Dynamic Changes in Functional Network Connectivity Involving Amyotrophic Lateral Sclerosis and Its Correlation With Disease Severity

Abstract

Background: Aberrant static functional connectivity (FC) has been well demonstrated in amyotrophic lateral sclerosis (ALS); however, ALS-related alterations in FC dynamic properties remain unclear, although dynamic FC analyses contribute to uncover mechanisms underlying neurodegenerative disorders.

Purpose: To explore dynamic functional network connectivity (dFNC) in ALS and its correlation with disease severity.

Study type: Prospective.

Subjects: Thirty-two ALS patients and 45 healthy controls.

Field strength/sequence: Multiband resting-state functional images using gradient echo echo-planar imaging and T1-weighted images were acquired at 3.0 T.

Assessment: Disease severity was evaluated with the revised ALS Functional Rating Scale (ALSFRS-R) and patients were stratified according to diagnostic category. Independent component analysis was conducted to identify the components of seven intrinsic brain networks (ie, visual/sensorimotor (SMN)/auditory/cognitive-control (CCN)/default-mode (DMN)/subcortical/cerebellar networks). A sliding-window correlation approach was used to compute dFNC. FNC states were determined by k-mean clustering, and state-specific FNC and dynamic indices (fraction time/mean dwell time/transition number) were calculated.

Statistical tests: Two-sample t test used for comparisons on dynamic measures and Spearman's correlation analysis.

Results: ALS patients showed increased FNC between DMN-SMN in state 1 and between CCN-SMN in state 4. Patients remained in state 2 (showing the weakest FNC) for a significantly longer time (mean dwell time: 49.8 ± 40.1 vs. 93.6 ± 126.3; P < 0.05) and remained in state 1 (showing a relatively strong FNC) for a shorter time (fraction time: 0.27 ± 0.25 vs. 0.13 ± 0.20; P < 0.05). ALS patients exhibited less temporal variability in their FNC (transition number: 10.2 ± 4.4 vs. 7.8 ± 3.8; P < 0.05). A significant correlation was observed between ALSFRS-R and mean dwell time in state 2 (r = -0.414, P < 0.05) and transition number (r = 0.452, P < 0.05). No significant between-subgroup difference in dFNC properties was found (all P > 0.05).

Data conclusion: Our findings suggest aberrant dFNC properties in ALS, which is associated with disease severity.

Level of evidence: 2 TECHNICAL EFFICACY: Stage 3.

Keywords: amyotrophic lateral sclerosis; dynamic; functional network connectivity; independent component analysis; resting-state functional magnetic resonance imaging.