For a study, researchers expected that the effect of blood lipid-related metabolites on primary open-angle glaucoma (POAG) would vary depending on the lipoprotein particles and lipid sub-fractions studied. Therefore, the investigators looked at the relationships between blood levels of lipoprotein particles and lipid subfractions, and POAG. Individuals enrolled for the population-based Epidemiology of Eye Disease study’s baseline visit (n=8,503). All individuals were subjected to comprehensive standardized ocular and systemic exams. A nuclear magnetic resonance metabolomics platform was used to quantify a total of 130 blood lipid-related compounds. The analyses were divided into two parts. First, they used regression analysis followed by Bayesian network modeling to determine if and which lipid-related metabolites were directly correlated with POAG. Second, they used 2-sample Mendelian randomization (MR) analysis to see if there is any causal association between the detected lipid-related metabolites and POAG. They conducted GWAS on high-density lipoprotein (HDL) 3 cholesterol (after normal inverse transformation) and utilized the top variants related to HLD3 cholesterol as instrumental variables (IVs) in the MR analysis.
POAG affected 175 (2.1%) of the participants. First, a logistic regression model revealed that total HDL3 cholesterol was inversely linked with POAG, but phospholipids in very big HDL were favorably associated. Using Bayesian network analysis, the study group discovered that only total HDL3 cholesterol was directly associated with POAG (odds ratio [OR], 0.72 per 1 standard deviation increase in HDL3 cholesterol; 95% CI, 0.61–0.84), regardless of age, gender, intraocular pressure (IOP), body mass index (BMI), education level, systolic blood pressure, axial length, or statin medication). Using 5 GWAS IVs and the inverse variance weighted MR technique, they discovered that higher levels of HDL3 cholesterol were linked with a lower risk of POAG (OR, 0.91; 95% CI, 0.84–0.99, P = 0.021). Other MR techniques produced consistent OR estimates, including weighted median, model-based estimator, and contaminated mixing approaches. POAG was not linked with any of the regular lipids (blood total, HDL, or low-density lipoprotein [LDL] cholesterol). Overall, the findings indicated that the connection between HDL3 cholesterol and POAG may be causative and specific and that dysregulation of cholesterol transport might play a role in POAG pathogenesis.
Reference:www.aaojournal.org/article/S0161-6420(21)00731-4/fulltext
