Functional Connectivity Alterations of the Central Autonomic Network in Multiple Sclerosis: Links to Fatigue and Aerobic Training Effects.

Abstract

Background and objectives: Autonomic dysfunction is common in multiple sclerosis (MS), and autonomic symptoms are independent contributors to MS fatigue. Aerobic training (AT) has been shown to reduce fatigue severity and to promote functional network reorganization in MS. We explored resting-state (RS) functional connectivity (FC) of the central autonomic network (CAN) in MS, its relationship with fatigue severity, and its modifications after AT vs non-AT.

Methods: Seventy-five patients with MS (38 with relapsing-remitting [RR] and 37 with progressive [P]) underwent 3T RS functional MRI at baseline and at the end of AT (24 training sessions, 2-3 times per week for 2-3 months) or non-AT of equivalent duration. Sixty-seven matched healthy controls (HCs) served as a reference for baseline RS FC. At both visits, all patients were assessed using the Modified Fatigue Impact Scale (MFIS). Seed-based RS FC used core CAN modulatory regions: left/right ventromedial prefrontal cortex (vmPFC), mid-cingulate cortex (MCC), amygdala, hypothalamus, and anterior and posterior insula.

Results: Compared with HCs, patients with MS showed increased baseline RS FC of the right amygdala, anterior insula, and vmPFC with superior frontal gyri, precentral gyri and right precuneus. Considering disease phenotype, we found a divergent behavior of baseline RS FC in the interaction analysis (p < 0.05, corrected), with patients with RRMS showing predominantly increased RS FC of the amygdala, anterior insula, and MCC with supplementary motor areas (SMAs) and deep gray matter, whereas patients with PMS showed widespread baseline RS FC decrease, particularly involving posterior insula connections with the cerebellum and parietal regions, and MCC connections with frontotemporal areas. In patients with RRMS, increased baseline CAN RS FC inversely correlated with fatigue severity (r = range -0.33/-0.54, p range = 0.046/0.001), while no significant correlations were found in patients with PMS. In patients who underwent AT, decreased hypothalamic network RS FC at follow-up vs baseline was observed (p < 0.001, significant time-by-treatment interaction vs non-AT), with more prominent changes in patients with RRMS compared with patients with PMS. No correlations were found between longitudinal RS FC modifications and concomitant fatigue changes.

Discussion: CAN dysregulation is present in MS, with distinct RS FC abnormalities characterizing patients with RRMS and PMS and correlating with fatigue severity. AT modulates CAN RS FC, particularly in patients with RRMS.