📝 SummarXiv

Abstract

Objective: Differences in WM structural connectivity were sought by comparing graph measures from young women with normal weight ($D_\mathrm{norm}$) and obesity ($D_\mathrm{over}$), studying FA-based inter-tract correlation matrices. Methods: Undirected binary adjacency matrices were created from WM inter-tract correlation matrices. Sixteen global graph measures were compared between the $D_\mathrm{norm}$ and $D_\mathrm{over}$ groups. At the nodal level, four graph measures were compared. Optimal community structures were calculated using the Louvain community detection algorithm. This was done considering all network, as well as the left and right hemispheres separately. Results: At the global level and considering all network, the $D_\mathrm{over}$ group had significantly reduced global efficiency compared to the $D_\mathrm{norm}$ group. Considering only the right hemisphere, none of the graph measures were significantly different between the two groups. Studying only the left hemisphere, the $D_\mathrm{over}$ group obtained significantly reduced measures compared to the $D_\mathrm{norm}$ group in five global graph measures. At the nodal level, the $D_\mathrm{over}$ group obtained significantly reduced graph measures in WM tracts when studying all network and the left hemisphere separately. These tracts were mainly involved in the reward network. The optimal community structures calculated were consistent with the results reported. Conclusion: The $D_\mathrm{over}$ group showed altered and reduced structural connectivity in WM tracts compared to the $D_\mathrm{norm}$ group. The reported tracts are involved in reward processing, inhibitory control, executive decision-making, and cognitive processing. In addition to obtaining results consistent with those reported in the literature, it was additionally observed that the results were lateralized to the left hemisphere.