蜜芽app

Age-related macular degeneration (AMD) is the most common cause of blindness among the elderly in European countries. In a new report, researchers used metabolomics to identify novel circulating metabolites associated with early-intermediate and late AMD and their interaction with genetic and lifestyle risk factors.¹

The risk of AMD is associated with genetic variants and lifestyle factors such as smoking and diet. The interaction between genotype, phenotype, and lifestyle or environment may be reflected in the levels of metabolites in fluids or in tissues.² Consequently, high-throughput analysis of metabolites (metabolomics) can provide evidence of disease pathways, biomarkers of disease prognosis, or response to therapy.² 

Researchers in the EYE-RISK Consortium previously showed that the levels of 60 metabolites in blood were associated with AMD.³ They also found that levels of 34 metabolites were associated with seven AMD-associated gene variants.³ 鈥淗owever, it is still unclear how these metabolites reflect genetic and lifestyle factors, and whether they are biomarkers for a certain stage of disease,鈥� the authors of the new report wrote.¹

In this new follow-up study, the authors characterized plasma metabolites from individuals with AMD and their association with genetic and lifestyle factors. 

Study design and patient characteristics

The researchers obtained their data from the European Eye Epidemiology consortium and EYE-RISK project. The studies included were three population-based studies and one case-control study from The Netherlands and one case-control study from Germany.¹

Data sets included fundus photographs and levels of 146 metabolites measured with a nuclear magnetic resonance array from plasma. Genetic risk scores were determined for 49 AMD-associated variants. Lifestyle risk scores were calculated from smoking status and daily fruit, vegetable, and fish consumption assessed with a questionnaire. 

The metabolomic analysis was performed in 5923 individuals. Results included 3850 people without AMD, while early-intermediate AMD was diagnosed in 1114 individuals and late AMD in 959. Genotype and lifestyle data were available for 3836 individuals.

The AMD groups differed from controls at baseline in several respects. Individuals with early-intermediate AMD were older than controls (74.7 卤 8.1 years vs 72.4 卤 7.2 years; P < .001). They also had a slightly lower mean BMI (25.5 vs 26.0 kg/m²; P < .001), and fewer were current smokers (11.9% vs 14.9%; P = .04) compared with controls. 

Individuals with late AMD were also older than controls (77.7 卤 8.2 years vs 72.4 卤 7.2 years; P < .001). Unlike in the early-intermediate group, more individuals in the late AMD group were current smokers compared with controls (15.2% vs 14.9%; P = .017). Fewer of the late AMD group had hypertension (34.3% vs 43.7%; P < .001) compared with controls. 

Metabolomic profiles by AMD stage

The researchers found that the metabolomic profiles depended on the stage of AMD. In the early-intermediate AMD group, the levels of 61 metabolites in plasma were significantly different compared with controls. Most (94%) were involved in lipid pathways. For example, total cholesterol in VLDL was reduced (odds ratio [OR] 0.9; false discovery rate [FDR] P = 1.1 x 10^-4), but total cholesterol in HDL and HDL2 (ORs 1.2; FDR P < 7.6 x 10^-5) and apolipoprotein A1 (OR 1.2; FDR P < 1.3 x 10^-3) were increased in early-intermediate AMD compared with controls. 

Among the non-lipid metabolites, albumin, glycoprotein acetyls, the amino acid isoleucine, and citrate were detected at significantly lower levels in early-intermediate AMD compared with controls (ORs 0.9; FDR P < 1.4 x 10^-2).

Late AMD was associated with alterations in the levels of eight metabolites. Six out of eight of these metabolites were also significantly different between late AMD and early-intermediate AMD. The non-essential amino acid tyrosine and the essential amino acids histidine, leucine, phenylalanine, and valine were reduced in plasma from individuals with late AMD compared to controls (ORs 0.6-0.8, FDR P < 1.5 x 10^-3). In contrast, the ketone bodies acetoacetate and 3-hydroxybutyrate were increased compared with controls (OR 1.4; FDR P < 2.5 x 10^-8). 

