Oscillatory vs. non-oscillatory subthalamic beta activity in Parkinson's disease

Jesús Pardo-Valencia,Carla Fernández-García,Fernando Alonso-Frech,Guglielmo Foffani
DOI: https://doi.org/10.1113/JP284768
Abstract:Parkinson's disease is characterized by exaggerated beta activity (13-35 Hz) in cortico-basal ganglia motor loops. Beta activity includes both periodic fluctuations (i.e. oscillatory activity) and aperiodic fluctuations reflecting spiking activity and excitation/inhibition balance (i.e. non-oscillatory activity). However, the relative contribution, dopamine dependency and clinical correlations of oscillatory vs. non-oscillatory beta activity remain unclear. We recorded, modelled and analysed subthalamic local field potentials in parkinsonian patients at rest while off or on medication. Autoregressive modelling with additive 1/f noise clarified the relationships between measures of beta activity in the time domain (i.e. amplitude and duration of beta bursts) or in the frequency domain (i.e. power and sharpness of the spectral peak) and oscillatory vs. non-oscillatory activity: burst duration and spectral sharpness are specifically sensitive to oscillatory activity, whereas burst amplitude and spectral power are ambiguously sensitive to both oscillatory and non-oscillatory activity. Our experimental data confirmed the model predictions and assumptions. We subsequently analysed the effect of levodopa, obtaining strong-to-extreme Bayesian evidence that oscillatory beta activity is reduced in patients on vs. off medication, with moderate evidence for absence of modulation of the non-oscillatory component. Finally, specifically the oscillatory component of beta activity correlated with the rate of motor progression of the disease. Methodologically, these results provide an integrative understanding of beta-based biomarkers relevant for adaptive deep brain stimulation. Biologically, they suggest that primarily the oscillatory component of subthalamic beta activity is dopamine dependent and may play a role not only in the pathophysiology but also in the progression of Parkinson's disease. KEY POINTS: Beta activity in Parkinson's disease includes both true periodic fluctuations (i.e. oscillatory activity) and aperiodic fluctuations reflecting spiking activity and synaptic balance (i.e. non-oscillatory activity). The relative contribution, dopamine dependency and clinical correlations of oscillatory vs. non-oscillatory beta activity remain unclear. Burst duration and spectral sharpness are specifically sensitive to oscillatory activity, while burst amplitude and spectral power are ambiguously sensitive to both oscillatory and non-oscillatory activity. Only the oscillatory component of subthalamic beta activity is dopamine-dependent. Stronger beta oscillatory activity correlates with faster motor progression of the disease.
What problem does this paper attempt to address?