Biological Versus Technical Reliability of Epigenetic Clocks and Implications for Disease Prognosis and Intervention Response.

Abstract

Epigenetic clocks are widely used to estimate biological age and predict health outcomes, but their translational value depends not only on accuracy but also on reliability. Using the TranslAGE platform, we evaluated both technical and biological reliability across 18 DNA methylation-based aging biomarkers, including chronological, mortality, and pace-of-aging clocks. Most clocks demonstrated excellent technical reproducibility across replicate assays on EPIC and 450 K arrays. However, several showed sensitivity to experimental factors such as slide position and DNA extraction protocols. In contrast, PC-based clocks, particularly PCGrimAge and SystemsAge, remained robust across conditions. Biological reliability, assessed across repeated measures collected within short intervals under varying conditions such as meals, stress, and environmental exposures, was substantially lower. Most clocks showed only low to moderate stability which further decreased when adjusting for immune composition. Notably, technical reproducibility did not predict biological reliability. We further show that reliability directly impacts downstream applications. Clocks with higher reliability produced more stable associations with cognitive outcomes and more consistent responses to interventions, while less reliable clocks yielded variable or misleading results. These findings highlight a critical limitation in current epigenetic clocks. Improving biological reliability will be essential for advancing their use in clinical and interventional settings.