Clustering classification of cyclists according to the acute fatigue outcomes produced by an ultra-endurance event

Abstract

This study aimed to analyze the differences between clusters obtained by the acute effect of fatigue after an ultra-endurance event in the internal and external load of cyclists. 26 volunteers participated in the study, and they were divided into the experimental group (N = 18; height: 177 ± 8 cm; body mass: 78.6 ± 10.3 kg) and the control group (N = 8; height: 176 ± 10 cm; body mass: 78.0 ± 15.7 kg). The experimental group completed a 12 h non-stop cycling event. Jump height, lactate, plasma antioxidant capacity, pain perception and fatigue perception were measured before and after the event. Cyclists of the experimental group were classified considering their training characteristics (recreational vs. competitive) and by conducting a non-supervised K-means clustering. The differentiation of cyclists according to training characteristics resulted in a lower distance covered by recreational than competitive cyclists (279.4 ± 39.7 km vs. 371.0 ± 71.7 km; ES ≥ 0.8; p < 0.01), although no differences were observed in the remaining variables between groups (p > 0.05). The clustering analysis provided two clusters. Cluster 2 suffered a greater jump height reduction (-3.3 ± 1.6 vs. 1.2 ± 0.8; ES ≥ 0.8; p < 0.001) and increased pain and fatigue perception (ES ≥ 0.5; p < 0.05) after the race than Cluster 1. In conclusion, counter-movement jump can differentiate the fatigue produced by a cycling ultra-endurance event and therefore, this non-invasive technique is useful in fatigue monitoring and recovery planification.