Associations between morphometric similarity network and brain gene expression in type 2 diabetes mellitus

Abnormal functional and structural connectivity in brain networks is commonly observed in patients with type 2 diabetes mellitus (T2DM) and is often associated with cognitive impairment. In this study, we employed the Morphometric Similarity Network (MSN) method, which is based on seven morphometric features derived from structural and diffusion magnetic resonance imaging, to investigate structural differences in the brains of T2DM patients by quantifying structural similarities between brain regions. Globally, morphometric similarity (MS) was significantly reduced in T2DM patients. Regionally, MS was decreased in the left sensorimotor network and the right salience/ventral attention network, while it was increased in the bilateral visual network. Notably, the increased MS in the bilateral visual network was negatively correlated with memory function in T2DM patients. Furthermore, using the Allen Human Brain Atlas (AHBA; http://human.brain-map.org), which provides transcriptome data from postmortem adult brains, we linked MSN changes to regional gene expression patterns. Transcription-neuroimaging association analyses identified 298 genes whose expression was significantly spatially correlated with T2DM-related MSN abnormalities. Many of these genes are involved in biological processes such as central nervous system development and neurotransmitter transmission, offering valuable molecular and cellular insights into MS abnormalities and cognitive decline in T2DM. These findings shed light on the neural and genetic mechanisms underlying T2DM-related brain changes and cognitive impairment, providing new perspectives for future research and potential therapeutic approaches.