Retinal Parameters, Cortical Microinfarcts and Functional Cognitive Impairment

AbstractBackgroundQuantitative changes in retinal vessels and thinning of optic nerves have been associated with subclinical and clinical age‐related brain pathologies. However, data on the association between both retinal vascular and neuronal parameters with cortical cerebral microinfarcts and how these factors jointly contribute to functional cognitive impairment is lacking. We investigate the association of retinal vascular and neuronal changes with cortical CMIs on 3T MRI and to explore their interaction in relation to functional cognitive impairment in a memory‐clinic population.MethodA total of 538 participants (mean age:72.6±7.8 years) were included. Retinal vascular parameters (caliber, tortuosity, fractal dimension) were measured from retinal fundus photographs using a semi‐automated computer‐assisted program. Retinal nerve fiber layer (RNFL) and ganglion cell‐inner plexiform layer (GC‐IPL) thicknesses were obtained from spectral domain‐optical coherence tomography. Cortical CMIs were defined as hypointense lesions on T1‐weighted images, <5mm in diameter, restricted to the cortex, and perpendicular to the cortical surface. Functional cognitive impairment was assessed on Clinical Dementia Rating Sum‐of‐Boxes (CDR‐SoB) score.ResultLarger venular caliber (Rate ratios (RR):1.21, 95%CI:1.03‐1.42), increased venular fractal dimension (RR:1.25, 95%CI:1.07‐1.46), increased venular tortuosity (RR:1.08, 95%CI:1.02‐1.15) and thinner GC‐IPL (RR:1.22, 95%CI:1.12‐1.32) were associated with increasing CMI counts. A significant interaction term was found between venular fractals, venular tortuosity, thinner GC‐IPL and RNFL and cortical CMIs (p<0.05) on CDR‐SoB scores. Among individuals in highest tertile of retinal parameters, a significant interaction was observed between venular tortuosity (RR:1.14, 95%CI:1.09‐1.19, p‐interaction<0.001), GC‐IPL (RR:1.23, 95%CI:1.13‐1.34, p‐interaction<0.001) and RNFL (RR:1.48, 95%CI:1.31‐1.68, p‐interaction=0.002) with CMIs on CDR‐SoB scores.ConclusionWe showed that retinal vascular and neuronal parameters are associated with cortical CMIs independent of risk factors. Moreover, there was an interaction between retinal parameters and cortical CMIs on CDR‐SoB. Further studies may be warranted to evaluate the potential clinical utility of retinal parameters in risk prediction for CMIs.