Abstract
Purpose :
Metabolomics, the high-throughput analysis of a wide range of metabolites in body fluids, has great potential to uncover biomarkers and pathways that contribute to disease pathophysiology. Studies into metabolomics in age-related macular degeneration (AMD) are scarce and limited by relatively small sample sizes. We therefore performed the largest metabolomics study in AMD to date, to identify associations between amino acids, glycolysis measures, ketone bodies, fatty acids and lipoprotein subclasses with AMD.
Methods :
A total of 7,753 samples from five cohorts (EUGENDA, Rotterdam Study, ALIENOR, MARS, CORRBI) were analyzed by NMR-based metabolomics at Nightingale Health to quantify 146 individual metabolite measurements and 79 derivative values (ratios or totals). Data was quality controlled, followed by cohort-specific logistic regression and random effects meta-analysis to determine the association of each metabolite and derivative with AMD, adjusting for age, gender, and cohort-specific confounders. The significance threshold was determined to be 0.05 after Benjamini-Hochberg correction for multiple testing.
Results :
Measurements of 2,307 AMD patients and 4,294 control individuals passed all quality control checks. Meta-analysis identified 64 measured metabolites and 19 derivative values significantly associated with AMD. Extra-large (XL-) and large (L-) HDL subclasses were associated with increased risk for AMD (e.g. phospholipids in XL-HDL OR=1.21, 95% CI=1.12-1.30, pFDR=6.9x10-5), while small (S-) and medium (M-) HDL subclasses and all VLDL subclasses were associated with decreased risk of AMD (e.g. cholesterol esters in L-VLDL OR=0.85, 95% CI=0.76-0.94, pFDR=0.008). The ratio of dietary omega-6 fatty acids to total fatty acids was associated with increased risk of AMD (OR=1.13, 95% CI=1.05-1.21, pFDR=0.004), and two dietary amino acids (Phe, Leu) were associated with decreased risk of AMD (OR=0.91, 95% CI=0.85-0.97, pFDR=0.015 and OR=0.91, 95% CI=0.85-0.98, pFDR=0.031, respectively).
Conclusions :
The most significant associations were observed for HDL and VLDL lipoprotein subclasses, underscoring an important role for lipid metabolism in AMD pathogenesis. Interestingly, directions of effect differed among HDL subclasses, indicating that HDL particle size and composition are relevant in AMD. Associations with dietary fatty acids and amino acids highlight the role of nutrition in AMD.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.