Abstract

Introduction: Although deep brain stimulation often improves levodopa-responsive gait symptoms, robust therapies for gait dysfunction from Parkinson's disease remain a major unmet need. Walking speed could represent a simple, integrated tool to assess DBS efficacy but is often not examined systematically or quantitatively during DBS programming. Here we investigate the reliability and functional significance of changes in gait by directional DBS in the subthalamic nucleus.Methods: Nineteen patients underwent unilateral subthalamic nucleus DBS surgery with an eight-contact directional lead in the most severely affected hemisphere. They arrived off dopaminergic medications >12 h preoperatively and for device activation 1 month after surgery. We measured a comfortable walking speed using an instrumented walkway with DBS off and at each of 10 stimulation configurations at the midpoint of the therapeutic window. Repeated measures of ANOVA contrasted preoperative vs. maximum and minimum walking speeds across DBS configurations during device activation. Intraclass correlation coefficients examined walking speed reliability across the four trials within each DBS configuration. We also investigated whether changes in walking speed related to modification of step length vs. cadence with a one-sample t-test.Results: Mean comfortable walking speed improved significantly with DBS on vs. both DBS off and minimum speeds with DBS on. Pairwise comparisons showed no significant difference between DBS off and minimum comfortable walking speed with DBS on. Intraclass correlations were ≥0.949 within each condition. Changes in comfortable walk speed were conferred primarily by changes in step length.Conclusion: Acute assessment of walking speed is a reliable, clinically meaningful measure of gait function during DBS activation. Directional and circular unilateral subthalamic DBS in appropriate configurations elicit acute and clinically significant improvements in gait dysfunction related to PD. Next-generation directional DBS technologies have significant potential to enhance gait by individually tailoring stimulation parameters to optimize efficacy.