Brain Functional Connectome Gradients Analysis Study of Chronic Back Pain

<bold>Objective</bold> The aim of this study was to investigate the relationship between changes in functional network hierarchy and pain in patients with chronic back pain (CBP) by analyzing the differences in functional network connectivity gradients between the CBP group and the healthy control group. <bold>Methods</bold> Data (including basic clinical information such as VAS scores, years of pain, and resting functional MRI images) from the patients with CBP and healthy controls (HC) were obtained from the OpenPain database. Participants were grouped into 32 pairs matched by age and sex. The rs-fMRI images were pre-processed to obtain Fisher-Z transformed functional connectivity matrices, and BrainSpace was used to analyse the gradient scores of seven functional networks (sensorimotor network, visual network, dorsal attention network, ventral attention network, limbic network, frontoparietal control network and default mode network) in the two groups. The differences in the gradient scores of the functional networks between the two groups were analyzed statistically. Besides, the correlation analysis was applied to explore the relationship between the different functional network gradient scores and the VAS scores. <bold>Results</bold> Both ends of the functional network hierarchy (i.e., sensorimotor and default mode networks) were significantly constricted in the CBP group compared to the HC group, and gradients in the cortical range were suppressed. Compared to the HC group, the CBP group had higher gradient scores of sensorimotor, visual, dorsal attention, and ventral attention network, and lower gradient scores of frontoparietal control, limbic, and default mode network. There were significant differences in the increased gradient scores of sensorimotor network (<italic>P</italic>&lt;0.01) and ventral attention network (<italic>P</italic>&lt;0.001), and in the decreased gradient scores of limbic network (<italic>P</italic>&lt;0.05) and default mode network (<italic>P</italic>&lt;0.001). In addition, correlation analysis showed that the gradient scores of sensorimotor network in the CBP group was positively correlated with VAS scores (<italic>r</italic>=0.432, <italic>P</italic>&lt;0.05), which suggested that higher sensorimotor network gradient scores represented more severe pain responses. Besides, there was a negative correlation between the gradient scores of default mode network (<italic>r</italic>=-0.404, <italic>P</italic>&lt;0.05) and VAS scores, which suggested that lower gradient scores of default mode network was associated with more severe pain symptoms. <bold>Conclusion</bold> The macroscale network hierarchy in the CBP group was significantly altered and the alteration was closely related to pain symptoms. The results not only enrich the research of gradient analysis, but also provide a new perspective for further research on the neural mechanisms of pain in patients with CBP.