Motor-related oscillatory activity in schizophrenia according to phase of illness and clinical symptom severity

Abstract

Magnetoencephalography (MEG) measures magnetic fields generated by synchronised neural current flow and provides direct inference on brain electrophysiology and connectivity, with high spatial and temporal resolution. The movement-related beta decrease (MRBD) and the post-movement beta rebound (PMBR) are well-characterised effects in magnetoencephalography (MEG), with the latter having been shown to relate to long-range network integrity. Our previous work has shown that the PMBR is diminished (relative to controls) in a group of schizophrenia patients. However, little is known about how this effect might differ in patients at different stages of illness and degrees of clinical severity. Here, we extend our previous findings showing that the MEG derived PMBR abnormality in schizophrenia exists in 29 recent-onset and 35 established cases (i.e., chronic patients), compared to 42 control cases. In established cases, PMBR is negatively correlated with severity of disorganization symptoms. Further, using a hidden Markov model analysis, we show that transient pan-spectral oscillatory "bursts", which underlie the PMBR, differ between healthy controls and patients. Results corroborate that PMBR is associated with disorganization of mental activity in schizophrenia.

Keywords: Oscillatory bursts; Post-movement beta rebound; Psychosis; Schizophrenia; Transient beta events.

Fig. 1

HMM output for a single virtual sensor timeseries, in one representative healthy control participant. A) Broadband (1–48 Hz) regional timecourse data (arbitrary amplitude) with beta burst state identified by the HMM (in red, with trials demarcated with red lines). B) Time-frequency decomposition of the same data generated using a Morlet wavelet transform (trials demarcated in red). C) Spectra showing the component frequencies of the three states identified from the HMM. The state most highly correlated with the beta envelope (the beta burst state 3) is shown in red. D) Raster-plot showing the binarised occurrences of the burst state for all trials, note how the MRBD and PMBR time periods can be easily distinguished. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Visuo-motor task source localisation. Localisations shown for beta desynchronization (blue) and beta rebound (yellow/red) (radiological view; left on right) for A) CT group, B) RO patients and C) ES patients. Images were thresholded to show > 90% of maximum intensity for PMBR and minimum intensity for MRBD. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

Visuo-motor source power results A) Scale shows power relative to baseline. Top line shows the right MRBD, middle line shows the left MRBD and bottom line shows the left PMBR for controls, recent-onset (RO) and established (ES) patients. B) Left: A representation of the beta frequency (13–30 Hz) time course for the controls, recent-onset patients and established patients separately, showing percentage change of the PMBR from baseline. Mean and SEM shown. Right: Bar chart of the difference between baseline and PMBR windows as a percentage change. Mean and CI shown. Graphs do not include outliers.

Fig. 4

PMBR relationship with PANSS. The plot shows the disorganization PANSS subscale score factor correlated with the PMBR percentage change from baseline for the established patient group.

Fig. 5

PMBR relationship with antipsychotic exposure score. The plot shows the relationship between the PMBR and antipsychotic medication score for recent-onset (RO; blue) and established patients (ES; red). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

Characteristics of HMM identified task-related bursts. A) Power spectral density of chosen burst state, red for controls (CT), black for recent-onset patients (RO) and blue for established patients (ES). Mean and SEM shown. B) Raster plots of binarised bursts across time per trial. Top plot shows controls, middle plot shows recent-onset patients, bottom plot shows established patients. C) First column: time courses of HMM burst characteristics, top to bottom, amplitude, burst state probability and burst duration. Mean and SEM shown. Second column: Percentage change of the relevant HMM burst characteristic within the PMBR (2.5–4.5 s) relative to the baseline window (6–9 s) for each time course. Mean and CI shown. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)