Validation of a commercially available mobile application for velocity-based resistance training.

Background: Velocity-based training (VBT) is commonly used for programming and autoregulation of resistance training. Velocity may also be measured during resistance training to estimate one repetition maximum and monitor fatigue. This study quantifies the validity of Metric VBT, a mobile application that uses camera-vision for measuring barbell range of motion (RoM) and mean velocity during resistance exercises. Methods: Twenty-four participants completed back squat and bench press repetitions across various loads. Five mobile devices were placed at varying angles (0, ±10, and ±20°) perpendicular to the participant. The validity of Metric VBT was assessed in comparison to Vicon motion analysis using precision and recall, Lin's concordance correlation coefficient, and Bland-Altman plots. Proportional bias was assessed using linear regression. Results: Metric VBT accurately detected over 95% of repetitions. It showed moderate to substantial agreement with the Vicon system for measuring RoM in both exercises. The average Limits of Agreement (LoA) for RoM across all camera positions were -5.45 to 4.94 cm for squats and -5.80 to 3.55 cm for bench presses. Metric VBT exhibited poor to moderate agreement with the Vicon system for measuring mean velocity. The average LoA for mean velocity were 0.03 to 0.25 m/s for squats and -5.80 to 3.55 m/s for bench presses. A proportional bias was observed, with bias increasing as repetition velocity increased. Conclusions: Metric VBT's wide LoA for measuring RoM and mean velocity highlights significant accuracy concerns, exceeding acceptable levels for practical use. However, for users prioritizing repetition counts over precise RoM or mean velocity data, the application can still provide useful information for monitoring workout volume.