🧬 Aging Wiki

2026-06-03

20 min read

[2026-06-03] log

[2026-06-03] update | exercise.md — “Training frequency and body recomposition” section (ad-hoc)

Folded a new section into exercise answering a recurring practical question: how training frequency (e.g., daily running vs. alternating/recovery days) relates to the rate of body recomposition. Tangentially aging-relevant — low visceral adiposity + preserved lean mass protect against inflammaging, insulin resistance, and sarcopenia.

  • Framing: recomposition = two decoupled levers (fat loss vs. lean gain) responding to different stimuli; frequency is a weaker lever than volume, energy balance, and protein.
  • Evidence synthesized (7 new citations, DOI/PMID-sanity-verified):
    • Willis 2012 STRRIDE-AT/RT (doi:10.1152/japplphysiol.01370.2011, PMID 23019316) — aerobic/combined drive fat-mass loss; resistance adds lean mass, aerobic does not; combined doesn’t beat aerobic-alone on fat loss despite ~2× time.
    • Schoenfeld 2016 frequency meta (doi:10.1007/s40279-016-0543-8, PMID 27102172) — at equated weekly volume, RT frequency has no independent hypertrophy effect (volume-mediated).
    • Wilson 2012 concurrent-training interference meta (doi:10.1519/JSC.0b013e31823a3e2d, PMID 22002517) — interference scales with endurance frequency/duration; running > cycling.
    • concurrent2025 sequence/recovery meta (doi:10.3389/fspor.2025.1692399); ACSM 2026 RT Position Stand (doi:10.1249/mss.0000000000003897).
    • Mechanism clarification (user-prompted): Melanson 2009 (doi:10.1152/japplphysiol.00958.2009, PMID 19833807) — when energy balance is maintained, exercise does not induce negative fat balance; 24-h fat oxidation tracks energy balance, not exercise substrate use (“fat-burning zone” myth). Swift 2018 (doi:10.1016/j.pcad.2018.07.014, PMID 30003901) — exercise fat loss is energy-expenditure-driven, often blunted by compensatory intake. Added a paragraph reframing the STRRIDE aerobic fat-loss advantage as an energy-deficit effect (aerobic ~12 mi/wk = larger energy cost; ad-lib diets), NOT a fat-specific property of jogging; the surviving non-calorie modality effect is compositional (resistance preserves lean → higher fat fraction of weight lost).
  • Gap surfaced: dose-response-unclear — no RCT directly compares daily vs. alternating endurance, volume-matched, with recomposition as the outcome.
  • Linked existing protein-intake from cross-references.

[2026-06-03] verify | exercise.md recomposition section — PDF cross-check (wiki-verifier + main-agent precision fix)

Ran wiki-verifier on the new section against primary PDFs (scoped to the 6/3 additions; rest of page already verified 2026-05-08). verified: true retained, scope addendum expanded.

  • Read end-to-end: Willis 2012 STRRIDE-AT/RT (PMC3544497) and Melanson 2009 (PMC). Abstract-only (closed-access): Schoenfeld 2016, Wilson 2012, Swift 2018. Metadata-only: concurrent2025, ACSM 2026.
  • Willis 2012 — exact kg deltas inserted into [^willis2012], “not-PDF-checked” caveat removed. RT (n=44): fat mass −0.26 kg (p=0.429, NS), lean +1.09 (p<0.0001). AT (n=38): fat −1.66 (p=0.001), lean +0.10 (p=0.613, NS). AT/RT (n=37): fat −2.44 (p<0.0001), lean +0.81 (p=0.001). All directional claims confirmed. Body refined: combined was not statistically superior to aerobic-alone on fat/body mass despite ~2× time (combined’s fat point estimate was numerically larger, −2.44 vs −1.66, but NS) — fixed a potential apparent contradiction with the footnote numbers.
  • Schoenfeld 2016 — error corrected, then precision-refined. My original “no independent frequency effect / purely volume-mediated” framing was wrong (verifier catch). But the verifier’s correction over-attached the p-value: the significant ES 0.49 vs 0.30 (p=0.002) is the volume-confounded binary comparison; the volume-equated per-muscle-group analysis “could not be carried out” (underpowered), so twice>once at matched volume is the authors’ qualitative conclusion, not a powered result. Main agent re-pulled the abstract (efetch) and rewrote body + footnote to state this precisely, and scoped the finding to resistance hypertrophy (NOT transferable to aerobic/jogging frequency — which is the actual question).
  • Melanson 2009 footnote corrected: full population (LS/LT/OS, n=27), whole-room calorimetry; key finding = fat balance significantly more positive on exercise day (p<0.01) when energy replaced. Body claim was already correct.
  • concurrent2025 retyped “meta-analysis” → “semi-systematic review” (Europe PMC title).
  • Supersession: no post-2021 paper supersedes STRRIDE’s core conclusions for non-diabetic overweight/obese adults (Gong 2026 Maturitas combined-training MD −1.74 kg fat in sarcopenic obesity is consistent). literature-checked-through left at 2026-05-08.
  • Lesson reinforced: verifier abstract-only corrections can over-claim a statistic’s scope — cross-checked against the abstract directly (cf. verifier-introduced-error memory).

