Adequate Segmentation in Marker-Based Motion Capture Studies for Hyperflexion and Hyperextension Lumbar Exercises

The recent literature has debated the appropriate level of complexity for spine kinematic models. Multi-segmental analyses have been suggested to be more suitable for activities such as walking and running; however, studies focusing on sport-specific movements remain limited. This study compared four spine segmentation strategies for analyzing exercises simulating flexion and extension in acrobatic elements. Seventeen competitive university-level cheerleaders (male and female) participated in a motion capture study. Each athlete performed six exercises in the same order. Reflective markers were placed on the spinous processes of C7, T10, L1, L2, L3, L4, L5, and S1. From these, four models were constructed: (1) L1 and L5, (2) T10 and S1, (3) L1, L3, and L5, and (4) all lumbar vertebrae. Each model was fitted in the sagittal plane using a polynomial function and compared with the others via Pearson correlation. Model 3 (L1, L3, and L5) and Model 4 (all lumbar vertebrae) showed strong correlations across all trials, with Pearson coefficients approaching 1. These findings support the use of a two-segment representation of the lumbar spine (Model 3: L1-L3 and L3-L5) as a suitable approach for kinematic analysis of flexion-extension in acrobatic athletes.