Reliability and measurement error of a wearable sensor for tibial acceleration and cadence in recreational runners

Abstract Running is an increasingly popular sport; however, injuries remain common concern among runners. The impact of tibial acceleration and cadence at different running stages has been the focus of ongoing research. While wearable devices have been developed to analyze running biomechanics, their reliability requires further investigation. This study aimed to evaluate the reliability and measurement error of the TgForce ® device in assessing tibial acceleration and cadence in runners. A cross-sectional study was conducted with 43 rearfoot-strike runners. Participants ran on a motorized treadmill for three minutes while wearing the accelerometer attached to the anteromedial region of the tibia. The device measured tibial acceleration (g) and cadence (steps/min). Participants completed the protocol twice, with a seven-day interval, and were evaluated by two examiners. The device demonstrated good inter-rater reliability for vertical tibial acceleration (ICC: 0.79; SEM: 0.93 g; SDC: 2.58 g), sagittal tibial acceleration (ICC: 0.85; SEM: 1.14 g; SDC: 3.18 g), resultant tibial acceleration (ICC: 0.81; SEM: 1.08 g; SDC: 3.01 g), and cadence (ICC: 0.88; SEM: 8.60 steps/min; SDC: 23.86 steps/min). Intra-rater reliability was good for vertical tibial acceleration (ICC: 0.82) and moderate for sagittal tibial acceleration (ICC: 0.65), resultant tibial acceleration (ICC: 0.71), and cadence (ICC: 0.73). Proportional bias was observed in the inter-rater measurement of vertical tibial acceleration and the intra-rater measurement of sagittal tibial acceleration, while systematic bias was observed in the intra-rater measurements of sagittal and resultant tibial acceleration. In conclusion, the accelerometer demonstrated moderate-to-good reliability for assessing tibial acceleration and cadence in recreational runners. Caution is warranted when interpreting small changes in cadence due to the observed measurement error (SEM and SDC). Further research is needed to evaluate reliability under real-world running conditions.