Abstract
PURPOSE: Differentiating between Alzheimer's disease (AD) and frontotemporal dementia (FTD) can be challenging due to overlapping cognitive and behavioral manifestations. Evidence regarding non-invasive and early-stage biomarkers remains limited. Our aim was to identify retinal biomarkers for the risk of AD and FTD in populations without dementia and explore underlying brain structural mechanisms. METHODS: We included a total of 3,0573 UK Biobank participants without dementia, ocular disorders, and diabetes who underwent baseline retinal optical coherence tomography (OCT) imaging. Cox proportional hazards models were used to estimate the associations between macular OCT parameters and the risk of AD and FTD. Mediation analysis was used to explore the underlying mechanisms affected by brain structures. RESULTS: The mean age at recruitment was 55.27, and 46.10% of the participants were male. During a mean follow-up of 9.15 ± 2.59 years, 148 patients with AD and eight patients with FTD were identified. Reduced thickness of the ganglion cell-inner plexiform layer (GC-IPL) at baseline was associated with an increased risk of AD (HR, 1.033; 95% CI, 1.001-1.066; P = 0.044), while thinner retinal pigment epithelial in the inner superior subfield at baseline was associated with an elevated risk of FTD (HR, 1.409; 95% CI, 1.060-1.871; P = 0.018). Structurally abnormal visual pathways, including cortical and subcortical gray matter volumes, as well as white matter integrity, mediated the association between the GC-IPL thickness and AD risk. CONCLUSION: Our findings provide preliminary empirical support for a relationship between prodromal changes in retinal layers and a higher risk of AD or FTD, suggesting that macular OCT may serve as a non-invasive, sensitive biomarker of high-risk years before the onset of dementia.