Abstract Central fatigue affects 80% of Multiple Sclerosis (MS) patients, with 60% of them claiming it as the most disabling symptom. Current research often independently explores neurophysiological, structural, or functional imaging and biological underpinnings of fatigue, thus lacking a multidimensional perspective. Here, we used a multidimensional approach to investigate the functional, structural, and biological underpinnings of fatigue in MS and to assess the relative contribution of each factor. A cross-sectional study was conducted with 41 relapsing-remitting MS patients and 21 healthy controls (Female 14) (HC). MS patients, paired according to the score at the Expanded Disability Status Scale, were categorized as fatigued (MS-F: 19, Female 13, FSS ≥ 4) or non-fatigued (MS-NF: 22, Female 11, FSS 4). Over five phases, participants underwent Transcranial Magnetic Stimulation, resting-state Electroencephalography, structural and functional Magnetic Resonance, clinical assessments, and blood tests for neurofilament light chain, serum glial fibrillary acidic protein, and cytokine levels. Data were analyzed using both non-parametric and parametric tests, based on the data distribution. Finally, a decision-tree model was applied to predict patient group assignment. Neurophysiologically, the two patient groups differed in several domains. Those with fatigue had increased θ-band EEG power in frontocentral regions with eyes open. Transcranial Magnetic Stimulation findings indicated significantly lower Intracortical Facilitation (ICF) in the MS-F group. Neuroimaging revealed stronger functional connectivity between nodes of the Default Mode Network, between the left temporal node and the right prefrontal node, in the MS-F group. Furthermore, fractional anisotropy (FA) via Diffusion Tensor Imaging showed less myelin integrity in the corticospinal tracts and corpus callosum in these patients. No significant differences were observed in lesion load, brain volumes, clinical/psychological measures, or blood sample findings linked with neurodegeneration or inflammation; the only psychological variable that differed between the two groups was the depression scale score, with MS-F patients reporting higher scores than MS-NF patients. The decision tree analysis identified both ICF and significantly lower FA values as the most accurate predictors of fatigue, with a classification accuracy of 84.2%. Results highlight the importance of a multidisciplinary approach in defining central fatigue in MS, which would emerge through subtle, subclinical, regional abnormalities of myelin integrity and clearly manifest neurophysiological evidence of impaired glutamatergic activity in motor areas. They also suggest possible biomarkers for the diagnosis of fatigue, possibly useful for eventual targeting novel neuromodulatory treatments.
Benelli et al. (2026) studied this question.