Explainable machine learning framework for biomarker discovery by combining biological age and frailty prediction (Preprint)

Abstract

BACKGROUND Biological age (BA) and frailty are two distinct measures of health that offer insights into aging. The existing machine learning (ML) predictors of BA or frailty from blood-based biomarkers are incapable of comparing and analyzing these biomarkers, which may deepen our understanding of these two distinct pathways toward aging. OBJECTIVE This study aimed to develop a framework to compare and analyze biomarkers by combining BA and frailty ML predictors with eXplainable Artificial Intelligence (XAI) techniques. METHODS We utilized data from middle-aged and older Chinese adults (≥45 years) in the 2011/2012 wave (n=9702) and the 2015/2016 wave (n=9455, as external validation) of the China Health and Retirement Longitudinal Study (CHARLS). Sixteen blood-based biomarkers were used to predict BA and frailty. Four ML algorithms were employed in the training and validation, and performance metrics were compared to select the best models. Then, SHapley Additive exPlanations (SHAP) analysis was conducted on the selected models. RESULTS Gradient Boosting performed the best in the BA predictor, and CatBoost performed the best in the frailty predictor. Traditional ML feature importance identified cystatin C and glycated hemoglobin as the major contributors for their respective models. However, subsequent SHAP analysis demonstrated that only cystatin C was the primary contributor in both models, suggesting that it plays an important role in both pathways. CONCLUSIONS Our novel framework can compare and analyze biomarkers by integrating BA and frailty predictors and XAI techniques. The present approach leverages routine blood biomarkers and can easily incorporate additional biomarkers, providing a scalable and comprehensive toolset that offers a quantitative understanding of interesting biomarkers and complex physiological traits. CLINICALTRIAL N/A