[2026-06-03] ingest | sex-differential aging — batch 1 (female reproductive foundation)

User-prompted coverage campaign: an audit found the wiki’s aging coverage was reproductive-endocrine-blind (defaulting to a male-ish neutral), with female-specific biology especially thin. Plan = 5 seeding batches + sex-bias patches to existing disease pages + a capstone MOC, all verifier-validated. Batch 1 seeded 6 atomic pages in parallel (all verified: false, banners on, queued for verifier):

  • added: tissues/ovary.md — “fastest-aging organ” framing; Faddy-Gosden + Wallace-Kelsey follicle-depletion models; cohesin/aneuploidy, mtDNA, BRCA1-reserve link; perimenopausal endocrine cascade; ovarian-rejuvenation flagged investigational.

  • added: phenotypes/menopause.md — STRAW+10 staging; inhibin B→AMH→FSH→estradiol cascade; systemic downstream consequences (bone/CVD/vasomotor-KNDy/GSM/skin/metabolic/cognitive); surgical-vs-natural; grandmother hypothesis. (26 footnotes.)

  • added: molecules/compounds/estradiol.md — E2; PubChem CID 5757, ChEMBL135, DB00783; ESR1/ESR2 + GPER mechanism; bone/CVD/skin/brain/metabolic aging; timing-hypothesis; cancer-risk duality; R25 recency run (literature-checked-through 2026-06-03). Integrated Melville 2025 (Lancet Healthy Longevity, null overall MHT-dementia) vs older observational discordance.

  • added: molecules/compounds/progesterone.md — P4; PubChem 5994, CHEMBL103, DB00396; PR-A/PR-B + allopregnanolone/GABA-A; bioidentical-vs-synthetic-progestin distinction foregrounded (WHI used MPA); R25 run. Added progesterone-receptor-agonist mechanism class to intervention-classes.

  • added: cell-types/oocytes.md — CL:0000023; meiotic-arrest longevity; cohesin non-renewal aneuploidy model; mtDNA/SAC/BRCA1-ATM; oogonial-stem-cell controversy (#gap/contradictory-evidence).

  • added: cell-types/granulosa-cells.md — CL:0000501; FOXL2; two-cell/two-gonadotropin aromatase model; AMH/inhibin source; GC senescence + atresia; oocyte-GC coupling.

  • propagation: updated reproductive-system MOC (populated key-organs/tissues/cell-types/phenotypes; moved seeded items out of “planned”); resolved the #gap/unsourced sex-hormone-axes note on altered-intercellular-communication (now links menopause/ovary/estradiol/progesterone); added reciprocal links on osteoporosis + brca1. Removed a seeder-invented icd-10-secondary: frontmatter field from menopause (schema discipline — moved N95.x cluster to body prose).

  • audit correction: the initial Explore audit wrongly reported esr1.md/esr2.md existed — they do not; nor do gper/cyp19a1. These (the mechanistic core of estrogen signaling) are now top-priority stubs folded into the next wave.

  • gaps surfaced: estrogen-signaling proteins (ESR1/ESR2/GPER/aromatase), AMH/FSH/LH, systemic HRT, vasomotor-symptoms, GSM — all queued in later batches.

[2026-06-03] ingest | sex-differential aging — batch 2 (estrogen signaling + reproductive endocrine proteins)

Six type: protein pages (all verified:false, banners, queued for verifier). Resolves the estrogen-signaling stubs dangling off batch 1’s estradiol.

  • added: molecules/proteins/esr1.md — ERα (UniProt P03372, gene 2099). Dominant bone-protective ER; anchored on the two natural experiments (Smith 1994 ESR1-mutation man + Morishima 1995 aromatase-deficient man — both osteoporotic/unfused-epiphyses despite normal T → estrogen governs bone in both sexes). Endothelial ERα/CVD; ER+ breast-cancer antagonistic pleiotropy. druggability-tier 1.

  • added: molecules/proteins/esr2.md — ERβ (UniProt Q92731, gene 2100). Granulosa/prostate/colon/CNS distribution; modulates/opposes ERα proliferation; Esr2-KO subfertility. tier 2.

  • added: molecules/proteins/gper.md — membrane GPER/GPR30 (UniProt Q99527, gene 2852). Rapid non-genomic estrogen signaling (cAMP/EGFR/PI3K); G-1/G-15 probes; tamoxifen/fulvestrant are GPER agonists. tier 3. literature-checked-through 2026-06-03.

  • added: molecules/proteins/cyp19a1.md — aromatase (UniProt P11511, gene 1588). Rate-limiting androgen→estrogen step; the premenopausal-ovary→postmenopausal-adipose estrogen-source shift (why postmenopausal estrogen tracks adiposity); AIs as iatrogenic estrogen-deprivation model + AI bone loss. tier 1.

