Phase-amplitude markers of synchrony and noise: A resting-state and TMS-EEG study of schizophrenia

The electroencephalogram (EEG) of schizophrenia patients exhibits several well-known abnormalities. For brain responses evoked by a stimulus in a variety of behavioral paradigms, these alterations appear as a reduction of signal-to-noise ratio and of phase-locking, and as a facilitated excitability. Here we observe these effects in EEG using transcranial magnetic stimulation (TMS)-evoked potentials, comparing also to the resting-state EEG. To ensure veracity of our results we used three weekly sessions and analyzed both resting state and TMS-EEG data. In addition to confirming the known results for the stimulus response we also show a broadening of the amplitude distribution in the resting-state EEG of patients relative to that of controls, indicative of lower signal to noise. Specifically, we evaluated the instantaneous amplitude in narrow-band filtered EEG, using a form of mean-normalized variation (quartile coefficient, QC) as a marker for amplitude fluctuations. We find that on time scales of seconds to tens of seconds, amplitude fluctuations in the alpha and beta frequency bands in the resting state of healthy controls are larger than their fluctuations in schizophrenia patients. According to our marker, the narrow-band-filtered amplitude fluctuations in the patient group are more similar to the theoretical limit of narrow-band Gaussian noise. Our results thus support the neuronal noise hypothesis of schizophrenia, which states that the ability of neuronal populations to form locally and temporally correlated activity is reduced in schizophrenia due to inherent noise in all brain activity.