In multiple sclerosis (MS) the prevalence of cognitive impairment (CI) ranges 40–65%. Besides, high level of cognitive reserve (CR) is strongly associated with better cognitive functioning in patients with MS (PwMS). How the cognitive performance and the CR level in PwMS can be associated to the fluctuations of spontaneous EEG is not completely understood.
To compare scalp voltage maps (microstates) during resting between PwMS and healthy controls (HCs) and to investigate how the temporal parameters of microstates and the level of CR can predict the cognitive performance.
50 relapsing-remitting multiple sclerosis (RRMS) patients and 25 HCs, matched for age and sex, were enrolled. For PwMS, we administrated the Brief International Cognitive Assessment (BICAMS) and the Cognitive Reserve Index questionnaire (CRIq). All participants underwent to 15-min of high-density EEG recording (256ch), closed-eyes. EEG data were filtered 1-40Hz, ICA-corrected for artifacts and downsampled to 256Hz. Microstates analysis identified a set of voltage maps representing the EEG activity for all participants that were fitted on the corrected-EEGs to quantify: global explained variance (GEV), mean duration (MD), time coverage (TC) and occurrence (Frequ). Repeated Two-Way ANOVA (Bonferroni comparisons) was used to compare groups and microstates; stepwise multiple linear regression was performed to study the cognitive performance in correlation to the microstates and level of CR; alpha=0.05.
24% of PwMS had CI and 34% reported a low level of CR. Microstates analysis found 4 maps in both groups and PwMS showed a significant increase in Map-A (GEV/TC/Frequ) and Map-B (GEV), while Map-C (MD) and Map-D (MD/TC) were significantly decreased with respect to HCs. Multiple linear regression analysis showed two strong predicted models (p<0.001), respectively, for Brief Visual-Spatial Memory Test (BVMT) and Symbol Digit Memory Test (SDMT). BVMT was predicted by GEV of Map-C improved by CRI_L, reaching 44% of explained variance. SDMT was predicted by Frequ of Map-A and CRI_Edu.
Microstate analysis reveled altered fluctuations of EEG topographies in RRMS patients with low disability and at the first stage of disease. In particular, Map-C (salience network) and leisure as well as Map-A (auditory processing) and education, respectively, predicted BVMR and SDMT scores.