  • added: molecules/proteins/amh.md — anti-Müllerian hormone (UniProt P03971, gene 268). TGF-β ligand via AMHR2→SMAD1/5/8; granulosa-secreted follicle-pool gatekeeper; premier ovarian-reserve biomarker (Nelson 2023 SR n=28,858; precision-for-individual-menopause-prediction caveat framed). R25 run, literature-checked-through 2026-06-03.

  • added: molecules/proteins/fsh.md — FSHB (UniProt P01225; CGA alpha P01215 noted). Rising FSH = earliest endocrine signature of the transition; contested non-gonadal direct-action frontier given real space with both sides — bone (Sun 2006), fat (Liu 2017), brain/AD (Xiong 2022), all largely mouse/one-lab-cluster, needs-human-replication; Sauvé 2024 cautious-translation discordance framed. R25 run, literature-checked-through 2026-06-03.

  • propagation: updated reproductive-system MOC (moved the 6 into “Seeded members”). Bug fix: corrected a seeder-introduced wrong DOI on batch-1 tissues/ovary.md — [^dewailly2014] was 10.1093/humupd/dmu002 (resolves to an unrelated Han/O’Malley endometriosis paper) → corrected to 10.1093/humupd/dmt062 (the actual Dewailly 2014 AMH review; title/authors/volume were already correct). Caught by the AMH seeder cross-checking the same citation.

  • new stubs surfaced: [[lh]], [[amhr2]], [[inhibin]], [[testosterone]], plus pathway stubs [[hypothalamic-pituitary-gonadal-axis]], [[steroidogenesis]], [[gnrh-signaling]], [[estrogen-signaling-pathway]].

  • verifier note: ESR1 Smith-1994/Morishima-1995 BMD figures, FSH non-gonadal Nature/Cell papers (Sun/Liu/Xiong — Xiong PDF local), and AMH Durlinger-1999 gatekeeper claim are HIGH-priority PDF checks.

[2026-06-03] verify | sex-differential aging — batches 1+2 (all 12 foundation pages → verified:true)

User chose to verify the foundation before seeding further. Ran 12 wiki-verifier passes (2 waves of 6) against primary PDFs. Every page required corrections — strong vindication of verify-before-propagate; the sonnet seeders introduced a high rate of fabricated/wrong specifics. All 12 flipped to verified:true (2026-06-03, claude), most at partial scope (closed-access primaries abstract-only). Notable corrections:

Batch 1:

  • tissues/ovary.md → true. Entire follicle-count table was wrong — seeder conflated old “~7 million total germ cells” with non-growing-follicle counts; corrected to Wallace-Kelsey NGF model (~300k peak, ~295k birth, ~180k age 13, ~16k age 35; model mean menopause 49.6 y). Titus 2013 cell-type flipped (granulosa→oocytes, r=0.60). STRAW FSH threshold >40→>25 IU/L random draw. Supersession flagged: Winship 2024 EBioMedicine (PMID 39084071) — AMH unreliable as reserve marker in BRCA1 carriers.
  • phenotypes/menopause.md → true. STRAW+10 stage names wrong throughout (fixed to Fig 2); Wilson 2019 oophorectomy HR misrepresented — 1.81 was only the no-MHT subgroup, overall HR 1.02 (NS); Brincat 1983 “30% collagen loss in 5 yr” was fabricated (paper is cross-sectional, 48% more collagen in HRT users) — claim removed, unsourced.
  • molecules/compounds/estradiol.md → true. ERβ→ERα in osteoclasts; Campos 2024 (CEE specifically, not route, raises BP); Melville 2025 sign-inverted — seeder claimed it supported the timing-window hypothesis; paper found NO significant timing effect (#gap/contradictory-evidence); Chlebowski issue/pages wrong; mendelsohn 2002→2005.
  • molecules/compounds/progesterone.md → true. Wright 2014 ProTECT III fabricated CI (0.69–1.31 → actual relative benefit 0.95, 95% CI 0.85–1.06; metric was “relative benefit” not OR); Asi 2016 study count (2 cohorts pooled, not 3); Rodrigues 2026 DOI resolved (108998); Fournier CIs added.
  • cell-types/oocytes.md → true. Shang 2024 omitted the null live-birth result (OR 1.05, NS) — efficacy overstated; NAC wasn’t in that meta; n 2,617→2,218; Chiang/Zielinska ages + fragmentation % corrected.
  • cell-types/granulosa-cells.md → true. FOXL2 >97%→97% (86/89, Shah 2009); two false Camaioni 2022 citations removed (paper doesn’t cover the cited claims); citation issue/pages fixed; van Rooij n 120→119.

