State-specific dynamic functional connectivity disturbances in spinal cord injury-related neuropathic pain

This exploratory study aimed to characterize dynamic functional network connectivity (dFNC) alterations in spinal cord injury-induced neuropathic pain (SCI-NP) and dFNC-related metrics as biomarkers for treatment guidance. A total of 23 SCI individuals with NP (SCI-NP), 20 SCI individuals without NP (SCI-N), and 20 healthy controls (HCs) were enrolled. SCI-NP individuals were predominantly chronic (≥12 months post-injury), whereas SCI-N individuals were predominantly acute/subacute (<12 months post-injury). Resting-state fMRI data were analyzed using independent component analysis (ICA) combined with a sliding-window approach to characterize dFNC. Recurring connectivity states were identified using clustering algorithms. Group differences were assessed using ANCOVA, adjusting for age and sex. P values were FDR-corrected for multiple comparisons. Spearman's correlation analyses examined associations between dFNC metrics and pain and clinical scores. Twenty-two whole-brain independent components were identified, and seven functionally meaningful components were selected for dFNC analysis, which revealed six distinct connectivity states. Compared with the SCI-N group, the SCI-NP individuals showed increased State 1 connectivity between the medial visual network (mVN) and sensorimotor network (SMN) and between the default mode network (DMN) and SMN (both p < 0.05, uncorrected), reflecting enhanced cross-modal sensory integration and maladaptive plasticity commonly observed in chronic neuropathic pain. In contrast, State 6 exhibited significantly reduced DMN-visual network (VN) connectivity in the SCI-NP group (p < 0.05, uncorrected), representing a low-interaction resting configuration that may reflect preserved functional segregation preferentially maintained in the SCI-N group. The SCI-NP individuals had shorter mean dwell time (p = 0.04, uncorrected) and fraction of time (p = 0.03, uncorrected) in State 6 relative to SCI-N individuals. None of these differences survived FDR correction.These findings indicate that SCI-NP individuals exhibit state-specific dynamic connectivity disturbances, including maladaptive hyperconnectivity involving sensorimotor, visual, and default mode networks. Metrics of dFNC-including state-specific connectivity and dwell time may serve as sensitive imaging features for guiding targeted rehabilitation strategies in SCI-NP.