Only citrate was associated with both stages of AMD. There was a reduction in citrate in early-intermediate AMD (OR 0.9 FDR P<1.4 x 10^-2) as well as citrate reduction in late AMD (OR 0.8; FDR P < 2.7 x 1^-3) compared with controls.¹

The changes in these metabolites suggested that late AMD was associated with perturbations in several metabolic pathways. These included the biosynthesis of tyrosine, the metabolism of phenylalanine, and the synthesis and degradation of ketone bodies. Although some of the changes detected could indicate low blood glucose, there were no differences in mean fasting glucose or the frequency of diabetes between individuals with late AMD and controls.

Integrating risk factors and metabolomics

A total of 47 metabolites were associated with the genetic risk score for AMD. For individual genes, the high-risk ARMS2/HTRA1 genotype was associated with increased acetoacetate, phenylalanine, and citrate (尾s, 0.03-0.05; FDR P < 5.0 x 10^-3). The genetic risk score related to the extracellular matrix, however, was not associated with any metabolite.

The levels of 30 metabolites were associated with the total lifestyle risk score, particularly monounsaturated fatty acids and glycoprotein acetyls related to inflammation (尾s, 0.07-0.08; FDR P < 1.4 x 10^-4). Changes in the levels of eight metabolites, such as lower levels of citrate (尾, 鈥�0.09; FDR P < 3.1 x 10^-6), were associated with smoking. Vegetable intake alone was associated with the levels of 49 metabolites. These included higher levels of large HDLs (尾s, 0.16-0.20; FDR P < .05) and lower levels of cholesterols, fatty acids, apolipoproteins, VLDL subclasses, and 3-hydroxybutyrate (尾s, 鈥�0.18 to 鈥�0.30; FDR P < 4 x 10^-2). 

Based on the eight metabolites associated with late AMD, the researchers calculated a metabolic risk score to ascertain whether these metabolites were associated with AMD independent of genetic and lifestyle risk factors. After adjustment for the genetic risk score, lifestyle risk score, age, sex, and study location, the metabolic risk score was still associated with greater odds of developing late AMD (OR 1.6; 95% confidence interval [CI], 1.4-1.9; P < .001).

The researchers also calculated how much of the metabolic risk score mediated the association of genetic or lifestyle risk score with late AMD. They estimated that the metabolic risk score accounted for 5.3% (95% CI, 2.6%-9.0%) of the genetic risk score and 19.5% (95% CI, 7.0%-88.0%) of the lifestyle risk score.

Working toward a systems biology of AMD

The study confirms the group鈥檚 earlier report that a high fraction of metabolites associated with AMD are in lipid pathways, which is consistent with a role for systemic dyslipidemia in the development of drusen.^3,4

The approach also revealed previously undescribed metabolite associations. In early-intermediate AMD, the researchers identified novel positive associations with apolipoprotein A1 and HDLs and negative associations with albumin and glycoprotein acetyls. Although only eight metabolites were associated with late AMD, four of these were novel: a positive association with ketone bodies acetoacetate and 3-hydroxybutyrate and a negative association with the amino acids valine and histidine.

The researchers also concluded that circulating metabolite levels reflected some of the genetic and lifestyle risk for AMD. While their estimates showed that differences in diet and smoking accounted for approximately one-fifth of the lifestyle risk, they wrote that 鈥渁 large proportion of lifestyle risk remains unexplained.鈥� In addition, more research is needed to understand whether changes in metabolite levels independent of genetic and lifestyle factors have a direct role in the development of AMD.

The authors described the limitations of their report. While the study included the largest number of individuals for metabolic profiling to date, there were still few incident AMD cases. Lifestyle factors, rare genetic variants, and other possible confounding factors they did not assess could affect the results. 

Another limitation is that their nuclear magnetic resonance platform measured far fewer metabolites than mass spectrometry approaches. 鈥淯ntargeted investigations using robust mass spectrometry or NMR will likely result in more comprehensive metabolic associations,鈥� they wrote.

鈥淭he integration of metabolomic, genetic, and lifestyle factors in a systems biology approach holds great potential for elucidating biological processes involved in AMD,鈥� the authors concluded.

Published: September 12, 2023