Batch 2:

  • molecules/proteins/esr1.md → true. Smith 1994 + Morishima 1995 natural-experiment numbers confirmed against PDF/abstract (load-bearing); Gavin 2009 n 48→33; GenAge ID was 216, seeder wrongly set null.

  • molecules/proteins/esr2.md → true. DBD 96→95%, LBD 53→55%; false “luteinization” claim (Emmen 2005: luteinization normal in βERKO); ovulation “could not be induced”→“reduced”; Muka “cardioprotective”→“antiangiogenic”; fabricated rat-model claim removed.

  • molecules/proteins/gper.md → true. Gurrala model rats→mice + mRNA/protein discrepancy nuance; unsourced ADAM10/17 sheddase attribution removed; Filardo 2002 HCN/EPAC misattribution fixed.

  • molecules/proteins/cyp19a1.md → true. SOFT/TEXT P<0.05→P<0.001; trial mislabeled SOFT→TEXT+SOFT combined; Yu 2024 n badly understated (~2,000→4,951 GWAS + 205,427 UKB).

  • molecules/proteins/amh.md → true. Durlinger 1999 DOI was wrong (dmt .7179 → resolved to unrelated Hall/McDonnell ER paper; corrected to .7204, PMID 10579345); AMH peak age corrected (~25); OCP suppression 20–30%→14–55%; Nelson 2023 pages 262-282→327-346; BRCA1 caveat added.

  • molecules/proteins/fsh.md → true. Sun 2006 genotype error — estrogen-independence evidence is FSHβ+/− eugonadal mice, not FSHR heterozygotes (the crux of the claim); Liu 2017 both sexes (G_i/cAMP/Ucp1 mechanism); Gera 2022 over-attributed anti-fat/neuro claims (those are Liu/Xiong); Xiong δ-secretase→AEP; contested-mouse framing + gap tags confirmed airtight.

  • propagation: no leaked wrong values found in other pages (grep sweep clean — these are new pages, no study/ pages cite them yet). Minor: cyp19a1 carries an undocumented known-interactors: frontmatter field (harmless; next lint).

  • two wrong DOIs total this campaign (ovary/dewailly fixed during batch-2 propagation; amh/durlinger fixed in verification) — reinforces seeder-brief-doi-unreliable pattern.

[2026-06-03] ingest | sex-differential aging — batch 3 (male foundation + female menopausal symptoms)

Seven pages (all verified:false, banners, queued for wave-C verifier). Builds the male axis (deliberately contrasted with the abrupt female axis) + the two missing female clinical-symptom phenotypes.

  • added: molecules/compounds/testosterone.md — PubChem 6013, CHEMBL386630, DB00624. AR-agonism + aromatization-to-E2 arm; gradual partial decline vs abrupt menopause; bone benefit substantially via E2 (cyp19a1/esr1 natural experiments); TRAVERSE (Lincoff 2023) CV-safety + AF signal, T-Trials, Hudson 2023 meta; “Low T” over-treatment controversy. R25, literature-checked-through 2026-06-03. Added androgen-receptor-agonist mechanism class to intervention-classes (completes estrogen/progesterone/androgen receptor-agonist cluster).

  • added: phenotypes/andropause.md — late-onset hypogonadism; EMAS strict criteria (Wu 2010: ≥3 sexual symptoms + total T<11 nmol/L → only ~2% prevalence, vs “Low T” marketing); the menopause-contrast is the page’s spine; functional (reversible) vs intrinsic hypogonadism.

  • added: tissues/testis.md — continuous SSC renewal vs fixed oocyte pool (central contrast); the paternal-age effect (Kong 2012 ~2 de novo mutations/yr; selfish spermatogonial selection FGFR2/3; neurodevelopmental associations); Sertoli/blood-testis-barrier decline.

  • added: cell-types/leydig-cells.md — CL:0000178; male steroidogenic cell (theca analog); LH→cAMP→StAR cascade; age decline in number + per-cell capacity + LH-responsiveness → low T despite compensatory LH; INSL3 as Leydig-capacity biomarker; stem-Leydig biology. Petersen 2015 (number-stable) vs older histology flagged contradictory-evidence.

  • added: molecules/proteins/lh.md — LHB (UniProt P01229, gene 3972) + shared CGA alpha; LHCGR; rises in both sexes with gonadal failure; contested gonadotropin-hypothesis-of-AD (Casadesus) kept brief + gap-tagged. literature-checked-through 2026-06-03.

  • added: phenotypes/vasomotor-symptoms.md — KNDy-neuron/NK3R mechanism (estrogen withdrawal → KNDy hypertrophy → narrowed thermoneutral zone); SWAN median ~7.4 yr duration; VMS as CV/sleep risk marker beyond QoL; NK3R-antagonist breakthrough (fezolinetant SKYLIGHT, elinzanetant OASIS). R25, literature-checked-through 2026-06-03.

