Impacts of Chronic Kidney Disease on Retinal Neurodegeneration: A Cross-Cohort Analysis

Abstract

To assess the cross-sectional and longitudinal associations between chronic kidney disease (CKD) and ganglion cell-inner plexiform layer (GCIPL) thickness in a UK Biobank population and a Chinese cohort. Prospective observational cohort study and cross-sectional study. This study included 23,014 without neurodegenerative diseases individuals from the UK Biobank, and three years of annual follow-up data of 2,197 individuals from a Chinese cohort. Three groups were defined by estimated glomerular filtration rate (eGFR) based on serum creatinine classifying CKD severity as no CKD, Mild-CKD, and moderate-to-severe CKD (MS-CKD). The GCIPL thickness, measured using optical coherence tomography, was analyzed through linear regression over time to determine its decline rate in micrometers per year. Linear regression models were used to assess the correlation between renal function and both the baseline GCIPL thickness and the GCIPL decline rate. The cross-sectional analysis in a largely White population showed that poorer renal function negatively correlated with GCIPL thickness with the mean thinner of 0.15 μm (95% CI: -0.30 to -0.01; p=0.042) in Mild-CKD, and 0.83 μm thinner (95% CI -1.34 to -0.32; p=0.001) in MS-CKD, compared to that of controls without CKD. Longitudinal analysis in Chinese cohort showed that the GCIPL decreased more rapidly in persons with poorer renal function. After correcting for all confounding factors, the rate of GCIPL thinning was 0.30 μm/year (95% CI -0.41 to -0.19; p<0.001) more in the Mild-CKD group and 0.52 μm/year (95% CI -0.79 to -0.26; p<0.001) more in the MS-CKD group, compared to controls without CKD. This relationship sustained in individuals with diabetes or hypertension. Poor renal function was associated with a lower baseline GCIPL thickness in the UK population and a faster decline rate in Chinese participants. However, the detailed underlying mechanisms still need further exploration.