Association of individual-based morphological brain network alterations with cognitive impairment in type 2 diabetes mellitus

Objective To investigate the altered characteristics of cortical morphology and individual-based morphological brain networks in type 2 diabetes mellitus (T2DM), as well as the neural network mechanisms underlying cognitive impairment in T2DM.Methods A total of 150 T2DM patients and 130 healthy controls (HCs) were recruited in this study. The study used voxel- and surface-based morphometric analyses to investigate morphological alterations (including gray matter volume, cortical thickness, cortical surface area, and localized gyrus index) in the brains of T2DM patients. Then two methods, Jensen-Shannon divergence-based similarities (JSDs) and Kullback-Leibler divergence-based similarities (KLDs), were used to construct individual morphometric brain networks based on gray matter volume, to discover altered features of the topological network and extract abnormal key brain regions. Subsequently, partial correlation analyses were performed to explore the relationship between clinical biochemical indices, neuropsychological test scores, and altered cortical morphology and network indices.Results Brain regions with reduced gray matter volume and cortical thickness in T2DM patients were mainly concentrated in the frontal lobe, temporal lobe, parietal lobe, anterior cingulate gyrus, insula, lingual gyrus, and cerebellar hemispheres. The global attributes of the Individual-based morphological brain network were significantly reduced (Cp, Eloc, sigma), with an increase in the nodal efficiency of the hippocampus and the nodal local efficiency of the anterior cingulate gyrus, and the nodal local efficiency of the parahippocampal gyrus and transverse temporal gyrus were reduced. There was a correlation between these node attributes and cognitive scale scores.Conclusion This study demonstrated that patients with T2DM exhibit generalized cortical atrophy and damage to individual morphologic brain networks. It also identified overlapping and cognitively relevant key brain regions, primarily within the limbic/paralimbic network (especially the hippocampus and cingulate gyrus), which may serve as imaging markers for identifying cognitive deficits in T2DM. These findings offer new insights into the neural network mechanisms underlying T2DM-associated brain damage and cognitive impairment.