Reliability Of Vertical Oscillation During Running Using Marker-less Motion Capture

Vertical oscillation (VO) of the center of mass during running has been investigated relative to both running economy and injury risk. Recent technology such as wearable accelerometers and marker-less motion capture (ML MOCAP) has made the measurement of VO easier and enhanced its potential application in gait retraining. However, limited inquiries on the reliability of VO from ML MOCAP have been conducted. PURPOSE: To determine the reliability of VO from ML MOCAP when assessed across three speeds on three distinct data collection sessions. METHODS: Twenty-one experienced distance runners (14 F, 19.5 ± 1.4 yrs) performed three consecutive two-minute treadmill running trials at self-selected speeds (ratings of perceived effort of 3, 5 and 7-out-of-10). The same protocol was repeated two additional times with at least one day and no more than ten days between visits. ML MOCAP software processed video data (120 Hz) to determine 19-segment 3D model pose estimates and whole-body center of mass (COM) position throughout each trial. Segment properties (segmental mass and COM locations) were derived from previous work to calculate full body COM trajectory. VO for each trial was then computed as the difference between successive maxima and minimain the vertical direction. VO was averaged across all gait cycles within each trial (>35 per trial). Statistical analysis, conducted in JASP (ver. 0.17.1) included intraclass correlation coefficients, Bland-Altman limits of agreement (LoA) over 95% confidence intervals, standard error of measurement (SEM), and minimum detectable change (MDC) values. RESULTS: Mean VO (Session 1: 9.0 ± 1.3 cm; Session 2: 9.1 ± 1.4 cm; Session 3: 9.3 ± 1.6 cm) demonstrated excellent reliability (ICC = 0.909; LoA: [0.823,0.959]) across the 21 participants and three distinct testing sessions. Additionally, both the SEM (0.293 cm) and MDC (0.959 cm) were < 1 cm. CONCLUSION: The VO values (range: 6.5-13.1 cm) computed from ML MOCAP were consistent with previous reports that measured VO with other methods. Within a population of experienced distance runners and over a range of speeds (2.67 – 4.44 m/s), ML MOCAP was a convenient and reliable method to measure VO during treadmill running.