Abstract
Purpose :
The retina and optic nerve head are the only parts of the nervous system that can be directly visualized. Previous studies have demonstrated a relationship between retinal thickness and cognitive aging and Alzheimer’s Disease (AD), with lower retinal nerve fibre layer (RNFL) thickness predicting greater cognitive decline and AD symptoms over time. The purpose of this study is to test if retinal thickness parameters are associated with performance in cognitive tests in a large general adult population.
Methods :
We investigated optical coherence tomography (OCT) data and detailed results of cognitive function tests in a large population-based cohort (UK Biobank), a multisite cohort study of UK residents aged 40 to 69 years. The cognitive tests include verbal and mathematical reasoning tests, as well as problem-solving and logical thinking tasks. The OCT data include segmented macular layer thicknesses, including the ganglion cell layer (GCL), retinal nerve fibre layer (RNFL), ganglion cell-inner plexiform layer (GCIPL), and total retinal thickness.
To assess the relationship between the two, we built linear regression models, controlling for age and sex, with results of 17 different cognitive tests as the outcome and 12 OCT retinal thickness parameters as predictors. Significant results were obtained after Bonferroni correction (0.05/(12x16) = 0.00026)
Results :
Complete data from 54,665 UK Biobank participants were available. Significant associations were found between fluid intelligence and reasoning (the capacity to solve problems that require logic and reasoning ability, independent of acquired knowledge) and average RNFL thickness (p=6.22x10-06). The average RNFL parameter showed only low to moderate correlation (Pearson’s R in the range of =-0.11 to +0.33) with other retinal parameters measured. We observed no other pairwise association with any other retinal thickness parameters that remained significant after Bonferroni correction.
Conclusions :
This study provided evidence that retinal structure may be associated with specific higher cognitive functions. Our results demonstrated a stronger association between RNFL thickness and numerical as well as logical cognitive skills as opposed to verbal and more creative aspects of cognition. More work will be needed in the future to replicate and further extend these initial findings.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.