Brain glucose metabolism network and connectivity in patients with somatoform disorders / 中华核医学与分子影像杂志

ABSTRACT

Objective@#To observe the alteration of brain glucose metabolic network in patients with somatoform disorders (SFD).@*Methods@#18F-fluorodeoxyglucose (FDG) PET images of 18 SFD patients (10 males, 8 females; age (39.5±12.0) years; illness duration (3.67±3.20) years) and 21 matched healthy controls (13 males, 8 females; age (43.9±8.4) years) in Huashan Hospital of Fudan University from October 2011 to December 2012 were enrolled to construct the brain glucose metabolic networks for 2 groups (SFD group, control group) respectively. Then the global network properties (normalized clustering coefficient, normalized shortest path length, small-worldness and global efficiency) and local parameters (clustering coefficient and betweenness centrality of the node) were calculated using the graph theory. Differences between 2 groups were compared by permutation test with 1 000 permutations. The top 20% (18/90) were classified as Hub nodes based on the results of clustering coefficient and betweenness centrality of the node.@*Results@#Small-worldness of SFD patients was similar to that of healthy controls (σ>1). There were decreased tendency in normalized clustering coefficient and global efficiency, and increased tendency in normalized shortest path length in SFD patients, but without significant differences (P>0.05). Compared to healthy controls, the betweenness centrality of left pallidum, left amygdala, left precuneus and right angular gyrus increased (permutation test, P<0.05); the betweenness centrality of left middle temporal gyrus, right superior occipital gyrus decreased (permutation test, P<0.05); the clustering coefficients of bilateral pallidum, bilateral thalamus, and left amygdala decreased (permutation test, P<0.05). Most changed Hub nodes (16/24) belonged to limbic system.@*Conclusion@#The changes of topological properties of brain glucose metabolic network in SFD patients including the decreased tendency of small-worldness and global efficiency, as well as the altered Hub nodes, may provide valid imaging evidences for brain dysfunction of somatization symptoms.