Altered network recruitment in multiple sclerosis patients during resting state

Electroencephalography (EEG) allows a versatile recording of neuronal activity in neurological disorders. Signal analytical techniques like time-frequency analysis (TFA) can uncover latent information in the EEG of patients. Classically, EEG and TFA analysis relies on predefined frequency bands. Cluster-based permutation statistics provides an unbiased statistical comparison of high dimensional neurophysiological data, stabilizing the measured effect size and increasing the probability of uncovering the main effect present.

Resting state surface EEG recordings from 51 Multiple Sclerosis (MS) and 51 control patients were analyzed using TFA and compared by cluster-based permutation statistics. We further correlated results with disease characteristics, retinal nerve fiber layer (RNFL) thickness assessed by optical coherence tomography, evoked potentials (EPs), and neuropsychological data.

We detected increased power in the low and high beta frequency bands in MS compared to control patients in a subset of recording sites. Spectral power in the high beta band of resting state EEG recorded at O1 negatively correlated with the RNFL thickness. Furthermore, differences in low and high beta power were dependent on the EP score of MS patients.

Our results suggest a more desynchronized EEG activity in MS patients compared to controls, correlating with clinical markers of disease severity. The findings indicate differential brain network recruitment that may be the result of compensatory mechanisms and the basis for clinical impairment in MS.