Non-alcoholic fatty liver disease is associated with structural covariance network reconfiguration in cognitively unimpaired adults with type 2 diabetes.

Type 2 diabetes (T2D) is often accompanied by non-alcoholic fatty liver disease (NAFLD), both of which are related to brain damage and cognitive impairment. However, cortical structural alteration and its relationship with metabolism and cognition in T2D with NAFLD (T2NAFLD) and without NAFLD (T2noNAFLD) remain unclear. The brain MRI scans, clinical measures and neuropsychological test were evaluated in 50 normal controls (NC), 73 T2noNAFLD, and 58 T2NAFLD. The cortical thickness and graph theory properties of structural covariance network was calculated. Statistical analyses included one-way analysis of covariance with post hoc, partial correlation and mediation analysis. The nonparametric permutation test was performed to evaluate differences in topological properties of structural covariance network. We found T2NAFLD group had worse glucose and lipid profiles, more obesity and more severe insulin resistance, and poorer working memory compared to T2noNAFLD and NC. T2D patients demonstrated increase in cortical thickness compared to NC, but no difference between the two T2D groups. The structural covariance network integration decreased in T2D patients, with T2NAFLD exhibiting more obvious network reconfiguration at node level. Cortical thickness mediated the relationship between post-prandial glucose, waist-hip ratio, and working memory. The findings suggest that cortical thickening may be a compensatory response to reduced network integration, with NAFLD exacerbating regional structural network changes in T2D. This research advances our understanding of how these metabolic comorbidities contribute to cognitive decline, potentially guiding future therapeutic strategies for T2D patients with and without NAFLD.