Investigation of Molecular and Metabolic Mechanisms of Long-term Exercise Adaptations Through a Multi-omics Approach

Abstract

Elite athletes maintain their exceptional physical performance through high-intensity training, which significantly alters their blood metabolic profile. These metabolic changes are influenced by exercise type and duration. A systematic metabolic study of exercise is crucial for elucidating the physiological mechanisms underlying athletic performance and developing personalized approaches in sports science and health management. However, existing research on exercise physiology has primarily focused on detecting physiological differences between endurance athletes and non-athletes or analyzing pre- and post-training changes. These approaches fail to reflect long-term metabolic adaptations to exercise and lack the sensitivity required to accurately capture metabolic variations. Moreover, athletic groups do not exist as strictly binary classifications but rather along a continuous spectrum, necessitating a novel approach to better characterize these differences. This study seeks to determine how different training types influence the metabolic landscape of elite athletes and whether distinct metabolic patterns can be identified for endurance- and power-based training. To achieve this, multi-omics profiling was applied to provide a more precise and comprehensive evaluation of physiological adaptations to training. Serum samples were collected from 172 elite athletes who participated in national and international sports events and tested negative for doping. Targeted metabolomics-based mass spectrometry and untargeted proteomic profiling were performed. Orthogonal partial least squares discriminant analysis was used for multivariate analysis, while univariate linear models were employed to assess metabolic differences between high- and moderate-power and endurance athletes. Furthermore, metabolic and proteomic profiles cannot be fully explained by simple linear relationships. To address this, Lasso regression analysis was applied to identify the most significant features distinguishing high-power, high-endurance, and hybrid athletes. This study provides new insights into the metabolic pathways and biomarkers associated with athletic performance, contributing to the development of personalized training strategies and optimized recovery programs for elite athletes.