Reliability of the Main 2D Kinematic Variables of Running Evaluated Categorically in Amateur Runners, Aged 18 to 55 Years.

Abstract

Background: Running offers numerous health benefits, yet the prevalence of running-related injuries remains perfect. Evaluating running biomechanics through kinematic variables has gained attention due to their association with injuries. Currently, 3D video analysis is considered the gold standard, however, its complexity and cost limit widespread use. Consequently, 2D kinematic analysis using standard cameras has become popular, although research on its reliability is lacking. Hypothesis/Purpose: The purpose of this study was to evaluate the reliability of the main kinematic variables of running, assessed categorically using 2D video analysis, in amateur runners aged 18 to 55 years.

Study design: A cross-sectional analytical observational study.

Methods: Thirty-three amateur runners aged 18-55 underwent evaluation. Reflective markers were placed on key anatomical landmarks of the dominant lower limb to enhance motion tracking in 2D video recordings from sagittal and posterior frontal planes. The videos were analyzed by three independent raters: a consensus group (four persons), a separate physical therapist, and an athletic trainer. Pairwise inter-rater agreement was assessed among all three raters using the Kappa coefficient. Nine categorical kinematic variables related to running biomechanics were assessed.

Results: The variables that presented the greater concordance were the foot progression angle (K 0.847, 0.654, 0.792), heel whip (K 0.780, 0.847, 0.835), knee window (K 0.847, 0.847, 0.835) and overstriding (K 0.920, 0.780, 0.857). The variables with the lowest concordance were pelvic drop (K 0.257, 0.047, 0.597) and hindfoot eversion (K 0.604, 0.236, 0.604). Perfect concordance was obtained in seven of the nine variables between the consensus group and the physical therapist.

Conclusion: 2D kinematic analysis exhibited promising reliability for specific running variables, suggesting its potential in clinical settings. However, its effectiveness depends on both the evaluated variables and the expertise of the evaluators.

Level of evidence: Level 3.