Aging becomes a measurable rate: a pace-of-aging clock predicts dementia (DunedinPACE)
DunedinPACE, a third-generation DNA-methylation "pace of aging" clock, was associated with Alzheimer's diagnosis and prospectively predicted incident dementia (HR 1.27, 95% CI 1.07-1.49; Framingham Offspring, n=2,264), while first/second-generation clocks were not (Sugden, Belsky et al., Neurology, 6 Jul 2022; PMID 35794023). Aging is becoming measurable as a modifiable rate — the substrate for healthspan trials, and a biomarker market that must be policed for hype.
Across ADNI and the Framingham Offspring Cohort, a faster pace of aging by DunedinPACE tracked Alzheimer's diagnosis and worse cognition, and prospectively predicted incident dementia (HR 1.27, 95% CI 1.07-1.49) — where the older Horvath/Hannum/PhenoAge/GrimAge clocks were not consistently associated. It is a functional biomarker of the rate of biological aging, not just a chronological estimate.
Why it matters (spine): a credible, primary-sourced aging biomarker is the measurement layer healthspan interventions need — and the exact place where consumer hype outruns evidence, which is RDKG's hype-vs-verified instrument.
Falsifier: it fails to predict outcomes or to move with real interventions. · Confidence: 0.8 as a predictive biomarker; ~0.45 as a decision-grade surrogate endpoint for "slowing aging" (held at signal-tier for that use).
Provenance: Karen Sugden, Daniel W. Belsky, Terrie Moffitt & Avshalom Caspi (Duke / Columbia; credited), Neurology 2022;99:e1402-e1413, PMID 35794023, DOI 10.1212/WNL.0000000000200898.
So we should use it as an exploratory trial readout in HKG and build the trust layer that separates validated biomarkers from consumer hype.
Use DunedinPACE as an exploratory readout in longevity work, but do not sell consumer "biological age" scores as clinical endpoints — the validated-vs-hyped biomarker line is itself a product.
domains: biomarkers · longevity · epigenetics · aging-clock · health