Hierarchical integrated processing of reward-related regions in obese males: A graph-theoretical-based study

Abstract

Abnormal activities in reward-related regions are associated with overeating or obesity. Preliminary studies have shown that changes in neural activity in obesity include not only regional reward regions abnormalities but also impairments in the communication between reward-related regions and multiple functional areas. A recent study has shown that the transitions between different neural networks are nonrandom and hierarchical, and that activation of particular brain networks is more likely to occur after other brain networks. The aims of this study were to investigate the key nodes of reward-related regions in obese males and explore the hierarchical integrated processing of key nodes. Twenty-four obese males and 24 normal-weight male controls of similar ages were recruited. The fMRI data were acquired using 3.0 T MRI. The fMRI data preprocessing was performed in DPABI and SPM 12. Degree centrality analyses were conducted using GRETNA toolkit, and Granger causality analyses were calculated using DynamicBC toolbox. Decreased degree centrality was observed in left ventral medial prefrontal cortex (vmPFC) and right parahippocampal/hippocampal gyrus in group with obesity. The group with obesity demonstrated increased effective connectivity between left vmPFC and several regions (left inferior temporal gyrus, left supplementary motor area, right insular cortex, right postcentral gyrus, right paracentral lobule and bilateral fusiform gyrus). Increased effective connectivity was observed between right parahippocampal/hippocampal gyrus and left precentral/postcentral gyrus. Decreased effective connectivity was found between right parahippocampal/hippocampal gyrus and left inferior parietal lobule. This study identified the features of hierarchical interactions between the key reward nodes and multiple function networks. These findings may provide more evidence for the existing view of hierarchical organization in reward processing.