Altered Effective Connectivity in Supraspinal Somatosensory Networks Reveals Divergent Adaptation Mechanisms in Complete and Incomplete Spinal Cord Injury

BACKGROUND AND PURPOSE: Supraspinal somatosensory pathways neuroplasticity critically influences sensory recovery after spinal cord injury (SCI) and offers promising neuromodulation targets. However, distinctions between complete SCI (CSCI) and incomplete SCI (ISCI) patients remain unclear. We aimed to delineate injury severity-dependent neuroplasticity patterns in somatosensory pathways, and provide mechanistic insights for developing targeted rehabilitation strategies. MATERIALS AND METHODS: Resting-state effective connectivity (EC) within supraspinal somatosensory pathways was analyzed using spectral dynamic causal modelling in 17 CSCI patients, 17 ISCI patients, and 37 healthy controls. The primary somatosensory cortex (S1), second somatosensory cortex (S2), thalamus (THA), insula (INS), cerebellar lobule VI (CB6) and primary motor cortex were employed as the regions of interest. A fully connected model was specified for each participant, and group-level differences in EC were assessed using parametric empirical Bayes. Connections with a posterior probability > 0.95 were considered significant. Additionally, correlation analyses were performed between significant EC and sensory scores. RESULTS: Both CSCI and ISCI groups exhibited impaired basic somatosensory conduction, including dysfunction from the THA to the S1 and dysregulation along the S1-S2-INS pathway. However, their specific connectivity patterns diverged. The CSCI group showed weakened self-inhibition within the THA and S1. The ISCI group exhibited stronger inhibitory EC from S2 to INS and from CB6 to S1. CONCLUSIONS: CSCI and ISCI patients predominantly exhibit decreased and increased EC within the supraspinal somatosensory pathways, respectively. The CSCI patients showed reduced THA/S1 self-inhibition, whereas ISCI patients exhibited strengthened S2-INS/CB6-S1 connectivity, possibly compensating for sensory deficits and suggesting neuromodulation targets for somatosensory recovery.