While chronological age measures time since birth, biological age reflects physiological decline, which varies across individuals and is linked to genetics, environment, and disease. In dogs, differences in longevity between breeds provide an opportunity to build biological age predictors. These predictors, often based on DNA methylation patterns, are widely studied in aging research to estimate healthspan and lifespan potential.
Armero and colleagues co-analyzed data from three major methylation studies using different measurement platforms and tissues. Their aim was to test the feasibility of constructing reliable epigenetic clocks for dogs. The study found that while some methylation markers correlate with biological age, these signals are rare, inconsistent across datasets, and do not colocalize reliably. This inconsistency undermines the ability to generate universally applicable age estimators.
One of the key findings was that population stratification—the differences in genetic background across dog breeds—introduces strong confounding effects. As a result, methylation-based biological clocks in dogs require heavy parameterization and correction for breed-specific factors to avoid biased predictions. Furthermore, the results suggest that aging-related methylation markers are scattered and often dependent on study design and tissue type.
This work underscores both the promise and the complexity of developing epigenetic clocks in dogs. It highlights the need for large, diverse datasets that integrate breed differences and environmental contexts to better understand how aging manifests at the molecular level in domestic dogs. Ultimately, refining these tools could help identify mechanisms of aging and inform veterinary and comparative medicine.
Source: Armero, A. S., Buckley, R. M., Mboning, L., Spatola, G. J., Horvath, S., Pellegrini, M., & Ostrander, E. A. (2024). Co-analysis of methylation platforms for signatures of biological aging in the domestic dog reveals previously unexplored confounding factors. Aging, 16, 10724–10748.
