Comparing the responsiveness of epigenetic aging clocks
Researchers comparing the responsiveness epigenetic aging biomarkers found that generation 2+ DNA methylation clocks—especially DunedinPACE and PCGrimAge—showed the strongest, most consistent signals across longevity interventions. The Nature Medicine analysis of 51 human studies suggests these tools could speed trial readouts, once validated as surrogate endpoints.
Key Takeaways
- A new analysis pooled 51 longitudinal human intervention studies to test how DNA methylation (DNAm) aging biomarkers respond to longevity-related treatments.
- Pharmacological interventions produced the strongest DNAm responses and were among the few categories linked to significantly reduced epigenetic age.
- Generation 2+ clocks, notably DunedinPACE and PCGrimAge, were the most responsive and concordant; they still need surrogate-endpoint validation.
- TNF-reducing therapies and two Mediterranean diet approaches met the authors’ consistency checks across studies.
- Participant health status mattered: many biomarkers responded more in disease groups than in healthy cohorts.
Human lifespan trials can take decades. That lag is why aging researchers want proxies—biomarkers that show whether an intervention is working long before mortality data arrive. Epigenetic aging clocks, built from DNA methylation patterns, are among the most discussed candidates. Coverage of this work sits alongside other updates in our Longevity & Biohacking hub.
Why do researchers need responsive epigenetic aging biomarkers?
Even popular clocks are not yet validated as surrogate endpoints for human longevity trials, and evidence on how they move under real interventions has been thin. Lead author Raghav Sehgal argued that if such biomarkers eventually predict long-term health, scientists could judge anti-aging therapies in years or months instead of decades.
The team treated “DNAm biomarkers” as an umbrella for 16 epigenetic clocks plus 94 other methylation markers, then scored responses across lifestyle, pharmacological, supplement, and medical-procedure interventions drawn from existing longitudinal studies, as reported by Lifespan.io.
Which interventions moved DNA methylation clocks the most?
Pharmacological approaches—examples in the dataset included metformin, rapamycin, semaglutide, ketamine, and anti-TNF therapy—produced stronger DNAm biomarker responses than other categories and were the only group with significantly larger effect sizes. Pharmacological and lifestyle interventions were also the categories that showed significantly reduced epigenetic age.
The authors hypothesize that drugs may hit inflammation and metabolic pathways such as TNF, AMPK, and mTOR. For consistency, they required that an intervention shift DNAm biomarkers of a given generation in the same magnitude and direction within a study, and that a second study of the same intervention confirm those biomarkers. Anti-TNF therapies used in inflammatory arthritis and IBD, plus two Mediterranean diet types in healthy people, met both bars.
Are all epigenetic aging biomarkers equally reliable?
No. Concordance tests favored generation 2+ biomarkers. Sensitivity also varied: DunedinPACE responded strongly to lifestyle interventions but was moved by fewer pharmacological interventions than other second-generation clocks. Several markers reacted more in disease populations than in healthy ones; only DunedinPACE showed similar response levels in both settings.
Using Generation X (GenX) DNAm biomarkers, the researchers also probed organ- and pathway-level effects—for example, lower lung system scores after smoking cessation and metformin’s links to inflammatory, brain-related, and metabolic pathways—so trials can catch changes composite clocks might average out.
Bottom line from the Nature Medicine paper: prioritize generation 2+ tools such as DunedinPACE and PCGrimAge for future trials, while recognizing they still require formal validation as clinical surrogates.