trial; it is recorded as supplemental, outside the original 61 reports.

#### Primary-source reconciliation

- **PMID 42023167; DOI 10.1016/j.eclinm.2026.103874 (METFORAGING).** Primary randomized pilot: 40 older non-diabetic people living with HIV randomized (19/21); 35 completed 96 weeks. Primary PhenoAge EAA between-group difference −1.02 years (95% CI −5.30 to 3.26; p=.627); no statistically significant between-group effects on epigenetic biomarkers. Original 61-source ledger does not contain this new 2026 paper; provider-discovered and within cutoff, but supplemental only. Small, disease-context pilot with imprecise null does not demonstrate healthspan extension.
- **PMID 33428709; DOI 10.1093/gerona/glaa295 (DPP/DPPOS frailty).** Original primary report already in source ledger, same group as diabetes-prevention follow-up. Frailty findings do not demonstrate broad geroprotection in non-diabetic adults; only selected older prediabetes population and functional frailty outcomes. Consensus’s older review citation to preliminary/unpublished DPPOS information is superseded by consulting this published original report.
- **PMID 36281624; DOI 10.1111/acel.13733 (Parish & Swindell). Secondary source, not a primary comparison.** Consensus-cited systematic review/meta-analysis of animal lifespan studies. It reports heterogeneous effects and no significant overall lifespan extension in mice or nematodes (with subgroup-specific worm signal). Absent from the frozen selected-records ledger, and excluded from the original primary-report count. It is a useful contradictory synthesis for SC2 but does not alter SC3.
- **Supplemental outside-ledger finding: PMID 42620010; DOI 10.64898/2026.07.27.26359055; NCT02570672.** Musi et al., medRxiv preprint (posted 2026-07-28; explicitly not peer reviewed), one-site 2-year randomized trial in older adults with prediabetes/glucose intolerance (145 randomized; 141 ITT). Registered protocol DOI 10.1093/gerona/glz069 specifies Fried frailty status/score as the primary endpoint and deficit-accumulation frailty index (FI) as secondary. Preprint reports no clear positive primary functional result: Fried score worsened in year 1 then stabilized; the year-1 effect was driven by weight-loss item, disappearing when that item was removed. Secondary FI decreased modestly (unadjusted −0.006±0.0026/year, p=.0222; BMI-adjusted −0.005±0.0024/year, p=.0446). This is a plausible narrow functional signal in older adults with dysglycemia, not evidence for the exact broad claim in SC3 (“non-diabetic humans” generally), not replication, and not peer reviewed. It cannot be treated as a primary threshold crossing. Any score recalculation that rests on this preprint would require confidence cap Low under the stated framework rule for non-peer-reviewed work; current score does not rely on it.

#### Claim/score judgement

Preregistered SC3 is broad non-diabetic human aging/healthspan, and the preregistered primary-subclaim aggregation must remain controlling. The prediabetes trial’s narrow secondary FI signal cannot establish broad generalization, while its registered Fried primary endpoint was not clearly positive. METFORAGING is a small biomarker-focused pilot with a wide null estimate; DPPOS is disease-prevention context and not generalized aging. Score remains **Γ / Moderate**; no patch proposed. Original totals remain **61 primary reports, 65 verified claims**; the two new metformin publications are supplemental and do not inflate the original count. If later authorized to recalculate on the preprint, confidence must be Low if the final judgment rests on it; that is not proposed here.

This reconciles provider-discovered citations with original source records and protocols. It is not an independent blinded review. Consensus labels were not used to assign tier or score.

#### Primary record links

[PMID 42023167](https://pubmed.ncbi.nlm.nih.gov/42023167/) · [PMID 33428709](https://pubmed.ncbi.nlm.nih.gov/33428709/) · [PMID 36281624](https://pubmed.ncbi.nlm.nih.gov/36281624/) · [PMID 42620010](https://pubmed.ncbi.nlm.nih.gov/42620010/) · [registered protocol, DOI 10.1093/gerona/glz069](https://doi.org/10.1093/gerona/glz069)


This dated supplement adds source reconciliation to the original review. Its original narrative, registered scope, source inventory and grades remain historical records. No earlier adversarial approval has been verified for this edition at the supplement date. This provenance supplement has no separate approval and is not an independent or blinded adjudication. New sources described here are supplemental and are not counted in the original inventory.