Multimodal imaging of glymphatic dysfunction and retinal vascular changes as biomarkers for Alzheimer's disease.

BACKGROUND

Alzheimer's disease (AD) is associated with retinal vascular changes, while the relationships among the glymphatic system, retina vasculature, and cognition in AD remain unclear.

METHODS

Thirty-one AD patients and 24 healthy controls (HC) were enrolled. Glymphatic function was assessed using perivascular space (PVS) scoring in the hippocampus (Hip), basal ganglia, and white matter, along with free water content and diffusion tensor imaging along the perivascular space index. Retinal vascular parameters (RVPs) included fractal dimension (FD), vascular density (VD), and mean arterial/venous caliber, analyzed across subregions scaled by the optic papilla diameter (PD). Group differences in glymphatic system and RVPs between AD and HC groups were examined, and correlations with cognitive performance were assessed. Bagged Trees classifiers were used to distinguish AD from HC based on these features.

RESULTS

AD patients exhibited significantly increased Hip_PVS compared to HC. FD and VD (total, 0.5-1.0 PD, 1.0-1.5 PD) were significantly lower in AD and negatively correlated with Hip_PVS. Mini-Mental State Examination and Montreal Cognitive Assessment scores were negatively correlated with Hip_PVS, and positively correlated with FD, VD (total), VD (0.5-1.0 PD), and VD (1.0-1.5 PD). Mediation analysis revealed that Hip_PVS mediated the relationship between RVPs and cognitive impairment. The retinal combination model, incorporating FD, VD (total), VD (0.5-1.0 PD), and VD (1.0-1.5 PD), achieved an AUC of 0.768 for distinguishing AD from HC, increasing to 0.877 when combined with Hip_PVS.

CONCLUSIONS

Hip_PVS-mediated glymphatic dysfunction may link retinal vascular changes to cognitive decline in AD, enhancing diagnostic effectiveness with combined retinal and glymphatic biomarkers.