Everyone knows a 60-year-old who moves like someone ten years younger, and a 45-year-old who looks worn out. Epigenetic clocks are an attempt to put a number on that hunch. They do it by reading chemical tags on DNA rather than counting birthdays.
Under the hood, the tags in question are methyl groups, small chemical attachments that sit at particular spots along the genome and help decide which genes are switched on or off. They leave the genetic code itself untouched. Some sites pick up methylation steadily over the decades; others lose it. An epigenetic clock is an algorithm that reads the level at a chosen set of those sites and folds them into one figure, an estimated age.
From calendar age to health
The approach goes back to a 2013 model that used 353 methylation sites across many tissue types and predicted chronological age with striking accuracy. Steve Horvath, a professor at the University of California, Los Angeles, is credited with inventing aging clocks.
That first generation was trained to do one thing: guess how old you are. The second generation was trained on health and mortality data instead, and the output changed character. It stopped saying how many years you have lived and started hinting at how well you are ageing. When a person’s biological age runs ahead of the calendar, researchers call it age acceleration. Pooled evidence now links it, across many studies, to age-related disease and earlier death.
A further refinement is the pace of aging. It estimates the rate at which someone is ageing right now, a speedometer rather than an odometer.
Beyond methylation
DNA is not the only place to look. Nearly every protein in the body carries glycans, chains of sugar molecules, and the pattern changes with age. On immunoglobulin G, the most abundant antibody, the sugars galactose and sialic acid gradually drop away, which nudges the antibody toward a more inflammatory state. Glycans shift fairly fast and can respond to lifestyle changes over months, so they suit tracking whether an intervention is doing anything.
Proteins themselves can serve as the readout too. Novo Nordisk drew blood from 10,052 people, half on a placebo, and measured them with proteomic clocks, which use levels of key proteins to predict age and the risk of dying. Eli Lilly ran a smaller study with epigenetic clocks. Both drugmakers reported slower biological ageing in people taking their weight-loss drugs, around “two to three years” overall according to a Novo project leader, and as much as 4 years on a clock built from heart proteins. Horvath called two years “a pretty strong effect.” The companies presented the data at a conference on ageing research, and MIT Technology Review reported it. Nobody yet knows whether the same would hold in healthy people.
That matters for the field. Until now it was uncertain whether the clocks were true biomarkers. A drug with broad, well-demonstrated benefits that moves the readout is the kind of evidence their makers have wanted for a decade.
What the numbers cannot do
Men’s Fitness, in a piece published on Oct 25, 2026, was blunt about the limits. These are correlations across populations, not diagnoses. They do not tell an individual what is wrong, the way a blood pressure reading does. Results can differ between samples and laboratories, and two clocks applied to the same person can disagree. Most direct-to-consumer tests have not been validated to guide medical decisions. And whether the clocks measure ageing or merely track alongside it is still an open question.
The levers that slow biological ageing in the data are the familiar ones: regular physical activity, not smoking, enough sleep and keeping cardiometabolic health in check. A methylation score replaces none of that, nor a physical exam. Talk to your doctor before acting on any biological age result.
Sources
- Your Real Age Isn't on Your Birth Certificate: How DNA Clocks and Glycans Track Cellular Aging
- Weight-loss drugs show signs of slowing biological aging, say drugmakers | MIT Technology Review
