The Relationship Between Inner Retinal Thickness, Cognition, and Frailty in Community‐dwelling Older Adults: findings from a 6‐year follow‐up study

Background Thinning of the inner retinal thickness, including ganglion cell‐inner plexiform layer (GC‐IPL) and retinal nerve fiber layer (RNFL), has been related to cognitive decline. However, the relationship between GC‐IPL/RNFL thickness, physical frailty, and cognitive frailty remains unclear. Therefore, this longitudinal study aims to investigate the association of these retinal markers with cognition, frailty, and cognitive frailty. Method This is a 7‐year cohort study including 221 community‐dwelling elders in the ongoing Taiwan Initiate for Geriatric Epidemiological Research (TIGER) between 2015 and 2017 and received biennial assessments for cognition and physical frailty three consecutive times. The global and domain‐specific (memory, attention, executive function, and language) cognition were assessed using the Taiwanese version of the Montreal Cognitive Assessment and a series of neuropsychological tests. Physical frailty was assessed by modified Fried frailty phenotypes. The generalized linear mixed model was used to analyze the association between GC‐IPL/RNFL thickness and global/domain‐specific cognitive frailty adjusted for age, sex, years of education, apolipoprotein E ε4 status carriers, cigarette smoking status, presence of depressive symptoms, hypertension or diabetes mellitus, years of follow‐up, and practice effect. Result The performance of global cognition decreased as the mean GC‐IPL thickness of bilateral eyes deviated from the sample mean (76.6 µm) (quadratic GC‐IPL: β = ‐0.4×10−2; 95% confidence interval: ‐0.7×10−2 to ‐0.2×10−2). Similar associations were also found for logical memory with RNFL/GC‐IPL thickness. No significant association was observed between RNFL/GC‐IPL thickness and physical frailty. An increased odds of impaired logical memory‐physical frailty was found for GC‐IPL (quadratic GC‐IPL: adjusted odds ratio = 1.005; 95% confidence interval: 1.001 to 1.005), but no significant association was found between cognitive frailty and RNFL thickness. Conclusion Change in the GC‐IPL thickness was associated with impaired cognition and poor logical memory‐frailty performance. The inner retinal thickness may serve as a biomarker for cognition or cognitive frailty in non‐demented elders.