Relationships between Retinal Nerve Fiber Layer and Brain Changes in Age-Related Macular Degeneration

Abstract

 

Purpose

 

People with AMD, compared to sighted peers, have lower performance on cognitive tests, a two-fold increased risk of incident Alzheimer's, and steeper cognitive decline, yet it remains unknown whether AMD and cognitive impairment reflect similar neurodegenerative processes affecting both brain and retina. We use data from our ongoing prospective cohort study to investigate whether AMD patients with worse cognition or more altered brain connectivity have worse retinal damage measured by nerve fiber thickness (NFL).

 

Methods

 

We present interim data from 12 subjects (average age 73.3 ±7.4, 66% female), all with NVAMD in at least one eye, who underwent resting state MRI and these neurocognitive tests: Rapid Distinction Test (RDT), Wechsler Logical Memory, Item Recall, Verbal Fluency (Phonemic and Semantic), and the Fuld Object-Memory test. From macular Spectral Domain Optical Coherence Tomography (OCT), three-dimensional OCT image cubes were manually segmented for NFL thickness. Resting state MRI images were analyzed for strength of functional connectivity (FC) between relevant brain regions. Strong FC implies that metabolic activity of two regions rises and falls in concert. We assessed two visual processing centers and two regions from the default mode network (DMN), a brain network in which connectivity is strong in healthy persons but deteriorates with age and brain dysfunction. Two-way correlations were constructed between relevant variables.

 

Results

 

Half the subjects had NVAMD OU (we used NFL from the eye with better visual acuity in these subjects). Average binocular visual acuity was 0.37 (±0.21) LogMAR, and length of symptoms was 7.2 (±10) years. We found no relationship between NFL thickness and any of 8 cognitive test scores or FC measures.

 

Conclusions

 

The lack of significant correlations may indicate that NFL thickness is an imperfect marker of neuronal damage in NVAMD due to non-linear changes in thickness from fluid accumulation to scarring as the disease progresses. Alternatively, our findings may suggest that cognitive changes in AMD do not arise from concurrent neurodegenerative processes but are the result of reduced cognitive reserve or stimulation in the context of vision loss. To help distinguish between these possible explanations, future work should examine whether other biomarkers of retinal damage correlate to worse cognition or brain changes in AMD.