Abstract:
Freezing of gait (FoG) is a debilitating motor symptom in patients with Pakinson’s disease (PD), characterized by a sudden inability to walk forward. It is leading to severe impairments of patients’ mobility, quality of life and resulting in falls, injuries, and hospitalization. The missing understanding of the underlying pathophysiology thus goes along with unexploited therapeutic options.
Deep brain stimulation devices with sensing capability enable to record oscillatory neuronal activity in patients while walking and exhibiting freezing. This technology allowed us to measure local field potentials (LFPs) with Deep Brain Stimulation (DBS) electrodes in PD patients to investigate subthalamic activity during and preceding a FoG episode, along with gait kinematics measured with external sensors. To delineate pathological changes, we further investigated the modulation of subthalamic activity over a gait cycle and during internally generated voluntary stops as a control condition.
We recorded subthalamic activity in 12 PD patients while performing different tasks such as sitting, standing, walking and walking with voluntary stops. To help trigger FoG episodes, the patients were off medication and completed the gait experiments in a corridor, which was additionally narrowed by two obstacles. The collected data were then pre-processed, synchronized in time and checked for artefacts. FoG was identified based on video and kinematic data and the LFP signal subsequently divided into different epochs according to the different tasks. Time frequency domains and time-frequency analyses were calculated for Walking, FoG and Stops as well as for the transition periods preceding them. Non-parametric permutation testing was used to identify significant clusters during different conditions.
We found that the subthalamic nucleus (STN) shows a significant attenuation in alpha and beta band power during Walking compared to Standing. Furthermore, the STN displays gait-cycle related modulation of alpha and beta band power. FoG showed a numerical increase of activity in the lower beta band and a significant decrease in the high beta band. In the time frequency analyses, FoG was characterised by fluctuating power increases in the beta band. The alpha and beta attenuation, as it could be observed during walking, was no longer present. Moreover, our time-frequency analyses showed that numerical increase and fluctuations in alpha and low beta activity during the gait cycle could already be observed in the transition period before FoG occurs.
In contrast, Stop was characterized by a significant decrease of alpha activity. In the time-frequency analyses alpha and beta activity decreased shortly after the beginning of a Stop. Beta activity returned to normal around 380ms after Stop onset in sense of a rebound. During Pre-Stop, there was a significant attenuation in the alpha band range and the observed suppression of both the lower and upper beta bands was largely preserved from Walking.
We were able to demonstrate, that FoG differs completely in its STN activation patterns from those of normal walking and voluntary stops. The increased alpha and low beta activity in the STN could be a potential cause and indicator of disturbed movement flow and interruption of the STN's oscillatory dynamics required for physiological gait. We were able to show that these changes can be measured using LFPs in freely moving patients and are already visible in the transition phase before the actual occurrence of FoG. They displace the normal gait modulation and indicate the evolution of motor network failure. Those findings highlight the role of the STN in encoding effective forward walking and its diagnostic and potentially therapeutic utility in FoG. Beyond the pathophysiological insights into FoG, our findings provide potential to establish biomarkers that precede the worsening of the patient’s clinical state. This is a key step leading to closed loop stimulation and optimizing current DBS therapy.
Deep brain stimulation