Abstract
Purpose :
Age-related macular degeneration (AMD) is the leading cause of blindness among the elderly, yet untreatable in 90% of the cases. Metabolic dysregulation (e.g. high HDL) may be part of the complex aetiology of AMD, and there is a gap in understanding of pathogenesis that could be addressed with metabolomics. To identify plasma metabolites associated with or causal for AMD, we performed a cross-sectional study based on metabolomics data from the TwinsUK cohort with a hypothesis-free approach.
Methods :
Metabolomics data included over 900 plasma metabolites (Metabolon platform) measured in in 262 AMD cases (mainly early/intermediate AMD) and 981 controls (95 and 96% female, mean age 69.1±6.88 and 63.0±6.94, mean BMI 26.0±4.86 and 26.5±4.96, respectively) from the TwinsUK cohort, graded for the presence or absence of AMD using optical coherence tomography (OCT) retinal scans. After removing the metabolites with >20% missing values, mixed effects logistic regression model was applied to each of the remaining 702 metabolites. Bonferroni correction (<7.1x10-5) for multiple testing was used as the threshold for significance. Metabolite levels were then compared in 119 discordant twin pairs using paired t-tests. FInally, Mendelian randomization tests assessed the effect of varying metabolite levels on AMD.
Results :
With mixed logistic regression models, 2 metabolites were significantly and inversely associated with AMD after multiple testing correction: N-acetyl-β-alanine (p=1.40x10-5) and 1-stearoyl-2-arachidonoyl-GPC (18:0/20:4, p=2.45x10-5). In a discordant twin model, both metabolites had significantly lower mean levels in AMD twins compared to their twin controls (mean difference=-0.081, p=0.0012 and -0.075, p=0.0037, respectively). However, Mendelian randomization for the only metabolites suitable instruments were available (1-stearoyl-2-arachidonoyl) did not demonstrate causality for AMD.
Conclusions :
This study found associations for two metabolites that were corroborated in discordant twins’ analysis and were consistent with previous research suggesting dysregulation of β-alanine and glycerophospholipid metabolic pathways in AMD compared to healthy individuals. Future work will need to independently replicate these results and further identify other AMD biomarkers to establish new therapeutic avenues and pinpoint individuals at high risk for AMD.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.