Background

Measuring the rate of aging and verifying the efficacy of longevity-promoting substances are central tasks in the field of geroscience. Because human lifespan spans several decades, directly confirming the efficacy of anti-aging drugs or therapies through mortality or lifespan extension requires astronomical time and cost. To overcome this challenge, epigenetic clocks that estimate biological age by tracking DNA methylation (DNAm) changes have been proposed as an alternative.

However, the various DNAm clocks developed to date have not been systematically validated for consistent reactivity to aging-modulating therapies. Academic efforts have largely been limited to asserting the utility of specific clocks based on small-scale clinical trial results. Standardized criteria to overcome technical discrepancies between different experimental environments and aging clock models, and to verify whether methylation changes can function as surrogate endpoints reflecting extended healthspan, have not yet been established. This has acted as a bottleneck, delaying the clinical entry of promising anti-aging candidates.

Key Findings

A collaborative research team led by Dr. Albert Higgins-Chen and Dr. Morgan Levine from the Yale School of Medicine developed an open platform, TranslAGE, to evaluate the validity of methylation biomarkers and conducted a large-scale benchmarking study. The researchers compiled data from 51 longitudinal intervention studies, ranging from dietary interventions, exercise, and calorie restriction to semaglutide or tumor necrosis factor (TNF) inhibitor treatments. The resulting database, TranslAGE-Response, includes analysis results from 3,128 human blood samples.

Based on this, an analysis was conducted comparing 16 representative DNAm clocks and 94 methylation-based biomarkers. The analysis revealed that second- and third-generation clocks, designed to predict survival or health status, responded significantly more sensitively to therapeutic interventions than first-generation clocks, which merely predict chronological age. Specifically, DunedinPACE and PCGrimAge showed the most consistent and superior trends in reducing aging rates across various health management and drug intervention studies.

Conversely, some promising anti-aging candidates failed to elicit consistent responses. Senolytic drugs, which inhibit cellular aging, showed variable results across studies, and rapamycin or nicotinamide riboside (NR) did not induce statistically significant changes in aging rates on methylation clocks. Notably, the reactivity of methylation biomarkers was more pronounced in clinical cohorts with underlying conditions compared to healthy adult groups.

Implications and Outlook

This benchmarking study is significant in that it transparently compared the reliability of epigenetic clocks that had previously been used indiscriminately. By providing a standardized evaluation tool, it offers clear guidelines for future anti-aging drug developers on which indicators to adopt in clinical trials.

Nevertheless, the research team cautiously noted that a decrease in methylation clock values does not necessarily guarantee a fundamental delay in biological aging or an extension of lifespan. It remains possible that biochemical values may have improved temporarily due to short-term drug use or lifestyle changes. In other words, current clocks serve as intervention-sensitive endpoints but have not yet reached the level of being perfect surrogate endpoints that fully prove mortality reduction. The challenge remains to demonstrate, through follow-up research, the long-term correlation between methylation clock changes and actual clinical benefits.