Aberrant brain structural–functional connectivity coupling associated with cognitive dysfunction in different cerebral small vessel disease burdens

Abstract Aims Emerging evidence suggests that cerebral small vessel disease (CSVD) pathology changes brain structural connectivity (SC) and functional connectivity (FC) networks. Although network‐level SC and FC are closely coupled in the healthy population, how SC‐FC coupling correlates with neurocognitive outcomes in patients with different CSVD burdens remains largely unknown. Methods Using multimodal MRI, we reconstructed whole‐brain SC and FC networks for 54 patients with severe CSVD burden (CSVD‐s), 106 patients with mild CSVD burden (CSVD‐m), and 79 healthy controls. We then investigated the aberrant SC‐FC coupling and functional network topology in CSVD and their correlations with cognitive dysfunction. Results Compared with controls, the CSVD‐m patients showed no significant change in any SC‐FC coupling, but the CSVD‐s patients exhibited significantly decreased whole‐brain ( p = 0.014), auditory/motor ( p = 0.033), and limbic modular ( p = 0.011) SC‐FC coupling. For functional network topology, despite no change in global efficiency, CSVD‐s patients exhibited significantly reduced nodal efficiency of the bilateral amygdala ( p = 0.024 and 0.035) and heschl gyrus ( p = 0.001 and 0.005). Notably, for the CSVD‐s patients, whole‐brain SC‐FC coupling showed a significantly positive correlation with MoCA ( r = 0.327, p = 0.020) and SDMT ( r = 0.373, p = 0.008) scores, limbic/subcortical modular SC‐FC coupling showed a negative correlation ( r = −0.316, p = 0.025) with SCWT score, and global/local efficiency ( r = 0.367, p = 0.009 and r = 0.353, p = 0.012) showed a positive correlation with AVLT score. For the CSVD‐m group, whole‐brain and auditory/motor modular SC‐FC couplings showed significantly positive correlations with SCWT ( r = 0.217, p = 0.028 and r = 0.219, p = 0.027) and TMT ( r = 0.324, p = 0.001 and r = 0.245, p = 0.013) scores, and global/local efficiency showed positive correlations with AVLT ( r = 0.230, p = 0.020 and r = 0.248, p = 0.012) and SDMT ( r = 0.263, p = 0.008 and r = 0.263, p = 0.007) scores. Conclusion Our findings demonstrated that decreased whole‐brain and module‐dependent SC‐FC coupling associated with reduced functional efficiency might underlie more severe burden and worse cognitive decline in CSVD. SC‐FC coupling might serve as a more sensitive neuroimaging biomarker of CSVD burden and provided new insights into the pathophysiologic mechanisms of clinical development of CSVD.