Robustness of mediolateral stability measures to different operationalizations of the center of mass and base of support.

Abstract

Mediolateral stability metrics may be highly sensitive to how the center of mass (CoM) and base of support (BoS) are defined. We therefore evaluated the impact of different CoM and BoS operationalizations on three gait stability metrics: the mediolateral margin of stability (MoS), its variability, and the probability of instability. Our objective was to identify methodological limitations of these definitions. Two existing datasets were analyzed: regular walking in people with hereditary spastic paraplegia, and narrow base walking in healthy participants. Multiple CoM methods (full-body, reduced-marker, pelvis-based, and ground reaction force) and BoS definitions (center of pressure, lateral malleolus marker, and foot-midpoint) were used to compute mean mediolateral MoS, its variability, and the probability of instability. Stability metrics were compared to a dataset-specific reference condition: full-body with center of pressure in the HSP dataset, and reduced-maker with foot-midpoint in the narrow base walking dataset. Compared to the reference condition, alternative CoM and BoS operationalizations resulted in maximal changes of 1.3 cm (mean) and 0.4 cm (variability) for CoM methods, compared to 4.5 cm (mean) and 0.1 cm (variability) for BoS definitions. The probability of instability changed by up to 35 % points. Particularly during narrow base walking, differences in probability of instability were large, since a small mediolateral MoS increased the sensitivity to variations in CoM and BoS definitions. Our findings underscore the importance of selecting appropriate CoM and BoS definitions when calculating mediolateral stability measures, to allow detection of subtle differences in mediolateral stability and improve comparability between studies.