Abstract
Background:
Impairments in postural stability are common following concussion. However, subjectively-scored assessments of postural control, such as the Balance Error Scoring System, may not accurately reflect postural instability. Objective measures, such as gait performance, can more accurately identify ongoing impairments in postural control throughout concussion recovery. Recent technological advances make gait evaluations more feasible across different clinical settings.
Purpose:
To compare performance on smartphone measured single- and dual-task gait evaluations among adolescent concussion participants and healthy controls across two timepoints.
Methods:
We evaluated concussed participants at initial (n=13; 14.6±1.3 years; 54% female; 5.5±1.6 days post-concussion) and follow-up visits (33.2±3.3 days post-concussion), and healthy control participants (n=21; 15.5±1.6 years; 48% female) at two timepoints 27.3±0.7 days apart on single- and dual-task gait performance. Single-task gait trials consisted of walking at a self-selected pace for 20 meters. Single-task cognitive trials involved responding to a cognitive task while standing in place. For dual-task trials, participants completed gait and cognitive tasks simultaneously. During each trial a smartphone affixed to the lumbar spine measured gait velocity and cognitive response accuracy and response time. We conducted a 2x2 mixed effects ANOVA to examine changes between groups across the testing timeline.
Results:
Concussion participants reported significantly greater symptom severity than control participants at the initial timepoint (23.1±13.5 vs 7.4±8.5; p=0.001); however, this difference did not persist to the second timepoint (p=0.99). The concussion group demonstrated significant improvements in single-task gait velocity between timepoints; although, between-group differences in gait velocity were not observed at either timepoint (Figure 1). No differences in cognitive accuracy between groups were identified. However, concussed participants performed worse than healthy controls on cognitive response time at both time points in single- and dual-task testing (Figure 2). Further, response time improved between testing periods for single-task, but not dual-task, conditions.
Conclusion:
Smartphone gait evaluations indicate improvement in single-task gait velocity during concussion recovery. However, no significant differences in gait velocity were found between concussed and control participants at either time point in single- or dual-task gait conditions. Interestingly, concussion patients had slower cognitive responses than healthy controls at both time points, despite no differences in cognitive accuracy or symptom severity at time 2. Our preliminary findings suggest that smartphone gait evaluations are effective in measuring changes in gait velocity and cognitive performance following adolescent concussion, as well as detecting subtle differences in gait and cognitive performance between concussed and control patients throughout recovery.
Concussion and control group means at Time 1 and Time 2 for average gait velocity. Error bars represent the 95% confidence interval for the mean.
Concussion and control group means at Time 1 and Time 2 for Stroop task response times. Error bars represent the 95% confidence interval for the mean.