Correlation between cortical and subcortical neural dynamics on multiple time scales in Parkinson’s disease

Complex amplitude dynamics of dominant alpha oscillations (8–13 Hz) in the cortex can be captured with long-range temporal correlations (LRTC) in healthy subjects and in various diseases. In patients with Parkinson’s disease (PD), intra-nuclear coherence was demonstrated in dominant beta rhythms (10–30 Hz) in the basal ganglia. However, so far the relation between cortical LRTC (across tens of seconds) and subcortical coherence (millisecond scale) is unknown. We addressed these “multiscale interactions” by simultaneous recordings of surface electroencephalography (EEG) and deep local field potentials (LFP) from the bilateral subthalamic nucleus (STN) in eight patients with severe PD eligible for deep brain stimulation, who performed a lexical decision task on medication. In the continuous data set LRTC up to 20 s were calculated in the amplitude envelope of 8–13-Hz EEG oscillations (across whole scalp), and subcortical coherence was assessed with measures being insensitive to volume conduction artifacts (imaginary part of coherency; iCOH) in 10–20 and 21–30-Hz oscillations in STN–LFP. We showed a significant positive correlation across patients between cortical LRTC (8–13 Hz) and subcortical iCOH selectively in 10–20-Hz oscillations in the left STN. Our results suggest a relation between neural dynamics in the most dominant rhythms in the cortex and basal ganglia in PD, extending across multiple time scales (milliseconds vs. tens of seconds). Furthermore, the investigation of multiscale interactions might contribute to our understanding of cortical–subcortical neural coupling in PD