Quantitative Assessment of the Retina Using OCT and Associations with Cognitive Function

Purpose: To determine the association of retinal thickness with cognitive function in Japanese persons. Design: Cross-sectional, population-based survey. Participants: A total of 1293 Japanese persons aged 65 to 86 years who resided in the Saku area in the Japan Public Health Center–Based Prospective Study participated in the eye and mental health screening. Methods: Participants underwent comprehensive ophthalmic assessment, including fundus photography, measurement of intraocular pressure, and determination of refraction status. We assessed the thickness of the macular retinal nerve fiber layer (mRNFL), ganglion cell-inner plexiform layer (GC-IPL), and ganglion cell complex (GCC, which includes the retinal nerve fiber layer and GC-IPL), and the full thickness in the macula and peripapillary retinal nerve fiber layer (ppRNFL) using spectral-domain (SD) OCT. Cognitive tests consisted of the Mini-Mental State Examination, Wechsler Memory Scale Revised logical memory I/II subtest, clock drawing test, and Clinical Dementia Rating Scale. These were used to designate the participants in the following 3 groups: Normal, those with mild cognitive impairment (MCI), and those with dementia. Multivariable logistic regression models were used to analyze associations between retinal thickness and cognitive function after adjusting potential confounding factors. Main Outcome Measures: Association of retinal thickness with cognitive function. Results: Among the 1293 potential subjects, 114 were excluded for a diagnosis of depression, 64 were excluded for retinal disease, and 140 were excluded for scanning errors or suboptimal OCT images. The remaining 975 participants (mean age, 73.2 years) were included in this analysis. Significant differences were found in the 3 groups in all layers and GCC thickness, but not in ppRNFL thickness. After adjusting for age, sex, educational status, and refraction, full macular thickness and GCC thickness were inversely associated with the presence of dementia, but ppRNFL thickness was not. Furthermore, GC-IPL, GCC, and full macular thicknesses were all associated with the presence of dementia in the inferior sectors. Conclusions: Macular thickness was associated with the presence of dementia, but ppRNFL was not. Our results suggest that OCT measurements of the macula could be superior to those of the ppRNFL in assessing neurodegenerative changes and a potentially useful diagnostic biomarker of cognitive function.