  • added: phenotypes/genitourinary-syndrome-menopause.md — GSM (Portman 2014 consensus term); estrogen-dependent urogenital-sinus tissue atrophy + vaginal-microbiome/pH shift; chronic-progressive vs VMS-remitting contrast; vaginal estrogen/prasterone/ospemifene; laser-device evidence contested (Jaber 2025 null sham-RCT). literature-checked-through 2026-06-03.

  • propagation: reproductive-system MOC reorganized into Female axis / Male axis / shared-gonadotropins; new stubs surfaced (sertoli-cells, spermatogonial-stem-cells, androgen-receptor, insl3, testosterone-replacement-therapy).

  • queued for wave-C verification (testosterone TRAVERSE/T-Trials numbers, testis Kong-2012 mutation rate, leydig Luo-2001 StAR/P450scc %, VMS Rance/SWAN, GSM Portman/Santen — all HIGH/medium).

[2026-06-03] verify | sex-differential aging — batch 3 (wave C; all 7 → verified:true)

7 wiki-verifier passes vs primary PDFs. Again every page needed corrections (sonnet seeder error rate stays high). All flipped verified:true (2026-06-03, claude), mostly partial scope.

  • molecules/compounds/testosterone.md → true. TRAVERSE MACE HR wrong direction 1.07→0.96 (95% CI 0.78–1.17); n ~5,200→5,246; treatment duration conflated w/ follow-up (21.7 mo tx, 33 mo f/u); AKI signal omitted (added); T-Trials n 788→790 + primary-endpoint nulls clarified; Hudson 2023 IPD-vs-aggregate-n conflation + QoL/IIEF effect sizes. Supersession: Braga 2025 meta (arrhythmia RR 1.53) corroborates AF signal.
  • phenotypes/andropause.md → true. Prevalence 2%→2.1% (63/2,966); 2nd EMAS symptom corrected (“decreased frequency of sexual thoughts”, not “libido”); enrolled-vs-analysis n clarified. Supersession flagged: EAU-2025 threshold now <12 nmol/L (vs EMAS <11) + Yeap 2026 mortality-anchored thresholds.
  • tissues/testis.md → true. SSC 23-divisions/yr misattributed (Kong→Goriely); spurious per-father division arithmetic removed; Kong de-novo rate precision 2.01/yr (94% variance); FGFR3 figures corrected; stale citation counts struck. Supersession: Mol Psychiatry 2025 +1.50 DNMs/yr corroborates.
  • cell-types/leydig-cells.md → true. Luo 2001 StAR/P450scc figures recontextualized (steady-state Western, not max-LH-stim); Anand-Ivell 2024 design (review of FAMAS+EMAS) + INSL3 threshold framing (≤0.4 ng/mL = LCI threshold, range 0.4–2.3).
  • molecules/proteins/lh.md → true. Mature LHB chain 141→121 aa; Zhang 2024 mechanism confabulated — wiki claimed amyloid/tau reduction; paper actually shows EGR1/cholinergic (fabrication, corrected); Sims 2023 design (genetic KO only, female/age-specific); GnRH frequency maintained not declined (Vermeulen); post-menopausal LH fold-rise overstated + wrong citation (Anawalt is male-only); Casadesus COI (Voyager) added; Bowen 2015 leuprolide RCT (null) added.
  • phenotypes/vasomotor-symptoms.md → true. Prevalence 70–80%→70%; Torres 2026 DOI resolved (10.1038/s41574-026-01247-8); Thurston follow-up median 19y (not 22) + p=0.026; SKYLIGHT-1 issue/pages + n 522→527; OASIS issue/pages; elinzanetant “hepatotoxicity” reversed — OASIS reported NO liver toxicity (corrected).
  • phenotypes/genitourinary-syndrome-menopause.md → true. REVIVE figures badly wrong — n 3,520→3,768; four QoL/HCP-discussion figures wrong/inverted (corrected to 70% dryness, 61.6% discussed, 45.2% on treatment); Santen absorption nuance (modern LC-MS/MS postmenopausal range vs outdated <20 pg/mL RIA); Jaber 2025 secondary outcomes (some improved, some worsened); n=36, halted early. Supersession flagged: HA-injection RCT (Maturitas 2025, n=116) strengthens non-hormonal options.
  • propagation: andropause already had correct TRAVERSE n (no inherit); genomic-instability doesn’t cite paternal-age (candidate future add, not error); menopause VMS/GSM cross-refs de-staled (“when this page is seeded” → live pointers). No leaked numeric errors.
  • third confabulated mechanism this campaign (lh/Zhang-2024, after estradiol/Melville sign-inversion + earlier fabrications) — abstract-only seeding of closed-access mechanism papers is the recurring failure mode; verifier full-text/abstract re-read catches it.

