Application of Deep Learning Algorithms for Identifying Hidden Patterns in Oxidative Stress Biomarker Data and Analyzing the Effects of Combined Aerobic–Resistance Training on Metabolic Health in Sedentary Individuals: A Multilayer Biomarker‑Based and Feature Discovery Study
Application of Deep Learning Algorithms for Identifying Hidden Patterns in Oxidative Stress Biomarker Data and Analyzing the Effects of Combined Aerobic–Resistance Training on Metabolic Health in Sedentary Individuals: A Multilayer Biomarker‑Based and Feature Discovery Study
Samin Nekahi1
1) Master of Exercise Physiology (Health and Fitness), Islamic Azad University, Taft Branch, Taft, Iran.
IOI Code
:
ABEB-ABECBH
Publication :
4th International Conference on Recent Advances in Engineering, Innovation and Technology - Belgium(eitconf.com)
Abstract :
Physical inactivity is a major contributor to oxidative stress, chronic low‑grade inflammation, and metabolic dysfunction, significantly increasing the risk of cardiometabolic diseases. Oxidative stress results from an imbalance between the production of reactive oxygen species (ROS) and the body’s antioxidant defense capacity, leading to cellular damage and impaired metabolic regulation. Understanding how structured exercise interventions influence oxidative biomarkers—and how advanced computational methods can uncover hidden biological patterns—has become increasingly important. The aim of this review was to examine the effects of combined aerobic–resistance training on oxidative stress biomarkers and metabolic health, and to explore the application of deep learning approaches for multilayer biomarker pattern recognition and feature discovery. This study reviewed physiological mechanisms linking oxidative stress to metabolic health and evaluated evidence regarding exercise-induced adaptations in sedentary individuals. In addition, deep learning models—including convolutional neural networks, transformer architectures, and autoencoder-based frameworks—were conceptually integrated to analyze complex, nonlinear relationships among oxidative biomarkers, inflammatory indicators, and metabolic variables. A multilayer analytical approach was employed to identify key predictive features and biomarker clusters associated with training responsiveness. Findings indicate that combined aerobic–resistance training significantly reduces lipid peroxidation markers, enhances antioxidant enzyme activity, improves redox balance, and positively influences metabolic parameters such as insulin sensitivity and cardiorespiratory fitness. Deep learning–based analyses revealed interconnected biomarker clusters linking oxidative stress, immune signaling, and metabolic regulation. Certain biomarkers, including glutathione ratios and superoxide dismutase activity, emerged as strong predictors of adaptive response to training. Individuals with higher baseline oxidative imbalance tended to exhibit greater physiological improvements following structured exercise. In conclusion, integrating combined exercise interventions with advanced computational modeling provides a powerful framework for understanding oxidative stress regulation and metabolic adaptation in sedentary populations. This approach supports the development of personalized, biomarker‑guided exercise strategies for improving metabolic health.
Keywords : oxidative stress combined training deep learning biomarkers metabolic health