Classifying Endurance Runners by Kinematics Strengthens Relationships of Muscle-Tendon Unit Function with Running Economy and Performance

PURPOSES: Running kinematics and sex may alter the primary muscle and tendon demands for propulsion and shock absorption, potentially modifying how muscle-tendon unit (MTU) function relates to the energy cost of running and performance. This study examined whether clustering runners by running kinematics strengthens these relationships and how they differ by sex. METHODS: One hundred twenty endurance runners (60 females) underwent assessments of MTU function, including slow (30°/s) and fast (180°/s) isokinetic knee extension (KE) and flexion (KF) torque and performance in countermovement, drop, and rebound jumps. These variables were summarized using principal component analysis. Running kinematics and energy cost of running were assessed at 11 km/h and 80% of vV̇O2max, and the season's best time (World Athletics [WA] score) was recorded. Runners were clustered via k-means based on foot strike angle and KE/KF range of motion. Stepwise regression analyses tested whether MTU-related associations differed between cluster-specific and whole-cohort models. RESULTS: K-means identified three or four clusters per sex at each running velocity condition. In whole-cohort analyses for each sex, only weak to fair correlations were found between MTU function and either energy cost (r = -0.204 to 0.267) or WA score (r = -0.276 to 0.175). Clustering significantly (P < 0.05) strengthened these relationships in a cluster-specific manner (adjusted R2 = 0.159 to 0.595). This pattern was more evident at 80% of vV̇O2max, particularly in male clusters. Across conditions, MTU-related variables were selected in a greater proportion of female clusters (57.1%) than male clusters (31.2%). CONCLUSIONS: Kinematic-based clustering strengthens sex- and cluster-specific associations of MTU function with the energy cost of running and performance, supporting more individualized performance enhancement strategies.