Reliability of Triaxial Accelerometry for Measuring Load in Men’s Collegiate Ice-Hockey

Van Iterson E.H., Fitzgerald J.S., Dietz C.C., Snyder E.M., Peterson B.J.

Objective:
The primary aim of this study was to evaluate the test-retest reliability of PlayerLoadTM in ice hockey players during the performance of tasks simulating game conditions using wearable microsensor technology.

Participants:
Eight male ice hockey players from Division I collegiate level participated in the study.

Methodology:
Participants wore Catapult Optimeye S5 monitors during the repeat performance of nine ice hockey tasks that simulated game conditions. These tasks included acceleration (forward/backward), 60% top-speed, top-speed (forward/backward), repeated shift circuit, ice-coasting, slap-shot, and bench-sitting. Test-retest reliability was assessed using the coefficient of variation (CV), intraclass correlation coefficient (ICC), and minimum differences (MD).

Results:
The test-retest CVs and ICCs for PlayerLoadTM were as follows:

Forward acceleration: CV = 8.6, ICC = 0.54
Backward acceleration: CV = 13.8, ICC = 0.78
60% top-speed: CV = 2.2, ICC = 0.96
Forward top-speed: CV = 7.5, ICC = 0.79
Backward top-speed: CV = 2.8, ICC = 0.96
Repeated shift test: CV = 26.6, ICC = 0.95
Slap-shot: CV = 3.9, ICC = 0.68
Coasting: CV = 3.7, ICC = 0.98
Bench-sitting: CV = 4.1, ICC = 0.98
Raw differences between bouts were not significant for any of the ice hockey tasks (P > 0.05). The raw differences between bouts were lower than the minimum differences for each task, indicating good reliability.

Conclusion:
The data suggest that PlayerLoadTM demonstrates moderate-to-large test-retest reliability in the environmental setting of male Division I collegiate ice hockey. These findings are important as PlayerLoadTM is routinely used to assess on-ice physical activity in male collegiate ice hockey without prior assessment of reliability.