Background: Generating appropriate balancing reactions in response to unexpected loss of balance during walking is important to prevent falls. The purpose of this study was to assess dynamic balancing responses following pushes to the pelvis in groups of post-stroke and healthy subjects.
Methods: Forty-one post-stroke subjects and forty-three healthy subjects participated in the study. Dynamic balancing responses to perturbations triggered at heel strike of the left or right leg, directed in the forward, backward, inward and outward directions during slow treadmill walking were assessed. Responses of the healthy group provided reference values used to classify responses of the post-stroke group into two subgroups; one within the reference responses (“inside” subgroup) and the other that falls out (“outside” subgroup). A battery of selected clinical outcome measures (6-Minute Walk Test, 10-Meter Walk Test, Timed-Up-and-Go test, Four Square Step Test, Functional Gait Assessment, Functional Independence Measure and One-legged stance test) was assessed for the post-stroke group to examine whether any of these outcome measures could discriminate between both subgroups.
Results: Both subgroups of stroke subjects were comparable in terms of clinical outcome measures but their capacity to react to unexpected loss of balance during walking differed considerably. The “inside” subgroup was able to appropriately modulate centre-of-pressure and ground-reaction-force both under the impaired and non-impaired leg. The “outside” subgroup showed limited modulation capacity under the impaired leg; their responses utilised a stepping strategy in which the non-impaired leg was placed such as to make a longer step (forward perturbation), to make a shorter step (backward perturbation) or to make a cross-step (outward perturbation). Consequently, peak centre-of-mass displacements following perturbations were significantly higher in the “outside” subgroup compared to the “inside” subgroup. Responses in both subgroups following inward perturbations did not differ considerably. One-legged stance test showed the largest potential to discriminate between both subgroups of stroke subjects.
Conclusions: The One-legged stance test could be used to obtain an indication of the abilities of each particular post-stroke subject to counteract an unexpected loss of balance. This may be relevant in clinical practice for the identification of post-stroke subjects who could benefit from perturbation-based training.