Biological age prediction using a DNN model based on pathways of steroidogenesis

Aging involves the progressive accumulation of cellular damage, leading to systemic decline and age-related diseases. Despite advances in medicine, accurately predicting biological age (BA) remains challenging due to the complexity of aging processes and the limitations of current models. This study introduces a method for predicting BA using a deep neural network (DNN) based on pathways of steroidogenesis. We analyzed 22 steroids from 148 serum samples of individuals aged 20 to 73, using 98 samples for model training and 50 for validation. Our model reflects the often-overlooked fact that aging heterogeneity expands over time and uncovers sex-specific variations in steroidogenesis. This study leveraged key markers, including cortisol (COL), which underscore the role of stress-related and sex-specific steroids in aging. The resulting model establishes a biologically meaningful and robust framework for predicting BA across diverse datasets, offering fresh insights and supporting more targeted strategies in aging research and disease management.