Unexpected walking perturbations: Reliability and validity of a new treadmill protocol to provoke muscular reflex activities at lower extremities and the trunk

Abstract

Instrumented treadmills offer the potential to generate standardized walking perturbations, which are particularly rapid and powerful. However, technical requirements to release adequate perturbations regarding timing, duration and amplitude are demanding. This study investigated the test-retest reliability and validity of a new treadmill perturbation protocol releasing rapid and unexpected belt perturbations to provoke muscular reflex responses at lower extremities and the trunk. Fourteen healthy participants underwent two identical treadmill walking protocols, consisting of 10 superimposed one-sided belt perturbations (100 ms duration; 2 m/s amplitude), triggered by a plantar pressure insole 200 ms after heel contact. Delay, duration and amplitude of applied perturbations were recorded by 3D-motion capture. Muscular reflex responses (within 200 ms) were measured at lower extremities and the trunk (10-lead EMG). Data was analyzed descriptively (mean ± SD). Reliability was analyzed using test-retest variability (TRV%) and limits of agreement (LoA, bias ± 1.96^∗SD). Perturbation delay was 202 ± 14 ms, duration was 102 ± 4 ms and amplitude was 2.1 ± 0.01 m/s. TRV for perturbation delay, duration and amplitude ranged from 5.0% to 5.7%. LoA reached 3 ± 36 ms for delay, 2 ± 13 ms for duration and 0.0 ± 0.3 m/s for amplitude. EMG amplitudes following perturbations ranged between 106 ± 97% and 909 ± 979% of unperturbed gait and EMG latencies between 82 ± 14 ms and 106 ± 16 ms. Minor differences between preset and observed perturbation characteristics and results of test-retest analysis prove a high validity with excellent reliability of the setup. Therefore, the protocol tested can be recommended to provoke muscular reflex responses at lower extremities and the trunk in perturbed walking.

Introduction

Experimentally induced stumbling during locomotion is often used to investigate kinematic and neuromuscular activation patterns in response to disturbances of dynamic postural control (Forner Cordero et al., 2003, Marigold and Misiaszek, 2009, Savin et al., 2010). Though perturbations are induced in a variety of ways, computer controlled treadmills offer the potential to generate standardized walking perturbations, which are particularly rapid and powerful (Berger et al., 1984, Forner Cordero et al., 2003, Granacher et al., 2006, Horstmann et al., 1987, Sessoms et al., 2014, Sloot et al., 2015). Such characteristics are specifically required for investigations in muscular reflex activities following the perturbation (Granacher et al., 2006, Müller et al., 2016, Sloot et al., 2015). For valid identification of reflex activities, discriminated from ongoing perturbation effects, the application of short timed stimuli is crucial (Sloot et al., 2015). Also, as neuromuscular responses and recovery strategies are related to the specific gait phase, a distinct timing of stimuli application is inevitable (Forner Cordero et al., 2003, Schillings et al., 2000). Though meeting these requirements is technically challenging, data on reliability and validity of applied perturbation characteristics during perturbed treadmill walking are rare (Horstmann et al., 1987, Sessoms et al., 2014, Sloot et al., 2015). In a recently published pilot investigation the authors were able to provoke neuromuscular reflex activities and kinematic changes in perturbed treadmill walking (Müller et al., 2016). While the induced perturbations lead to both a displacement of the body and to an increase of muscle EMG activity, the reliability and validity of the underlying perturbation protocol remained unknown. Therefore, the present study aimed to evaluate the technical reliability and validity of the underlying treadmill perturbation protocol, designed to specifically provoke muscular reflex responses at the trunk and lower extremities.