Validity of an instrumented sock and on-shoe sensor to provide biometric feedback to runners

Abstract

Study Design: Validity study.Background: Run form analysis and retraining are commonly utilized in orthopaedic and sports medicine environments. New tools in running technology may assist clinicians in providing feedback to runners outside the clinic. These devices provide runners the capacity to analyse multiple characteristics of their stride, but little is known regarding the accuracy of their output.Objective: Investigate the validity of an instrumented sock and on-shoe sensor as compared to high-speed video analysis.Methods: Forty recreational runners completed a 10-minute treadmill run at a self-selected pace while wearing both devices. High-speed video allowed for the analysis of average running cadence and percentage of time in each foot strike pattern. Other variables of interest included running speed and distance. Intraclass correlation coefficients (ICC2,1) were calculated for all variables. A categorical classification of dominant foot strike pattern was utilized, with Cohen’s kappa describing agreement of the devices to video analysis.Results: The instrumented sock’s ICCs were 0.40 for speed, 0.46 for distance, 0.97 for cadence and 0.40 for foot strike pattern. The on-shoe sensor generated ICCs of 0.86 for speed, 0.86 for distance, 0.99 for cadence and 0.91 for foot strike pattern. Cohen’s kappa values for foot strike pattern agreement were 0.35 for the sock and 0.79 for the on-shoe sensor.Conclusion: The instrumented sock demonstrated valid measurement of running cadence, but potential product design limitations reduced accuracy on the other study variables. The on-shoe sensor utilized in this study provided valid post-run feedback across measures of speed, distance, cadence and dichotomous classification of foot strike pattern.