Validation of Accelerometry Data to Identify Movement Patterns During Agility Testing

Purpose: The purpose of this study was to develop an algorithm for the detection and timing of foot contact/off timing for each lateral repetition and assess the validity/reliability of the calculated timings. Methods: Participants performed a modified Edgren Side Step Test in which they moved laterally along a 4-m path as quickly as possible while wearing an accelerometer on each ankle. Time of completion of each attempt was recorded using a stopwatch and digital video was obtained. Accelerometer-based (ACC) events were determined for the start of the test (START), foot contact at the end-line (FC) and the lifting of the foot when transitioning to the other direction (FO). Based on these ACC events the Overall, Split (ST) and Lag (LT) times were determined and compared to either the stopwatch or video-based timings (p < 0.05). The ACC event criterion was then applied by independent reviewers to assess inter/intra-rater reliability of identifying the events. Results: There was no significant difference in ACC (12.37 ± 2.19 s) based Overall Time compared to the Stopwatch (12.42 ± 2.25 s, p = 0.34). Bland-Altman plots for ST and LT revealed very good agreement between the ACC time to the Video (ST: Bias = 0.11 s, LOA -0.57 to 0.79; LT: Bias = -0.11 s, LOA -0.43 to 0.22). Intra and inter-rater reliability was moderate to excellent for all reviewer identified events. Conclusions: This study demonstrates methodology to identify ACC based timings during an agility test. The inclusion of an accelerometer supplements standard timing options with the added benefit of assessing sided split and lag times.