[2026-06-03] ingest+verify | sex-differential aging — batch 4 (distinctively-sexed mechanisms; seeded + wave-D verified, all → verified:true)

Four synthesis/mechanism pages, seeded then same-day verified:

  • processes/x-chromosome-inactivation.md → true. XCI biology (XIST/PRC2/escape) + the aging quartet: mosaic-advantage (female-longevity link), age-related XCI erosion (Liu 2023 — escape variability increases, not mean drift), autoimmunity (TLR7/CD40LG escape; Dou 2024 Cell XIST-RNP-as-autoantigen — verified vs local PDF, added ΔRepA-Xist + SJL/J-vs-C57BL/6J strain dependence + DM/SSc/SLE patient scope), skewed-XCI/CHIP. Carrel-Willard ~15–25% escape confirmed. 2025 preprint correctly labeled.
  • hypotheses/female-longevity-advantage.md → true (evidence-aggregating MOC). Phenomenon (Zarulli 2018 — survives even famine/epidemic/slavery, infant-mortality-driven; Lemaître 2020 — 18.6% median across 101 mammal species, 7.8% human, senescence-RATE not consistently sex-biased) + 8 candidate mechanisms each cross-linking atomic pages (XCI, epigenetic clocks/Horvath 2016, estrogen, iron, immune, telomere, mother’s-curse countervailing, behavioral). Correction: Trinidad slavery dates 1772–1783→1813–1816 (1772-73 was the Swedish famine); Horvath cerebellum-null nuance added.
  • hypotheses/mothers-curse.md → true (conceptual-frame). Maternal-mtDNA-inheritance → male-harming variants escape purifying selection. Evidence-for (Innocenti 2011 Drosophila sex-sieve; LHON 37%/13% M/F penetrance) vs strong evidence-against (Cayuela 2023 128-mammal: not supported, wild-only; Watson 2022 copepod: harmed females; Kochar 2025; Edmands 2024 “patchy”). Author order + LHON n corrected. key-evidence frontmatter repointed to studies/ stubs (main agent fixed a mis-pathed molecules/proteins link first).
  • processes/reproductive-aging-tradeoffs.md → true (empirical counterpart to disposable-soma-theory). Parity-longevity U-shape (confounded); late-last-birth→longevity (Sun/Perls 2015 OR 2.08); pregnancy biological-age acceleration + postpartum reversal (Poganik 2023); parity-telomere contested (Houminer-Klepar 2023 meta non-sig). Corrections: Ryan 2018 clock (Hannum+Horvath→Horvath only); Poganik cohort n (54 enrolled vs per-trimester 9/22/20 + chip-limited postpartum cohort); broken footnote ref fixed. Supersession noted: Xu 2026 NHANES (supports, not supersedes).
  • stubs surfaced for future seeding: iron, ferroptosis, xist (set is-noncoding-rna:true), mitochondrial-dna, clonal-hematopoiesis, + study stubs (zarulli-2018, lemaitre-2020, innocenti-2011, cayuela-2023, watson-2022, nielsen-2024).
  • these 4 belong in the forthcoming sex-differences-in-aging capstone MOC (not the organ-focused reproductive-system MOC).

[2026-06-03] ingest+verify | sex-differential aging — batch 5 (sex-bias patches to disease pages) + capstone MOC

Batch 5 — patched 5 EXISTING verified disease pages with a sourced ## Sex differences section each (added, not rewritten; flagged pending in verified-scope, then verified + flag cleared). All remain verified:true:

  • phenotypes/alzheimers-disease.md — ~⅔ female; APOE ε4×sex (Neu 2017: women OR 4.37 vs men 3.14 at 65–75, P=.002); perimenopausal-window/bioenergetic hypothesis (Mosconi 2017, Mishra-Brinton 2022); tau-sex (Coughlan 2026, Li 2026); FSH (Xiong 2022, mouse). Verifier fix: Snyder 2016 issue 10→11.
  • phenotypes/cardiovascular-aging.md — premenopausal protection → post-menopausal convergence (El Khoudary 2020 AHA); HFpEF/INOCA female predominance (Lam 2019); plaque erosion-vs-rupture (Sato 2022). Verifier fix: “~decade later” softened to “several years” (AHA wording) + removed an unsourced mean-age.
  • phenotypes/skin-aging.md — estrogen-dependent dermal collagen; correctly rejects the unsupported “30%/5yr” figure (Brincat 1983 is cross-sectional, ~48% more collagen in HRT users; Brincat 1987 ~1–2%/yr); Rittié 2008 topical-E2 collagen stimulation in sun-protected-not-photoaged skin (n=70). Verifier fix: Brincat 1987 vol/issue/pages.
  • phenotypes/frailty.md — the male–female health–survival paradox (women higher FI, lower mortality; Ahrenfeldt 2019 OR 1.56; Shi 2014 FI submaximal 0.52♀/0.44♂; Cohen 2018). Verifier fixes: Ahrenfeldt pages 1099-1109→1025-1036 + CI added.
  • phenotypes/immunosenescence.md — stronger female immunity vs autoimmunity trade-off (Klein-Flanagan 2016 ~80% autoimmune patients female; TLR7/CD40LG X-escape); Márquez 2020 men greater epigenomic immune-aging. Verifier fix (fabrication): n=298→172, method “CITE-seq”→ATAC-seq/RNA-seq/flow, age range — patch-agent had invented the method+n.

