A Yale-led analysis in Nature Medicine reprocessed 51 human intervention studies (3,128 blood samples) through a single standardized pipeline, calculating 16 epigenetic clocks and 94 additional methylation biomarkers for each. After correction for multiple testing, only 13 interventions significantly lowered epigenetic age, five raised it, and 26 moved nothing. Anti-TNF therapy and Mediterranean diet protocols were the only interventions whose effects held up under both internal consistency and independent replication.
Key Points
- Only 13 of 51 interventions lowered epigenetic age after correction
- Senolytics failed both internal consistency and replication criteria
- Anti-TNF therapy and Mediterranean diet alone met both thresholds
Longevity Analysis
The measurement instrument now matters as much as the intervention: first-generation clocks such as Horvath and Hannum carried substantial biological noise, while DunedinPACE and PCGrimAge produced the most coherent signal, with DunedinPACE most responsive to lifestyle change and GrimAgeV2 to pharmaceuticals. That the two replicable results involve suppressing inflammatory signaling and a dietary pattern that lowers inflammatory burden points toward interference removal rather than addition as the more reliable lever on measurable aging. For anyone acting on a consumer biological age result, the practical implication is to select the biomarker that matches the intervention being tested, and to treat a single unreplicated shift as provisional rather than decisive.
Original published by Longevity.Technology, by Kyle Umipig.