DunedinPACE (Pace of Aging Clock)

DunedinPACE measures the pace of aging rather than a single age estimate. Daniel Belsky and colleagues built it from the Dunedin Study, a birth cohort followed from 1972-73, by tracking within-person change across 19 biomarkers (organ-system function including cardiovascular, metabolic, renal, immune, dental, and pulmonary measures) at ages 26, 32, 38, and 45, then trained a DNA methylation signature to predict each person's rate of decline [1]. The claim: instead of estimating how old someone's cells look, DunedinPACE estimates how fast someone is currently aging, expressed as years of biological decline per calendar year (1.0 = normal pace) [2][6].

The 2022 eLife validation paper reported that a faster DunedinPACE predicted earlier mortality, incident chronic disease, and cognitive and physical decline in independent cohorts including the Framingham Heart Study and the Health and Retirement Study, and that it outperformed earlier clocks including GrimAge in some of these prospective analyses [1]. Because it captures pace of change longitudinally in its training data, proponents argue it is more sensitive to short-term intervention effects than clocks trained purely as age estimators, a comparison drawn against first-generation and second-generation epigenetic clocks including GrimAge [3][4].

How to measure it: a blood draw undergoes methylation array analysis (Illumina EPIC), and the published Belsky et al. algorithm computes the pace score. Values are typically reported as a ratio, where 1.0 means aging at the population-average pace, values below 1.0 mean slower-than-average aging, and above 1.0 mean faster. Commercial providers (TruDiagnostic) offer it as part of standard longevity panels.

How to intervene on it: DunedinPACE is newer than Horvath or GrimAge, so fewer intervention trials report it. A CALERIE caloric-restriction sub-study (2% average restriction sustained over two years) reported a slower DunedinPACE in the restricted group compared to controls, one of the first RCT-level signals that a lifestyle intervention can shift a pace-of-aging clock, though the CALERIE trial was designed around caloric restriction generally rather than DunedinPACE specifically. Bryan Johnson's Blueprint protocol has reported a DunedinPACE around 0.69 through TruDiagnostic testing, framed publicly as roughly 31% slower aging [5]; this is an n=1 self-report using a commercial test, not a controlled study, and should be read as anecdote rather than evidence a specific protocol works in general.

Critics note DunedinPACE was trained on a single New Zealand birth cohort that is overwhelmingly of European descent, raising questions about generalizability to other ancestries and environments. It is also newer than GrimAge and Horvath, so it has less independent replication across diverse populations, and the biomarkers it was trained to track (like some inflammatory and metabolic markers) already have established links to lifestyle factors, meaning improvements could partly reflect known risk-factor changes rather than novel information.

The plain takeaway: DunedinPACE is a promising, mechanistically distinct pace-of-aging measure with early trial-level support from caloric restriction data, but it remains less validated across diverse populations than older clocks, and individual results, including widely publicized ones, should be treated as single-person anecdotes rather than proof a specific regimen slows aging.