87 How to identify fallers and non-fallers in people with HSP Biomechanical assessment In order to obtain the biomechanical measures of gait stability, a gait analysis was performed on an instrumented treadmill. We extracted the following biomechanical measures: i) variability of step length, step time and step width8, ii) medio-lateral margin of stability (MoS) 22, iii) adherence to the medio-lateral foot placement strategy through foot placement deviations (FPD) 23, and iv) local dynamic stability through maximum local divergence exponents (LDEs). 24 All measures are further explained below. To this end, 35 reflective markers were placed on anatomical landmarks according to the standard upper and lower body plug-in-gait model. Thereafter, the individual comfortable treadmill speed was determined: the treadmill speed was gradually increased until the participant stated that it felt comfortable. Then, treadmill speed was increased by 0.3 ms-1 before gradually decreasing the speed until participants again stated they walked at their comfortable speed. 10, 25 The average speed of both assessments was used as the individual comfortable gait speed during the two three-minute walking trials. Participants with HSP performed two additional trials of three-minute treadmill walking at a fixed speed of either 0.6ms-1 or 0.8ms-1. In contrast, healthy controls were asked to perform two additional three-minute walking trials at both fixed walking speeds. All participants were instructed to walk as comfortable as possible without holding onto the treadmill bars. Participants with HSP wore their own (orthopedic) shoes and orthotic devices if required. Kinematic data were collected using a 10-camera 3-dimensional motion capture system (Vicon Nexus, Oxford, UK) at a sampling rate of 100Hz. The treadmill was equipped with two force plates sampling at a rate of 2000Hz. Details of the kinematic data processing can be found in Supplementary Explanations 1. Prior to calculating the biomechanical measures, walking trials were checked for events where participants lost balance and needed to hold onto the bars. The steps occurring during these events were removed, in addition to two additional steps prior to and following the event. This was done to ensure that the biomechanical measures reflected the participants’ ability to independently recover from small perturbations, without these measures being influenced by the external support that the bars provided. Heel strikes were identified as the moment of heel marker position minima in the vertical direction during a gait cycle. Step length and step time were calculated between two subsequent heel strikes (i.e., from left heel strike to right heel strike, and vice-versa). Step width was calculated as the distance between the left and right heel markers at mid stance. The mean and standard deviation of the spatiotemporal measures were calculated over the valid left and right steps collected during two similar three-minute walking trials. 6
RkJQdWJsaXNoZXIy MjY0ODMw