Criterion Validity of Accelerometer-Derived Peak Velocity During Jump Squats

• Published in: Journal of strength and conditioning research Vol. 25; pp. S53 - S54
• Main Authors: Weiss, L W, Ferreira, L C, Feldmann, C R, Schilling, B K, Hammond, K G
• Format: Journal Article
• Language: English
• Published: Champaign Lippincott Williams & Wilkins Ovid Technologies 01.03.2011
• Subjects: Accelerometers; Kinematics; Load; Motor ability
• ISSN: 1064-8011; 1533-4287
• DOI: 10.1097/01.JSC.0000395662.44411.80

An accelerometer-based dynamometer has been proposed as an alternative to more sophisticated devices for measuring bar velocity during resistance exercise. PURPOSE: To determine the stability reliability, precision, and criterion validity of an accelerometer device for assessing peak velocity during low-load jump squats. METHODS: Fifty-two subjects (25 men, 27 women) with a minimum of three months squat training experience were recruited from a university population. All performed a one repetition maximum (1RM) back squat to 90 degrees of knee flexion with an external load ≥ body weight. Subjects performed duplicate load-spectrum countermovement jump squats (CMJ) and static jump squats (SJ) at 20, 30, and 40% of their back squat 1RM on two separate days in a counterbalanced sequence. A direct measure of peak velocity (linear velocity transducer) served as the criterion, with data sampled at 500 Hz. Predicted peak velocity was obtained using duplicate dynamometers, each containing a triaxial accelerometer directly affixed to each end of the barbell midway between the lateral-most aspect of the shoulder and the thumb side of each hand. Acceleration data were downsampled from 1.5 KHz to 500 Hz by averaging every three data points via proprietary software. Velocity was then determined by integration. Stability reliability was assessed using intraclass correlation (ICC), precision by coefficient of variation (CV), and criterion validity by bivariate correlation (r). If either reliability or precision did not meet an acceptable standard (ICC ≥ 0.70; CV ≤ 15%), then bivariate correlations were calculated from the average of values for the two test sessions. RESULTS: Stability reliability and precision were relatively good for the criterion (ICC = 0.83 to 0.92; CV ≤ 3.8%, n[asymptotically =]51) for all CMJ and SJ conditions. For the duplicate accelerometers, stability reliability was mostly below standard (ICC = 0.44 to 0.79, n[asymptotically =]50), while precision was acceptable for all conditions (CV ≤ 13.4%). Since velocity data were largely unreliable, criterion validity was calculated using the average of the two test sessions, and was found to be low to moderate (r = 0.34 to 0.65, n[asymptotically =]48) for the two accelerometers. Conclusion: The linear velocity transducer produced both reliable and precise measures of peak velocity for both types of low-load jump squats. The accelerometer devices produced precise peak velocity data that was suboptimal for reliability and marginally-valid at best for all loads of the two styles of jump squats. Practical Applications: Within the constraints of this study, the accelerometer device appears to be a convenient but marginally useful alternative to a linear velocity transducer for measuring peak velocity for low-load jump squats in the laboratory, although clinical applications may be more robust. Study supported in part by Myotest, Inc. [PUBLICATION ABSTRACT]