Biofeedback gait retraining under real-world running conditions

Abstract

Gait retraining has been used as an intervention to mitigate the risk of running-related injuries among distance runners. Lab-based gait retraining studies have used various feedback strategies to modify the running gait and have demonstrated promising results in changing biomechanical parameters associated with injuries. However, lab-based training is not accessible to the general running population, and there was limited evidence that supports the transfer of training effect to conditions that resemble real-world running. For these reasons, gait retraining in runners' natural training environments might be preferred. The main objective of this thesis was to optimize the training protocol for gait retraining under real-world conditions. To this end, five studies were conducted to address three specific aims: 1) identify the limitations of lab-based gait retraining protocols, 2) assess habitual gait adaptations in real-world running, and 3) establish the technical specifications for systems used to modify gait outside of the lab. Regarding the first specific aim, two studies were conducted to examine the transfer of training effect to untrained conditions, including overground and slopes. Results of both studies suggested that runners who regularly train overground and on slopes may not benefit fully from lab-based training. Based on such findings, gait retraining along overground running routes that include slopes was recommended. Our third study addressed the second specific aim, it examined the natural biomechanical adaptations while running on slopes. By analyzing real-world training records, changes in speed and cadence were observed along slopes. These habitual adaptations could interact with the training parameters in gait retraining, subsequently affecting the training effect. Therefore, an adaptive feedback model with the training target set based on different sloped conditions was recommended. With the existing wearable technology, tibial acceleration can be measured using wireless accelerometers outside of the lab. Tibial acceleration is a common parameter used as feedback during training and as an outcome measure to assess the effect of gait retraining. The fourth and fifth studies addressed the third specific aim and presented the technical considerations required for accurate and reliable tibial acceleration measurements under conditions that resemble training in the real world. Based on the findings of these two studies, accelerometers with an operating range wider than ±16-g were recommended for accurate tibial acceleration measurements. Also, it was recommended to measure a minimum of 100 consecutive strides during each session when evaluating training performance to ensure reliable measurements. To conclude the findings of the five studies, a gait retraining protocol designed for training under real-world conditions was proposed and evaluated in a proof-of-concept study. This final study demonstrated the feasibility of using adaptive feedback in real-world training using a wearable sensor system. Tibial acceleration was measured under real-world training conditions and was used as feedback to guide the runner in modifying the gait pattern while training along slopes. Overall, the findings of this thesis provided insights for further optimization of gait retraining protocols and the future development of feedback systems suitable for use under real-world conditions.