The exploration of immune system function changes in marathon athletes after high-intensity training by Agent-Based Model

Background: Marathon running imposes substantial metabolic demand. While the acute effects of exercise are well-documented, the chronic immunometabolic alterations associated with intensified training blocks in elite populations remain characterized by complex, non-linear dynamics. Contemporary debates persist regarding whether post-exercise lymphopenia represents immunosuppression (the "open window") or a functional redistribution of immune effectors. This study explored the chronic effects of a four-week high-intensity training (HIT) block on immune markers in professional marathon athletes and evaluated the utility of an Agent-Based Model (ABM) for visualizing these system behaviors. Methods: Twenty-two professional marathon athletes (14 male, 8 female) underwent a four-week intensified training protocol characterized by sustained time in the severe-intensity domain (blood lactate > 7.0 mmol/L). Peripheral blood samples were analyzed pre- and post-training for leukocytes, immunoglobulins (Ig), cytokines (IL-6, IL-8, IL-10, TNF-α), and lymphocyte subsets. Concurrently, a NetLogo-based ABM was developed to simulate theoretical immune system dynamics under metabolic constraints. Results: < 0.001). The ABM simulation qualitatively reproduced these emergent patterns, visualizing the non-linear contraction of T-cell populations consistent with the empirical data. Conclusions: A four-week block of high-intensity marathon training is associated with a state of immunometabolic perturbation characterized by reduced circulating leukocytes, CD4+/CD8+ imbalance, and an uncoupled inflammatory-resolution cytokine response. While plasma volume expansion may contribute to the observed lower cell concentrations, the specific suppression of IL-10 and CD4+ cells suggests a maladaptive response to chronic load. The agent-based model serves as an exploratory tool for visualizing potential immunological tipping points during intensified training, bridging the gap between reductionist data and complex system dynamics.