Epigenetic clocks

node

Statistical predictors of chronological age built from DNA methylation levels at selected CpG sites. Steve Horvath's 2013 multi-tissue clock used 353 sites and predicted chronological age across most human tissues with a median error of a few years; Hannum's blood-based clock appeared the same year. Second-generation clocks — PhenoAge, GrimAge — were trained on mortality and clinical measures rather than on chronological age and predict health outcomes better, and DunedinPACE was trained on the longitudinal rate of change of multiple organ-system measures, so it estimates a pace rather than an age. The gap between a clock reading and chronological age is associated with mortality risk. What no clock has yet shown is that it is a valid surrogate: that changing the reading changes the outcome. Until an intervention is shown to move both the clock and the clinical endpoint in the same direction in the same trial, a clock is a correlate of ageing rather than a measure of it, and the commercial tests sold on that basis are ahead of their evidence.

A node is not a place. Drawing it on a map would assert something about the world that no stored fact supports.