Altered Interhemispheric Functional Connectivity in Patients With Diabetic Retinopathy: A Resting-State Functional MRI Study.

Abstract

Objective: Cognitive impairment is a prevalent complication among patients with diabetes mellitus. It tends to be more prominent in patients with diabetic retinopathy (DR) compared with patients with diabetes without DR (NDR). However, the functional connectivity (FC) between bilateral cerebral hemispheres in both remains poorly understood. This study aimed to investigate altered brain connectivity in patients with DR and NDR.

Subjects and methods: We selected 26 patients with DR, 30 patients with NDR, and 30 healthy controls (HCs) to participate in resting-state functional magnetic resonance imaging (rs-fMRI) and high-resolution T1-weighted structural scans. We employed the DPABI toolbox in MATLAB to preprocess the acquired images and applied voxel-mirrored homotopic connectivity (VMHC) and FC analysis methods to estimate differences among the 3 groups. The patients also underwent neuropsychological assessment scales. We utilized partial correlation analysis to explore the associations between aberrant connections and clinical variables as well as neuropsychological characteristics in patients with DR. Receiver operating characteristic (ROC) analysis was conducted to assess the diagnostic performance of VMHC values in distinct brain regions for differentiating DR patients from NDR patients.

Results: The results showed significantly altered VMHC values across the 3 groups, including bilateral lingual gyrus (LING_B), superior temporal gyrus (STG_B), and postcentral gyrus (PoCG_B). Significant differences in FC values were found across the LING_B, right cuneus (CUN_R), STG_R, PoCG_B, right precentral gyrus (PreCG_R), right precuneus (PCUN_R), and middle temporal gyrus (MTG_L) among the 3 groups. Moreover, a negative correlation was noted between the VMHC values of LING_B and disease duration in patients with DR. Positive correlations were detected between FC values in PoCG_B and fasting blood glucose (FBG) levels. Furthermore, ROC analysis of the VMHC values demonstrated that combining all the differential regions achieved the highest area under the curve of 0.826.

Conclusions: Significant alterations in VMHC and FC may reflect the underlying neuropathology of cognitive dysfunction in DR and NDR. These altered connectivity patterns could serve as neuroimaging biomarkers, offering insights into the early diagnosis and intervention of cognitive impairments in DR patients.