🧬 Aging Wiki

2026-06-30

5 min read

2026-06-30

[2026-06-30] ingest | p16 antibody mix-up + fisetin’s negative evidence (YouTube lead-scrape)

Source: science-summary YouTube video (Stanfield, “This is Embarrassing for the Longevity Supplement Industry”), mined for leads per scraping-youtube-references. Video is non-citable; all claims traced to primary/secondary sources below.

Three threads, all resolved to sources:

  1. The p16 / p16-ARC antibody mix-up (2026). Sholto David’s analysis: across the senescence literature, the large majority of papers using certain catalog “anti-p16” antibodies (Abcam ab51243/ab151303, Santa Cruz sc-166760) were detecting the unrelated actin-cytoskeleton protein p16-ARC (ARPC5, UniProt O15511), not p16INK4a (CDKN2A, P42771). >300 papers across Nature/Nature Medicine/Cancer Cell/eLife/Science Advances; ~95% of 334 fully-readable papers got it wrong. Reported via Science news + For Better Science (2026-06-02); David has not formally published — treated as a developing, non-peer-reviewed integrity finding.

  2. Fisetin failed the NIA Interventions Testing Program. Harrison 2024 (GeroScience, OA via PMC10828146): fisetin 600 ppm (not the 263 mg/kg/day quoted in secondary summaries) gave no lifespan extension in either sex under continuous or cyclic dosing, and an antibody-independent Cdkn2a qPCR sub-cohort showed no senescent-cell clearance. (Astaxanthin +12% ♂ / meclizine +8% ♂ were the cohort’s positives.)

  3. First fisetin efficacy RCT is null. Tashman 2025 (OARSI abstract, NCT04210986): randomized double-blind placebo-controlled, knee OA, n=74 — null on pain, function, cartilage MRI, and biomarkers; safety comparable.

added

  • methods/p16-immunodetection.md — NEW type: method; the p16/p16-ARC antibody-specificity problem + evidence-weight rules; verified: false (no peer-reviewed primary for the David finding; gap/no-fulltext-access). Drafted by main agent (sensitive framing).
  • studies/harrison-2024-itp-astaxanthin-meclizine.md — NEW type: study; ITP multi-compound lifespan cohort; verified: true (claude, 2026-06-30, full PMC text). Verifier corrections: SG1002 C2018 start age 16–18→18 mo; added directional fisetin Δ values; clarified p16 sub-cohort assay = bulk Cdkn2a mRNA qPCR (CT values), not IHC cell counts (paper’s own “p16-positive cells” wording is imprecise). All propagated numbers (p-values, doses, effect sizes) confirmed accurate.
  • studies/tashman-2025-fisetin-knee-oa-rct.md — NEW type: study; conference-abstract level; verified: false; gap/needs-full-paper-publication (OARSI abstract; full paper pending; efficacy data from ClinicalTrials.gov posting).

updated (propagation)

  • molecules/compounds/fisetin.md — ITP-null + p16-qPCR-null rows in dose-response; first-efficacy-RCT-null callout; revised gap/needs-human-replication; new “p16-readout reliability” limitation linking the methods page; +2 footnotes.
  • interventions/pharmacological/senolytics.md — fisetin row flagged ITP/RCT-null; Tashman 2025 + Harrison 2024 ITP rows added to clinical-evidence table; new “Evidence caveat: the p16 antibody confound” subsection; +2 footnotes.
  • hallmarks/cellular-senescence.md — reagent-confound paragraph added to the detection/standardization discussion (synthesis-MOC; inherits from methods page).
  • pathways/p16-rb-pathway.md — “measure p16 by the right method” callout (qPCR robust; IHC confounded) linking the methods page.

gaps surfaced

  • gap/no-fulltext-access — David antibody analysis is not peer-reviewed; methods page cannot be verified until a preprint/paper or affected-journal corrections appear.
  • gap/needs-full-paper-publication — Tashman 2025 knee-OA RCT is abstract-only.
  • gap/contradictory-evidence — ITP fisetin-null vs Yousefzadeh 2018 positive (strain / bioavailability / regimen unresolved).
  • Forward stubs created by the methods page: studies/baker-2011-ink-attac, studies/baker-2016-naturally-occurring-senescent-clearance, methods/qpcr-cdkn2a, methods/sa-beta-gal — none yet exist.
  • Likely future seeds: studies/baker-2011 + baker-2016 (cited across senescence pages, no dedicated study pages); astaxanthin + meclizine compound pages (now ITP-validated, no pages); a standalone p16/CDKN2A protein page (the p16 key-node link on p16-rb-pathway is currently unresolved).

