More than 80% of multiple sclerosis (MS) patients experience symptoms of fatigue. MS-related fatigue can only partly be explained by structural (lesions and atrophy) and functional (brain activation and conventional static functional connectivity) brain changes.
To investigate the relationship of dynamic functional connectivity (dFC) with present and future fatigue in MS patients and compare this with commonly used clinical and MRI parameters.
In 35 relapsing-remitting MS patients (age: 42.8, female/male: 20/15, disease duration: 11 years) and 19 healthy controls (HC) (age: 41.4, female/male: 11/8), fatigue was measured using the CIS-20r questionnaire at baseline and at a 6-month follow-up. Furthermore, disability (EDSS) was assessed for patients. All subjects underwent structural MRI and resting-state functional MRI at baseline. We calculated global static functional connectivity (sFC) and assessed dynamic connectivity using a tapered sliding-window approach by calculating the summed difference (diff) and coefficient of variation (cov). Moreover, we calculated connectivity between basal ganglia and cortical regions previously associated with fatigue in MS (medial prefrontal cortex, posterior cingulate cortex, and precuneus). We performed hierarchical regression analyses with forward selection to identify the most important predictors of fatigue at baseline and follow-up.
Patients were more fatigued than HCs at baseline (MS: 74.36 ± 29.33; HC: 46.72 ± 17.06; p=0.001) and follow-up (MS: 69.91 ± 27.01; HC: 45.11 ± 19.84; p=0.002). No difference in sFC was found between patients and controls. Patients had higher baseline global dFC than controls (p<0.05) but no difference in basal ganglia-cortical dFC. Basal ganglia-cortical dFC-cov added 12.5% extra explained variance (standardized β=-0.353, p=0.032) on top of EDSS (standardized β=0.380, p=0.022) to a regression model for baseline fatigue in patients (adjusted R2=0.211, p=0.011). Post-hoc analysis revealed lower basal ganglia-cortical dFC-cov in patients with severe fatigue at baseline (0.89 ± 0.06) compared to non-fatigued patients (0.93 ± 0.05; p=0.036).
Less dynamic connectivity between the basal ganglia and the cortex is associated with greater fatigue in MS patients, independent of disability status. These findings may reflect less efficient network reconfigurations of those connections as a potential additional neural correlate of fatigue in MS.