Capstone — frameworks/sex-differences-in-aging.md (type: framework, navigational MOC, no verified block). Slices the atomic layer by sex: §1 female-longevity phenomenon, §2 female axis, §3 male axis, §4 shared gonadotropins, §5 distinctively-sexed mechanisms, §6 sex-bias in 6 disease pages, §7 open gaps. Registered in index (new “Biological-axis overlay”) + frameworks/README.

Phase-1 complete: 24 new atomic pages + 5 disease-page patches + 1 capstone MOC, all verified against primary sources. Seeder error rate stayed high throughout (every page corrected; ≥4 fabricated mechanisms/methods, 2 wrong DOIs, multiple sign/genotype/n errors) — verify-each-batch was essential and caught all of it. New mechanism classes added: estrogen/progesterone/androgen-receptor-agonist. See project-sex-differential-aging-campaign memory for the remaining-gap list (iron/ferroptosis, XIST, sertoli/spermatogonial cells, HRT/TRT interventions, PCOS/endometriosis/Turner).

[2026-06-03] ingest+verify | sex-differential aging — batch 6 (hormone-replacement interventions; seeded + verified, all → verified:true)

User-requested residual-gap wave; user flagged a prior that hormone therapies may be net “pro-aging” — seeders/verifiers briefed to synthesize evidence NEUTRALLY (disease vs biological-aging endpoints; timing hypothesis; formulation/route). 4 class-level intervention pages:

  • interventions/pharmacological/hormone-replacement-therapy.md → true. Systemic MHT. WHI saga (Rossouw 2002 CEE+MPA vs Anderson 2004 CEE-alone), timing-hypothesis trials (ELITE/Hodis 2016, DOPS/Schierbeck 2012, KEEPS-Cog null), Manson 2017 18-yr mortality, Melville 2025 dementia-null. Bottom line: not a validated geroprotector, not uniformly pro-aging — timing/formulation-dependent symptom+bone therapy with ~neutral mortality. Verifier corrections: ELITE progesterone dose 400mg→45mg (9-fold), ELITE CIMT placebo-vs-effect confusion, DOPS HR 0.52→0.48, Manson CEE-alone 1.00→0.94, fabricated KEEPS “p=0.63” removed, VTE HR 2.13→2.11.
  • interventions/pharmacological/testosterone-replacement-therapy.md → true. T-Trials (Snyder 2016)/Budoff 2017 plaque/TRAVERSE (Lincoff 2023)/Hudson 2023 meta. Bottom line: real bounded benefits for genuine hypogonadism; broad “Low-T” anti-aging use unsupported + CV/erythrocytosis risk; not a geroprotector. Verifier corrections: Hudson n 49-RCT/5993→17-study/3431-IPD (testosterone page already correct, no fix), Braga 43→23 RCTs, Bhasin 2026 primary endpoint biochemical-recurrence→sexual-activity, unsourced TRAVERSE secondary HRs removed.
  • interventions/pharmacological/aromatase-inhibitors.md → true. AIs as a clean estrogen-DEPRIVATION model (accelerate bone loss/osteoporosis — pro-aging on that axis — yet life-saving in ER+ breast cancer). ATAC/BIG 1-98/ABCSG-16/IBIS-II/EBCTCG 2022. New aromatase-inhibition class. Verifier corrections: bhatnagar1993 DOI resolved to unrelated paper (fixed → Bhatnagar 1990 + Haynes 2003 for human data), ATAC HR 0.87→0.83, IBIS-II citation, BIG 1-98 fracture% (Rabaglio 2009).
  • interventions/pharmacological/selective-estrogen-receptor-modulators.md → true. Tissue-selective ER ligands — the attempt to decouple estrogen’s bone/lipid benefits from cancer risk (partial success). MORE/STAR/RUTH/PEARL/NSABP-P1. New selective-estrogen-receptor-modulation class. Verifier corrections: RUTH VTE HR 2.17→1.44, STAR endometrial RR 0.55→0.62, PEARL breast HR 0.21→0.19, MORE dose-stratification; “SMART” acronym likely fabricated (flagged in body).
  • propagation: grep-checked — no inherited wrong values elsewhere (DOPS/ATAC/bhatnagar corrections contained to new pages). interventions-by-hallmark MOC altered-intercellular-communication gap now filled by these 4. New mechanism classes: aromatase-inhibition, selective-estrogen-receptor-modulation.
  • synthesis takeaway (for user): none of the 4 hormone-therapy classes is a validated longevity/geroprotective drug; the “anti-aging hormone” marketing is unsupported. But “net pro-aging” is too simple: MHT (early-initiation) and TRT (true hypogonadism) have neutral-to-favorable mortality + real symptom/bone benefit; AIs are the clearest “pro-aging” case (deliberate estrogen removal → accelerated bone/possibly vascular aging) yet justified by cancer survival; SERMs partially decouple benefit from risk. The honest frame is timing/indication/formulation-dependent risk–benefit, not a blanket pro- or anti-aging verdict.

