Wearable Sensors to Quantify Performance and Fatigue during Tournament Competition among Elite Developmental Ice Hockey Players

Abstract

At the highest performance levels of ice hockey (e.g. senior/professional), rules govern the number of successive competition days and prevent multiple competitions on a single day. Yet, some youth leagues run weekend showcase events that include multiple competitions per day. It is generally assumed that multiple competitions per day and several competitions over multiple days will impart excessive fatigue and impair performance, but there is no evidence directly related to ice hockey. PURPOSE: Use player-worn sensors (PWS) to compare accelerations (ACC) and heart rate (HR) over 4 games in 3 days among elite youth ice hockey players in order to establish changes in cardiovascular stress and physical exertion associated with fatigue. METHODS: 33 elite, youth ice hockey players in two age categories (18U: N=17, 18.2 yrs ± .92 & 16U: N=16, 16.4 yrs ± 1.1) on two teams in a league showcase consented to procedures approved by the EMU-HSRC. Bioharness-3 (Zephyr, MD) recorded HR and ACC at 1 Hz over the 4 games (G1-G4) in the 3- day event. Peak ACC across multiple time frames (10, 15, 20, 30,[Midline Horizontal Ellipsis] 90, sec and 2, 2.5, 3, 5, 10, [Midline Horizontal Ellipsis], 45 min) were quantified and analyzed and HR was quantified and used in conjunction with ACC to determine exertion profiles for each on-ice session. MANOVAs for peak ACC and HR at each time point across G1 -G4 with Bonferoni post hocs and multiple games per day (M1, M2) for magnitude and time as main effects were performed using SPSS 23.0 (IBM, NY; α=0.05). RESULTS: HR (bpm) decreased G1 v G4 at 3 (187.5 ± 2.8 v 176.1 ± 2.8), 5 (178.7 ± 3.0 v 165.7 ± 3.0), 10 (170.4 ± 3.1 v 157.6 ± 3.1), 15 (167.1 ±3.0 v 153.3 ± 2.9) and 20 min (164.3 ± 3.0 v 150.5 ± 2.9)(p<.05). In addition, HR also decreased G1 v G3 for 10 (170.4 ± 3.1 v 158.5 ± 3.0), 15 (167.1 ±3.0 v 154.5 ± 2.9) and 20 min (164.3 ± 3.0 v 152.4 ± 2.9)(p<.05), although ACC were not different. Peak ACC (g’s) were lower for M2 vs M1 at 60 (0.632 ± 0.012 v 0.592 ± 0.012), 90 (0.551 ± 0.01 v 0.520 ± 0.01), 120 (0.495 ± 0.009 v 0.469 ± 0.009) and 180 sec (0.412 ± 0.009 v 0.386 ± 0.009)(p<.05). CONCLUSIONS: The decline in HR, but not ACC across games over 3 days indicates a cardiovascular adaptation as opposed to overt fatigue. On the other hand, the reduced ACC from 60 - 180 sec between games 1 and 2 in a day indicate reduced shift capacity and an overall decline in performance indicative of fatigue during the second contest.