Subthalamic beta power and gait in Parkinson's disease during unsupervised remote monitoring

Background: Brain-sense enabled deep brain stimulation (DBS) in Parkinson's disease (PD) allows understanding the neurophysiological role of basal ganglia in gait dysfunction. Subthalamic nucleus (STN) local field potentials (LFPs) indicate pathological PD gait changes, but previous evidence is based on data obtained in controlled settings. This study aims to examine the relationship between STN oscillations and sensor-derived gait parameters in PD, in an unsupervised, real-life environment. Methods: Patients with PD and STN-DBS (Medtronic Percept™-PC) were enrolled. Collected data included motor assessment, medications and stimulation parameters, lead locations, and continuous LFPs recordings of pre-selected frequency of interest in the beta range. Participants were all equipped with a single inertial measurement unit, placed at their lower back to record 10-h continuous data on step/stride variability and regularity, and duration. Results: A total of 10 patients and 256 walking episodes (mean duration of 88.9s, cumulative walking duration of 37.05min, mean number of steps taken for episode of 70) were collected and analyzed. STN beta power showed a significant correlation with normalized gait parameters. Right STN beta was related to step (r = 0.23, p < 0.01) and stride (r = 0.22, p < 0.001) time variability. Similar but non-significant trends were observed between gait parameters and left STN beta power (r = 0.11, p = 0.06 and r = 0.11, p = 0.06). Conclusions: This pilot study confirms that beta dynamics are indicative of pathological gait in both controlled and uncontrolled environments. Studies with larger sample sizes and prospective design will be helpful to validate and expand our pilot work toward diagnostic tools and translations into clinical practice.