Retinal ganglion cell-inner plexiform layer, white matter hyperintensities, and their interaction with cognition in older adults

• Published in: Frontiers in aging neuroscience Vol. 15; p. 1240815
• Main Authors: Wang, Ruilin, Wu, Xinmao, Zhang, Zengyi, Cao, Le, Kwapong, William Robert, Wang, Hang, Tao, Wendan, Ye, Chen, Liu, Junfeng, Wu, Bo
• Format: Journal Article
• Language: English
• Published: Lausanne Frontiers Research Foundation 15.11.2023; Frontiers Media S.A
• Subjects: Acuity; Adults; Aging; Aging Neuroscience; cognition; Cognition & reasoning; Cognitive ability; Dementia; Disease; Hippocampus; Hospitals; Magnetic resonance imaging; Medical imaging; MRI; Neuroimaging; older adults; Older people; Population; Retina; retina thickness; Substantia alba; Vascular diseases; Visual acuity; white matter hyperintensity; China
• ISSN: 1663-4365; 1663-4365
• DOI: 10.3389/fnagi.2023.1240815

Purpose We explored the interaction of optical coherence tomography (OCT) parameters and white matter hyperintensities with cognitive measures in our older adult cohort. Methods This observational study enrolled participants who underwent a comprehensive neuropsychological battery, structural 3-T brain magnetic resonance imaging (MRI), visual acuity examination, and OCT imaging. Cerebral small vessel disease (CSVD) markers were read on MR images; lacune, cerebral microbleeds (CMB), white matter hyperintensities (WMH), and enlarged perivascular spaces (EPVS), were defined according to the STRIVE standards. Retinal nerve fiber layer (RNFL) and ganglion cell-inner plexiform layer (GCIPL) thicknesses (μm) were measured on the OCT tool. Results Older adults with cognitive impairment (CI) showed lower RNFL ( p  = 0.001), GCIPL ( p  = 0.009) thicknesses, and lower hippocampal volume ( p  = 0.004) when compared to non-cognitively impaired (NCI). RNFL ( p  = 0.006) and GCIPL thicknesses ( p  = 0.032) correlated with MoCA scores. GCIPL thickness ( p  = 0.037), total WMH ( p  = 0.003), PWMH ( p  = 0.041), and DWMH ( p  = 0.001) correlated with hippocampal volume in our older adults after adjusting for covariates. With hippocampal volume as the outcome, a significant interaction ( p  < 0.05) between GCIPL and PWMH and total WMH was observed in our older adults. Conclusion Both GCIPL thinning and higher WMH burden (especially PWMH) are associated with hippocampal volume and older adults with both pathologies are more susceptible to subclinical cognitive decline.