Do epigenetic clocks differ so much that your result is essentially random?
Different epigenetic clocks sometimes lead to opposing conclusions, so your result does not necessarily say anything absolute. If you want to use a clock to track lifestyle changes, choose one of the newer composite clocks and never let a single measurement carry more weight than your own experienced health.
Two clocks that claim to measure the same thing can produce completely different results. For example, epiTOC2 does a good job of measuring stem cell division (correlation of 0.92 with laboratory measurements), whereas HypoClock barely scores on that exact same task (correlation of 0.30, not significant). When predicting atrial fibrillation, GrimAge and PhenoAge remained significant after correcting for known risk factors, but the Horvath clock did not. And an inflammatory clock detected no accelerated ageing in people living at -60 degrees, whereas earlier research using a different clock did find that. This is not a minor nuance: these are opposing conclusions.
Your result therefore partly depends on which clock you use. There is a good reason for this: each clock weights different aspects of DNA methylation. Mainstream clocks do measure how the average methylation level across the population changes with age, but they miss another aspect of ageing: the way people increasingly diverge from one another as they grow older. They therefore capture only part of the biological ageing process.
As age increases, so does the variation in methylation patterns between individuals. This 'epigenetic drift' means that an individual measurement in an older person inherently contains more noise than one taken in a younger person. Your result is therefore not purely random, but it does carry more uncertainty than it appears to at first glance.
What does this mean if you are considering taking a commercial test? Composite clocks that incorporate multiple health outcomes, such as GrimAge, appear to predict more robustly than first-generation clocks such as the Horvath clock, but even these provide only a snapshot. Whether a lower clock reading after a lifestyle change truly translates into a longer or healthier life has not yet been proven. Research that uses clocks as a measure for diet or sleep interventions largely lacks solid clinical trials, and several of the authors involved have financial interests in companies that sell these tests.
Claims are based on eight studies: two on clock comparisons for specific outcomes (PMID 32580750, 34587750), one longitudinal study involving 135 people (PMID 41482678), one on the Yakutian population (PMID 39769502), one on epigenetic drift (PMID 25913071), two on lifestyle interventions with declared conflicts of interest (PMID 40266024, 35778957), and one meta-analysis on trauma (PMID 29174058).