(1) ATHLETE MONITORING WEARABLE JUMP HEIGHT VALIDITY AND RELIABILITY

Purpose: To test the validity and reliability of jump height (JH) using a basketball- and volleyball-specific wearable to the gold standard of motion capture. Methods: Nineteen physically active young adults (21.3 ± 0.57yrs, 172 ± 10.3cm, 71.2 ± 17.5kgs) completed 2 visits within 2 weeks of each other. On visit one, following a warmup, an individual max jump height was determined by a Vertec. Once the max JH was determined, three ranges (45-55%, 75-85% & 95-105% of max JH) were calculated. During both visits, participants warmed up and were asked to wear a basketball- and volleyball-specific wearable (Kinexon Perform IMU, KINEXON Sports, Chicago, IL) in a pocket sewn into the waistband of shorts and another in a belt situated around waist level. Motion capture markers were placed on their feet, thorax, and back (1200 Hz; Qualisys, Göteborg, Sweden). Participants were asked to complete a minimum of five jumps for each JH range. JH ranges were randomized, and ample rest was provided between jumps. JHs were calculated by manufacturer software. JHs from motion capture were calculated using T12 or sternal notch marker by subtracting starting from peak height of the marker. Bland-Altman plots were used to measure agreement between the wearable technology and the gold standard for both shorts and belt locations at each of the three JH (50, 80, 100%). Bias and 95% confidence interval were recorded. Using the average of the five jumps per JH range, intraclass correlation coefficients (ICC (3,k)) established consistency from visit 1 to visit 2. ICC values were interpreted < 0.5 as poor, 0.5-0.75 as moderate, 0.75-0.9 as good, and >0.9 as excellent reliability. ICC 95% confidence intervals (CI) were also calculated. Results: Wearable JH from the shorts and belt had a near similar bias of -2 inches below the motion capture height. Specifically, JH for shorts was: -1.78 ± 0.62, -2.04 ± 0.83, -2.32 ± 0.97 inches, respectively for 50, 80, 100% max JH, compared to motion capture. Similarly at the belt, wearable JH bias compared to the motion capture was: -1.78 ± 0.69, -2.02 ± 0.88, -2.26 ± 1.06 inches, respectively for 50, 80, 100% max JH. ICC values for the shorts were 50% JH: 0.885 (CI: 0.643, 0.963), 80% JH: 0.954 (CI: 0.858, 0.985), and 100%: 0.971 (CI: 0.908, 0.991). In the belt location, ICCs were similar (50% JH: 0.878 (CI: 0.621, 0.961), 80% JH: 0.932 (CI: 0.788, 0.978), and 100% JH: 0.962 (CI: 0.882, 0.988)). Conclusions: Overall, the wearables underestimated jumps by about two inches. As the jump ranges converged toward maximum JH, both the mean difference and variability of bias increased slightly. There was no apparent difference in validity for the location of the wearable. Furthermore, the reliability of the wearable was good to excellent for all JH ranges in both locations. The ICC CI demonstrated that as jumps were closer to maximal, reliability was extremely strong. Overall, although the wearable may have had a bias of two inches, this bias was consistent across testing sessions. Practical Applications: The implication of an underestimated JH for athlete monitoring is the inability to compare it to other performance monitoring metrics from force plates or Vertec, which are also routinely used by sports scientists and strength and conditioning coaches. However, since the wearables have high reliability, a linear regression can be applied to correct for proper JH. Acknowledgements: None