Aging is a complex process characterized by various phenotypic changes, including epigenetic modifications, inflammation, and metabolic dysregulation. Recent advancements in the field of epigenetics have led to the development of DNA methylation clocks, which utilize DNA methylation profiles to estimate biological age and predict age-related outcomes. However, most existing clocks rely on linear models that may not adequately capture the complexity of survival-related DNA methylation signatures.
In this study, the authors constructed a heterogeneous stacked ensemble survival model using DNA methylation data from the Framingham Heart Study. They identified 190 CpG loci through elastic net Cox regression, which were then used to create a survival prediction model. This model was developed by integrating five complementary survival models using a neural network meta-learner.
The performance of the survival model was rigorously evaluated in an external validation cohort, revealing strong predictive capabilities for all-cause mortality. The results indicated that this ensemble model significantly outperformed PhenoAge and demonstrated statistical comparability to GrimAge. However, it is important to note that these findings were derived from cohorts of European ancestry and validated specifically in postmenopausal women aged 50-79 years, suggesting that the applicability of the model may be limited to demographically similar populations.
What this study does not show
This study does not establish the clinical utility of the ensemble DNA methylation clock for predicting health outcomes beyond mortality, nor does it provide evidence for its effectiveness in diverse populations.
Evidence scorecard
- Analytic validity: established
- Clinical validity: partial
- Clinical utility: not established
The evidence is based on primary research with human status unclassified.