Cerebellar Gray Matter Atrophy and Altered Structural Covariance Networks Are Associated with Cognitive Impairment in Type 2 Diabetes Mellitus

Meixia Long¹Jinyun Xue²Dingwen Hu¹Hui Chu¹Shijun Qiu^3*

• ¹Guangzhou University of Chinese Medicine, Guangzhou, China
• ²Other
• ³The First Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, China

Background The prevalence of type 2 diabetes mellitus (T2DM) is steadily increasing, with central nervous system complications commonly manifesting as mild cognitive impairment and dementia. However, the neuropathophysiological mechanisms underlying T2DM-related cognitive dysfunction remain poorly understood. Method This study employed voxel-based morphometry (VBM) and seed-to-voxel structural covariance network (SCN) analyses to investigate alterations in cerebellar gray matter volume (GMV) and SCNs in T2DM, as well as their associations with cognitive performance. Intergroup differences were assessed using two-sample t-tests with Gaussian random field correction. Results VBM analysis revealed significant GMV reductions in the bilateral cerebellar Crus I, left lobule I–IV, left Crus II, left lobule IX, and right lobule VIIb in T2DM participants. Seed-to-voxel SCN analysis further demonstrated decreased covariance between the left Crus I and the left middle temporal gyrus, middle occipital gyrus, and angular gyrus, along with increased covariance between the left lobule I–IV and the right caudate nucleus. Correlation analysis revealed that GMV of the left Crus I was positively associated with Clock Drawing Test scores, while GMV of the right Crus I correlated positively with auditory verbal learning test (AVLT) scores. In addition, GMV of the right lobule VIIb was positively associated with both AVLT and grooved pegboard test (GPT) scores, and GMV of the left lobule IX was positively correlated with GPT scores. With respect to network integrity, reduced SCN connectivity between the left Crus I and the default mode network (DMN) was negatively correlated with AVLT and the color word test performance, whereas enhanced SCN connectivity between the left lobule I–IV and the right caudate nucleus was negatively correlated with AVLT scores, while positively correlated with trail making test–A performance. Conclusions By integrating VBM and SCN approaches, this study demonstrated that cerebellar GMV atrophy and abnormal structural covariance in T2DM were closely associated with cognitive dysfunction. These findings highlight the role of disrupted cerebro-cerebellar connectivity in the pathophysiology of T2DM-related cognitive impairment.