Peak lower limb joint angles are weak predictors of hamstring length change during sprinting

Non-contact hamstring injuries (HSIs) commonly occur during the late swing phase of sprinting, when muscle-tendon units (MTUs) approach maximum length. Although sagittal-plane pelvis, hip and knee angles are often used as surrogate measures of overall hamstring lengthening, their predictive validity remains uncertain. This study investigated whether peak three-dimensional lower limb joint angles predict length change in the biarticular hamstring MTUs from peak hip flexion through to toe off (0-100 %) in fourteen intermediate-level male soccer players sprinting at maximal speed (8.56 ± 0.47 m·s^-1). Participant-specific musculoskeletal models were used to compute MTU lengths for the biceps femoris long head (BFlh), semimembranosus (SM), and semitendinosus (ST). To account for inter-subject temporal variability and enable accurate point-to-point comparisons across trials, dynamic time warping was applied for non-linear temporal registration. Statistical parametric mapping regression was used to assess associations between pelvis, hip and knee peak joint angles (sagittal, frontal, transverse) and length change of BFlh, SM, and ST. Peak sagittal angles were poor predictors while peak frontal pelvis angle was negatively associated with MTU length change during late swing (BFlh peak |r| = -0.371; SM peak |r| = -0.460). Frontal hip adduction was negatively associated with MTU length change from peak hip flexion to early stance (peak |r| -0.39 to -0.46). Internal hip rotation was associated with SM and ST lengthening (peak |r| = 0.51) from late swing to early stance. Knee extension angles showed no significant associations. These findings suggest caution when using single-plane joint angles as isolated indicators of hamstring MTU length.