Altered sensorimotor-occipital network topology in chronic insomnia: A graph-theoretical and edge-wise resting-state fMRI analysis
Zuguang Wen1, Yuefeng Kong1, Yaxuan Wen2
1First Affiliated Hospital of Guangxi Medical University, No. 6 Shuang-yong Road, Nanning, Guangxi 530021, China.
Objective:
The multi-level network aberrations in chronic insomnia (CI) and their neurobiological underpinnings remain elusive. Here, we applied graph-theoretical, edge-wise, and modular analyses to resting-state fMRI data to characterize systems-level functional network reorganization in CI patients.
Methods:
Resting-state fMRI was acquired from CI patients and healthy controls (HCs). Whole-brain functional networks were parcellated using the Dosenbach-160 atlas, and then graph-theoretical, edge-wise, and modular analyses were performed. Group differences in global, nodal, edge-wise, and modular metrics were examined via a general linear model, with age, sex, education, and framewise displacement as covariates. Global and nodal metrics were corrected using the Bonferroni method; edge-wise comparisons used network-based statistics (NBS; cluster-forming threshold p < 0.001, 5000 permutations, FWE-corrected p < 0.05).
Results:
Compared with HCs, CI patients showed significantly decreased average clustering coefficients, shorter path lengths, and reduced local efficiency, but increased global efficiency - a pattern indicative of a shift toward random network topology. Notably, normalized small-worldness scalar remained intact. Nodal clustering decreased in the left posterior occipital cortex. Edge-wise analysis revealed widespread connectivity disruptions involving parietal-posterior occipital, posterior insular-occipital, and dorsolateral prefrontal-temporal pathways. Modular analyses further demonstrated reduced connectivity both between sensorimotor and occipital networks, and within the sensorimotor network itself.
Conclusions:
CI entails multi-level functional network reorganization, characterized by global topological randomization, reduced occipital nodal clustering, and selective decoupling within sensorimotor and sensorimotor-occipital circuits. These findings suggest that CI involves widespread integration failure across large-scale systems rather than isolated regional dysfunction, with potential implications for understanding its neurocognitive sequelae.