[2026-06-03] ingest+verify | sex-differential aging — batches 7-9 (conditions + organs/cells + mechanisms; all seeded + verified → verified:true)

Final residual-gap wave. 13 new pages seeded + same-day verified (one planned page reconciled away — see below).

Batch 7 — female conditions (phenotypes): polycystic-ovary-syndrome (metabolic-aging acceleration + later-menopause/AMH paradox), endometriosis (estrogen-dependent + ovarian-cancer-subtype risk), premature-ovarian-insufficiency (early-estrogen-deprivation natural experiment), turner-syndrome (45,X / X-dosage natural experiment). Batch 8 — organs + cells: tissues/uterus, tissues/breast, tissues/prostate, cell-types/sertoli-cells (CL:0000216), cell-types/spermatogonial-stem-cells (CL:0000020), cell-types/theca-cells (CL:0000503). Batch 9 — mechanisms: molecules/metabolites/iron (FIRST type:metabolite page — schema adapted from compound, escalated for formalization), processes/ferroptosis, molecules/proteins/xist (lncRNA, is-noncoding-rna:true). The planned mitochondrial-dna page was NOT created — processes/mtdna.md + processes/mtdna-heteroplasmy.md already exist (verified); deleted the redundant seeded draft and added a mitochondrial-dna alias to mtdna.md so mitochondrial-dna resolves.

Verifier corrections (every page needed them; FOUR inverted findings this wave):

  • PCOS: Pedroso 2015 telomere finding inverted (claimed shorter TL → actually NS, no difference); Forslund 2022 menopause-timing claim false (claimed +2yr → NS p=0.65); Ryu 2021 n 46,000→6,811; La Marca DOI wrong (#gap/wrong-doi, contained — not in amh.md).
  • endometriosis: Barnard 2024 superficial-peritoneal claim inverted (claimed no risk → actually aHR 2.82); Kvaskoff CIs corrected.
  • POI: Roeters van Lennep 19→10 studies; Karamitrou I²=0% POI-subgroup context; Blümel pages; added Liu 2023 HF/AF.
  • Turner: Stochholm CI 2.35-3.45→2.18-3.55; Bondy pages.
  • uterus: Brinton n 160→313; Al-Hendy n; Wilson issue.
  • breast: Figueroa 2014 wrong cohort (Nurses’ Health Study → Komen Tissue Bank) + finding misattribution + TDLU-count direction error.
  • prostate: SASP cytokine list corrected (→IL-1α/IL-8/FGF-7/IGF-1); Berry decade-rows flagged; TRAVERSE added.
  • sertoli: Paul-Robaire 2013 cell-type misattribution (BTB proteins measured in spermatocytes, not Sertoli); Sakai 2018 deletion-claim unsupported.
  • SSC: Nie 2022 framing inverted (claimed SSCs drive inflammation → paper: SSC changes modest, somatic cells age); Maher 2018 method (WGS→targeted ultradeep).
  • theca: Nelson 1999 fabricated fold-changes removed; Burger 2002 + Thompson 1993 DOIs corrected; Thompson content misattributed.
  • iron: Zarulli rebuttal misrepresented (corrected to infant-mortality argument); Dixon over-attribution; brain-iron-rate sourcing; deferiprone PD trials added.
  • ferroptosis: Bersuker pages; PUFA-PE/ACSL4 attributions reassigned (Yang 2014 → Kagan/Doll 2017).
  • xist: XIST size 17kb→19kb human; “80+“→81 proteins.

propagation: fixed a 404 DOI on the verified x-chromosome-inactivation page (Zhao 2008 1149219→1163045, caught by XIST seeder) + added human/mouse XIST-size distinction; cross-page grep clean (wrong DOIs/inverted findings contained to own pages). Updated reproductive-system MOC (added uterus/breast/prostate organs, theca/sertoli/SSC cells, the 4 conditions, the 4 interventions) and the sex-differences-in-aging capstone MOC (folded all into §2/§3/§5; §7 open-gaps trimmed to proteins/pathways/sub-phenotypes/study-stubs).

CAMPAIGN COMPLETE. Total: 39 new atomic pages + 5 disease-page patches + 1 capstone MOC, all verified against primary sources. New type: metabolite precedent (iron) + 3 new mechanism classes (estrogen/progesterone/androgen-receptor-agonist) + aromatase-inhibition + SERM classes. Recurring lesson (now overwhelming): sonnet seeders fabricate at a high rate — ~every page corrected, multiple inverted findings, ~5 wrong DOIs, fabricated CIs/endpoints/methods. Verify-each-batch is non-negotiable for this workflow.

Graph View

Backlinks