Author(s): DAWSON, L., MCERLAIN-NAYLOR, S., DEVEREUX, G., BEATO, M., Institution: ST MARY'S UNIVERSITY TWICKENHAM, Country: UNITED KINGDOM, Abstract-ID: 2154

INTRODUCTION:
The importance of athlete monitoring in team sports is now widely recognised and commonly achieved using GNSS trunk-mounted player tracking systems. Systems allow for the output and interpretation of multiple external load metrics which practitioners use to inform the prescription and adaptation of training load. However, interunit reliability is important to allow comparisons to be made within and between athletes and teams. The aim of this study was to assess the interunit reliability of STATSports Apex units during repeated sprints of varied distances and changes of direction, to assess how reliability and metrics may change with the demands of repeated sprints.
METHODS:
Thirty-three female football players performed 5 sets of 2 x 20 m, 4 x 10m and 8 x 5 m repeated sprint protocols (200m total), while wearing two harnesses and STATSports Apex GNSS units, placed alongside each other, between the scapulae. The interunit reliability was assessed using intraclass correlation coefficient (ICC), for each protocol and all combined, for a selection of GNSS and accelerometer-based (ACC) metrics. One-way ANOVAs were used to compare the metrics reported for each sprint protocol.
RESULTS:
Across all protocols GNSS metrics demonstrated good to excellent interunit reliability (ICC: 0.845 - 0.999), while ACC metrics fatigue index (ICC: 0.495) and dynamic stress load (ICC: 0.484) were poor. Variation in reliability across individual sprint protocols was found for accelerations (2 x 20m ICC: 0.692, moderate; 4 x 10m ICC: 0.865, good; 8 x 5m ICC: 0.789, good), metabolic distance (2 x 20m ICC: 0.871, good; 4 x 10m ICC: 0.928, excellent; 8 x 5m ICC: 0.974, excellent), fatigue index (2 x 20m ICC: 0.393, poor; 4 x 10m ICC: 0.438, poor; 8 x 5m ICC: 0.544, moderate), and dynamic stress load (2 x 20m ICC: 0.445, poor; 4 x 10m ICC: 0.455, poor; 8 x 5m ICC: 0.539, moderate). Notably, despite increased acceleration and deceleration demands with more changes of direction, ACC metrics were not significantly different between any sprint protocols (p = 0.064 - 1.000).
CONCLUSION:
GNSS metrics show good to excellent interunit reliability, except for accelerations (moderate to good), therefore it is recommended that practitioners can confidently use APEX GNSS units for load monitoring of GNSS metrics. However, ACC metrics have demonstrated largely poor to moderate interunit reliability and highlighted difficulty distinguishing between the demands of varied sprint distances. It has been recognised that the ICC increases from 2 x 20m to 8 x 5m, suggesting that reliability may improve by increasing the amount of activity monitored. Therefore, short protocol duration may also explain the lack of significant differences identified between sprint protocols. It is recommended to avoid the interchanging of units between athletes and sessions, and interpret ACC metrics cautiously, with regard to the type and duration of activity being monitored.