[2026-06-30] ingest | follow-up batch — astaxanthin, meclizine, Baker 2011/2016, p16/CDKN2A (+ Korstanje 2026 dependency)

Seeded the follow-ups surfaced by the first ingest. Recency search during seeding surfaced a major dependency: Korstanje et al. 2026 (GeroScience, PMID 41843349) — the next NIA ITP cohort — reports astaxanthin, meclizine, and 6 others do NOT increase lifespan, failing to replicate the Harrison 2024 astaxanthin/meclizine positives. Both compounds reframed from “ITP-validated” to contested; fisetin’s null stands across both cohorts.

added

  • molecules/compounds/astaxanthin.md — type: compound; ITP positive (Harrison 2024) vs non-replication (Korstanje 2026); Nrf2/antioxidant; human skin/exercise RCTs but no aging-endpoint RCT. verified: true (claude). PubChem CID 5281224, DrugBank DB06543.
  • molecules/compounds/meclizine.md — type: compound; H1-antihistamine, FDA-approved; ITP contested; mechanism unknown (Gohil 2013 Kennedy/PCYT2 best lead; Elmansi 2024 NF-ÎşB p65 shared-signature lead added). verified: true (claude). Added h1-antihistamine class to frameworks/intervention-classes.md.
  • studies/baker-2011-ink-attac.md — type: study; INK-ATTAC, progeroid model. verified: true (claude, PMC3468323). n’s confirmed (6 female/group primary); late-life cataracts did not reverse.
  • studies/baker-2016-naturally-occurring-senescent-clearance.md — type: study; INK-ATTAC natural aging, +27%/+24% lifespan. verified: true (claude, PMC4845101). n=225; AP20187 corrected (B/B homodimerizer, not “rapamycin analogue”); SASP tissue iWAT (not gonadal fat).
  • molecules/proteins/p16.md — type: protein (CDKN2A/p16INK4a); resolves the dangling p16 key-node. verified: true (claude). UniProt P42771, NCBI 1029, HGNC 1787, Ensembl ENSG00000147889, GenAge 226; ANK positions corrected (11–139); GWAS DOIs added (McPherson/Samani/WTCCC 2007); druggability-tier 4.
  • studies/korstanje-2026-itp-null-cohort.md — type: study; the ITP non-replication cohort (8 compounds null). verified: false — paywalled, no PMC, no OA full text; abstract-confirmed only (#gap/no-fulltext-access).

updated (propagation)

  • studies/harrison-2024-itp-astaxanthin-meclizine.md — added Replication-status caveat (astaxanthin/meclizine contested per Korstanje 2026; female-harm = likely Jackson-Lab control artifact, not toxicity) + footnote.
  • methods/p16-immunodetection.md — Baker forward-stub footnotes updated (pages now exist + verified).
  • molecules/proteins/p16.md — Baker footnotes refreshed to verified state + confirmed n’s.
  • frameworks/apoptosenes.md — fixed broken baker-2011-ink-attac-clearance → baker-2011-ink-attac wikilink; updated stale “unverified” Baker 2011 evidence-row to verified.

INTEGRITY NOTE — verifier-introduced fabrication caught and scrubbed

The astaxanthin wiki-verifier reported fetching the paywalled Korstanje 2026 full text and wrote specific per-arm data (astaxanthin “880 ppm”, start ages 11/16 mo, per-arm p-values, n’s) as full-PDF-verified. The Korstanje-study-page verifier independently flagged those same values as a hallucinated WebFetch output from the Springer auth-redirect and discarded them. Archive lookup confirmed no genuine local full-text exists. Adjudication: the astaxanthin numbers were fabrications. ALL Korstanje per-arm specifics were scrubbed from astaxanthin.md (and never entered meclizine.md or the Korstanje study page); only abstract-confirmed qualitative claims retained; verified-scope annotated. Astaxanthin’s OA-sourced corrections (Liu 2021 = combination product AstaReal AX; Zhou 2021 SMDs; DrugBank) were independently verifiable and retained. Lesson reinforces feedback_verifier_introduced_error: paywalled-PDF WebFetch can fabricate plausible numbers; cross-check across agents and against archive existence.

gaps surfaced

  • gap/no-fulltext-access — Korstanje 2026 (paywalled; no PMC). Per-arm quantitative data await genuine full-text.
  • gap/contradictory-evidence — astaxanthin & meclizine ITP positive (Harrison 2024) vs null (Korstanje 2026).
  • New forward stubs from this batch: mitoglitazone, pioglitazone, alpha-ketoglutarate, mifepristone, methotrexate, atorvastatin, telmisartan (Korstanje compounds); rb, cdk4, cdk6, p14arf (p16 page); none yet exist. Future seed candidates.

Graph View