Network-based performance analysis of multidimensional performance characteristics in trained male youth combat athletes.

Youth combat sport performance relies on a complex interplay of biological maturation, anthropometrics, neuromuscular capacity, and executive cognitive function. Despite this, these domains are typically studied in isolation. To bridge this gap, the present study applied a network-based performance analysis to uncover the multidimensional organization of these characteristics in trained male youth combat athletes. We evaluated biological maturation, anthropometry, neuromuscular performance and executive cognitive function in eighty-eight trained male athletes (13-16 years) from six combat sport disciplines. A Gaussian graphical model was estimated with the Extended Bayesian Information Criterion graphical least absolute shrinkage and selection operator (EBICglasso; γ = 0.25). Nonparametric and case-dropping bootstrap procedures were used to evaluate network accuracy and stability. The estimated network contained 35 variables with 129 of 595 possible edges (network density = 21.7%). Anthropometric and maturational characteristics, particularly skeletal muscle mass, body mass, body height and maturity offset, were found to be the primary structural components, whereas Trail Making Test A completion time and Five-Key Task incongruent accuracy occupied highly connected positions within the network, linking cognitive, anthropometric and neuromuscular domains. The strongest conditional associations were observed between countermovement jump height from flight time and velocity (weight = 0.64; 95% CI: 0.57-0.77) and maturity offset and chronological age (weight = 0.59; 95% CI: 0.54-0.81). Strength centrality had acceptable stability (CS = 0.60). The findings suggest that performance in youth combat athletes is organized as an integrated multidimensional system of conditional associations. These results support the use of network-based performance analysis as a systems-oriented framework for multidimensional athlete profiling and hypothesis generation, while longitudinal studies are required to determine developmental trajectories and evaluate potential training applications.