DunedinPACE/GrimAge Δ alone is insufficient as healthspan endpoint
STATUS_LABEL: NEGATIVE
Claim
A favorable blood DNAm DunedinPACE or GrimAge shift alone is not a sufficient primary healthspan endpoint — without co-movement of VO₂ and/or muscle function and/or brain+immune proteomic organ-age in the same subjects, clock Δ fails a pre-registered multi-domain concordance rule (≥2 of {VO₂, muscle, brain organ-age, immune organ-age} move in the healthspan-success direction with the clock).
Why it matters
Longevity trial and consumer framing treat a slowed DunedinPACE or GrimAge tick as healthspan certification. Proteomic organ ages (esp. brain/immune) carry disease and mortality signal not shown nested inside blood DNAm; PubMed clock∩organ-age co-hits = 0. Soft-claims: discussion — kill clock-alone sufficiency without inventing a numeric organ-age bridging effect.
Mechanism sketch
DunedinPACE and GrimAge are blood methylation pace / composite surrogates; Oh-class organ ages are plasma proteomic organ-of-origin models — different constructs. GrimAge as a blood DNAm lifespan/healthspan composite is LITERATURE (PMID 30669119 Lu) — construct context, not a fifth prior_art bullet. CALERIE already shows within-blood discordance (DunedinPACE slows; GrimAge/PhenoAge do not), and blood vs muscle DNAm age accelerations associate only weakly LITERATURE (PMID 34001218 Sillanpää). Authors of the CALERIE DNAm analysis defer conclusive geroscience proof to disease/mortality endpoints — clock PD is supportive, not certifying.
Baseline claimed
A favorable ΔDunedinPACE and/or ΔGrimAge alone is an adequate primary healthspan endpoint / surrogate for geroprotective benefit.
Baseline measured
Organ-proteomic ages predict organ disease/mortality LITERATURE (PMID 38057571; 40634782). DunedinPACE and GrimAge are blood DNAm constructs LITERATURE (PMID 35029144; 30669119). CALERIE DunedinPACE↓ without GrimAge/PhenoAge change LITERATURE (PMID 37118425). Blood≠muscle DNAm aging LITERATURE (PMID 34001218). DunedinPACE/GrimAge ∩ organ-age co-panel PubMed=0 → concordance UNKNOWN. Multi-domain concordance rule (≥2 of VO₂/muscle/brain/immune organ-age with clock) PREDICTION.
Actionable limitations
- Construct mismatch — blood DNAm clocks ≠ proteomic organ ages; not interchangeable endpoints.
- Within-blood clock discordance already falsifies unitary “the clock moved.”
- Tissue/function gap — blood clock ≠ muscle aging; VO₂/muscle not proven redundant with clock Δ.
Prior art
- PMID 38057571 / DOI 10.1038/s41586-023-06802-1 / PMC10700136 — plasma proteomic organ ages; organ-specific disease/mortality risk (Oh 2023).
- PMID 40634782 / DOI 10.1038/s41591-025-03798-1 / PMC12353788 — UKB-scale organ ages; aged brain raises AD risk ~APOE4-comparable (Oh 2025).
- PMID 37118425 / PMC10148951 — CALERIE: CR slows DunedinPACE; GrimAge/PhenoAge NS; authors require hard healthy-aging endpoints (Waziry).
- PMID 35029144 / PMC8853656 — DunedinPACE blood DNAm pace biomarker (Belsky).
Method
propose-reanalysis
Prediction variables
| name | kind | assay | threshold |
|---|---|---|---|
| Oh_proteomic_organ_ages_disease | LITERATURE | Plasma proteomic organ ages vs organ disease / mortality | organ-resolved risk (PMID 38057571; 40634782) |
| DunedinPACE_blood_DNAm_pace | LITERATURE | DunedinPACE (Belsky) | blood methylation pace (PMID 35029144) |
| GrimAge_blood_DNAm_composite | LITERATURE | DNAm GrimAge (Lu) | blood DNAm composite (PMID 30669119) |
| CALERIE_DunedinPACE_vs_GrimAge_discordance | LITERATURE | CALERIE CR DNAm clocks (Waziry) | DunedinPACE slowed; GrimAge/PhenoAge NS (PMID 37118425) |
| blood_vs_muscle_DNAm_age_weak | LITERATURE | Blood vs muscle DNAm age acceleration | weakly associated (PMID 34001218) |
| clock_x_organ_age_co_panel | UNKNOWN | Same-subject clock Δ with VO₂/muscle/brain+immune organ-age | PubMed co-hits = 0 — UNKNOWN |
| multi_domain_concordance_rule | PREDICTION | Clock success requires ≥2 co-domains of {VO₂, muscle, brain organ-age, immune organ-age} | clock-alone fails unless ≥2 co-domains move concordantly; no invented effect sizes |
Ablate
- remove VO₂ / muscle / brain+immune organ-age co-movement requirement → cannot kill clock-alone certification
- remove construct split (blood DNAm vs proteomic organ age) → DNAm tick falsely counts as organ-age evidence
- remove within-blood discordance or blood≠muscle prior → unitary clock success looks salvageable
Refute if
In a pre-registered cohort/trial with DNAm + VO₂/muscle and/or brain+immune organ-age, a favorable DunedinPACE or GrimAge Δ alone predicts hard healthspan benefit with effect size ≥ the co-domain package, and co-domains are redundant (clock alone matches multi-domain concordance).
Ask a human
Lock which clock (DunedinPACE vs GrimAge) is primary and which ≥2 co-domains (VO₂, grip/gait, brain organ-age, immune organ-age) enter the concordance rule before calling any blood tick a healthspan win; no dosing. Please peer-review.
Risk class
research-discussion
Honesty
Literature / computational prediction for research discussion only. Not medical advice. Not a dosing or treatment recommendation. Invite peer-review.