Objective: To identify and interrelate static and dynamic characteristics of the restoration of quiet standing balance in a representative sample of stroke survivors in the Netherlands during their inpatient rehabilitation. Design: Exploratory study using an inception cohort with findings related to reference values from healthy elderly persons. Setting: Rehabilitation center. Participants: Thirty-seven inpatients (mean age, 61.6y; mean time poststroke, 10.0wk) with a first hemispheric intracerebral infarction or hematoma who were admitted to retrain standing balance and walking. Intervention: Individualized therapy. Main Outcome Measures: Center of pressure fluctuations were registered under each foot and in the sagittal and frontal planes separately by using a dual-plate force platform. The first balance measurements took place as soon as patients were able to stand unassisted for at least 30 seconds as well as 2, 4, 8, and, 12 weeks later. Quiet standing was assessed under 4 conditions: with and without a visual midline reference, with the eyes closed, and while performing a concurrent arithmetic task. Results: The stroke patients showed excessive postural sway and instability, particularly in the frontal plane, compared with reference values. Frontal plane balance was, however, also most responsive to the effects of balance training and recovery (P<.001). The degree of visual dependency for frontal plane balance control showed a significant reduction in time (P<.02). Weight-bearing asymmetry, which was most pronounced in patients with disturbed sensibility or ankle clonus, diminished considerably during the first 4 weeks of the follow-up period (P<.02). Yet, a substantial degree of weight-bearing asymmetry persisted during the 8 weeks thereafter, and it continued to be aggravated by attentional distraction (P<.001). During the same period, static asymmetry (ie, the degree of pes equino-varus loading at the paretic side) and dynamic asymmetry (ie, the extent to which compensatory ankle moments are applied at the nonparetic side) did not show normalization at all, although motor selectivity of the paretic leg improved by 1 stage on the 6-stage Brunnstrom scale (P<.001) and the independency level of balance and walking skills improved by 2 points on the 6-point Functional Ambulation Categories (P<.001). Conclusions: Balance recovery in postacute stroke inpatients is characterized by a reduction in postural sway and instability as well as by a reduction in visual dependency, particularly with regard to frontal plane balance. These restoration characteristics may be important factors underlying the relearning of independent standing and walking abilities. The clear lack of normalization for measures reflecting static and dynamic aspects of postural asymmetry suggests that the functional improvements in balance and gait must be more related to other mechanisms than to the restoration of support functions and equilibrium reactions of the paretic leg.
