Epigenetic clocks

Epigenetic clocks are biological aging biomarkers derived from patterns of DNA methylation across the genome. The core discovery, made independently by Steve Horvath and Gregory Hannum in 2013, is that the methylation state of a specific set of CpG sites across the genome predicts chronological age with striking accuracy — and, more importantly, that deviation from this predicted age (biological age relative to chronological age) predicts mortality, disease risk, and the effects of interventions such as caloric restriction.

The Horvath clock, trained on methylation data from many tissue types, predicts age with a median error of approximately 3.6 years across a wide range of human tissues. Subsequent clocks — Hannum, PhenoAge, GrimAge — have been trained on different outcomes (chronological age, physiological aging markers, mortality) and make somewhat different predictions. What all share is the identification of methylation drift as a systematic, calibrated process whose deviation from expectation carries health information.

Epigenetic clocks have become central instruments in aging research and longevity interventions, including the growing field of biological age reversal that uses methylation clock readouts as outcome measures for interventions ranging from dietary protocols to partial cellular reprogramming. Whether clock age measures the cause of aging or merely a correlated readout of underlying aging processes remains unresolved — a distinction with significant consequences for whether clock reversal constitutes actual rejuvenation or merely cosmetic molecular bookkeeping.== The Systems View of Epigenetic Aging ==

From a systems perspective, the methylation sites tracked by epigenetic clocks are not merely passive markers but regulatory nodes in a gene regulatory network. The Horvath clock's CpG sites are enriched in promoters of developmental genes — the same genes that orchestrate cellular differentiation during embryogenesis. This suggests that aging is not a separate process from development but a continuation of the same regulatory logic, running in reverse or decelerating. The clock does not measure time; it measures the cumulative drift of a dynamical system whose attractor states are progressively lost.

The implication is that biological age, as measured by methylation clocks, may be a property of the system's information content rather than its chronological history. A cell with a young epigenetic profile may be informationally closer to its developmental attractor, while an old profile has wandered into a region of state space from which return is thermodynamically improbable. This reframes the rejuvenation question: partial reprogramming does not reverse time; it recompresses the system's regulatory information, pushing it back toward a younger attractor. Whether this is genuine rejuvenation or merely a transient metastable state is the central experimental question of the next decade.