Integrating social, way of life, and genetic profiles to foretell epigenetic age acceleration in older adults

Epigenetic age acceleration (EAA) is linked to social, way of life, and behavioural exposures, but their relative significance when thought-about collectively stays unclear. To deal with this hole, we take an integrative, multi-domain method to check the relative significance of demographic, socioeconomic, psychosocial, health-behavioural, medical, and genetic elements for EAA prediction. In this cross-sectional research, we analysed 4018 Health and Retirement Study members and used LASSO, Random Forest, and XGBoost to foretell EAA throughout completely different epigenetic clocks. To establish key predictors and quantify area contributions, we fitted all-predictor fashions and skilled domain-only, ablation, and permutation fashions. In all-predictor fashions, GrimAge confirmed the best predictability (5-fold CV R² = 0.41), adopted by DunedinPoAm (R² = 0.20), whereas Hannum, Horvath, and PhenoAge confirmed little sign. Overall, predictive sign for GrimAge and DunedinPoAm EAA was concentrated in well being behaviours, largely reflecting smoking-related sign, with smaller contributions from bodily exercise, gender, family imply earnings, and African American ancestry.