Relationship Between Retinal Layer Thickness and Genetic Susceptibility to Age-Related Macular Degeneration in Asian Populations

PURPOSE: For OCT retinal thickness measurements to be used as a prodromal age-related macular degeneration (AMD) risk marker, the 3-dimensional (3D) topographic variation of the relationship between genetic susceptibility to AMD and retinal thickness needs to be assessed. We aimed to evaluate individual retinal layer thickness changes and topography at the macula that are associated with AMD genetic susceptibility. DESIGN: Genetic association study. PARTICIPANTS: A total of 1579 healthy participants (782 Chinese, 353 Malays, and 444 Indians) from the multiethnic Singapore Epidemiology of Eye Diseases study were included. METHODS: Spectral-domain OCT and automatic segmentation of individual retinal layers were performed to produce 10 retinal layer thickness measurements at each ETDRS subfield, producing 3D topographic information. Age-related macular degeneration genetic susceptibility was represented via single nucleotide polymorphisms (SNPs) and aggregated via whole genome (overall) and pathway-specific age-related macular degeneration polygenic risk score (PRS(AMD)). MAIN OUTCOME MEASURES: Associations of individual SNPs, overall PRS(AMD), and pathway-specific PRS(AMD) with retinal thickness were analyzed by individual retinal layer and ETDRS subfield. RESULTS: CFH rs10922109, ARMS2-HTRA1 rs3750846, and LIPC rs2043085 were the top AMD susceptibility SNPs associated with retinal thickness of individual layers (P < 1.67 × 10(−3)), all at the central subfield. The overall PRS(AMD) was most associated with thinner L9 (outer segment photoreceptor/retinal pigment epithelium complex) thickness at the central subfield (β = −0.63 μm; P = 5.45 × 10(−)(9)). Pathway-specific PRS(AMD) for the complement cascade (β = −0.53 μm; P = 9.42 × 10(−)(7)) and lipoprotein metabolism (β = −0.05 μm; P = 0.0061) were associated with thinner photoreceptor layers (L9 and L7 [photoreceptor inner/outer segments], respectively) at the central subfield. The mean PRS(AMD) score was larger among Indians compared with that of the Chinese and had the thinnest thickness at the L9 central subfield (β = −1.00 μm; P = 2.91 × 10(−)(7); R(2) = 5.5%). Associations at other retinal layers and ETDRS regions were more heterogeneous. CONCLUSIONS: Overall genetic susceptibility to AMD and the aggregate effects of the complement cascade and lipoprotein metabolism pathway are associated most significantly with L7 and L9 photoreceptor thinning at the central macula in healthy individuals. Photoreceptor thinning has potential to be a prodromal AMD risk marker, and topographic variation should be considered. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.