Mendelian Randomization of Blood Lipids for Coronary Heart Disease

Study Questions:

Are high-density lipoprotein (HDL) and triglycerides causally related to coronary heart disease (CHD)?


Weighted allele scores were developed based on single nucleotide polymorphisms (SNPs) with established associations with HDL, triglycerides, and low-density lipoprotein (LDL). For each trait, two scores were established: 1) an unrestricted score, which included all independent SNPs associated with the lipid trait identified from a prior meta-analysis, and 2) a restricted score, filtered to remove any SNPs also associated with either of the other two lipid traits. Mendelian randomization meta-analyses were conducted in 17 studies, including 62,199 participants with 12,099 CHD events.


Both the unrestricted and restricted allele scores for LDL were associated with CHD. For HDL, the unrestricted allele score was associated with CHD (odds ratio [OR], 0.53 per 1 mmol/L higher HDL), but neither the restricted allele score nor the unrestricted HDL allele score adjusted for triglycerides, LDL, or statin use showed a robust association. For triglycerides, the unrestricted allele score and the restricted allele score were both associated with CHD (OR, 1.62 and 1.61) per 1-log unit increment.


The authors concluded that these findings support a causal effect of triglycerides on CHD risk, but a causal role for HDL cholesterol, though possible, remains less certain.


Low HDL and high triglyceride levels are associated with increased cardiovascular risk; however, therapeutic interventions that raise HDL and/or lower triglycerides have not demonstrated cardiovascular risk reduction. Thus, unlike the case with LDL, the causal relationship of HDL and triglycerides with cardiovascular disease remains unclear. Mendelian randomization studies have proven useful to clarify causal relationships between biomarkers of interest and clinical outcomes. The current study represents an improved analysis that uses multiple independent SNPs along with lipid-lowering medication data to identify causal relationships between LDL and triglyceride levels with CHD risk. However, evidence for causality is not strong for HDL in this study. Additional studies with SNPs that affect different properties of HDL may be useful to address this controversy.

Clinical Topics: Diabetes and Cardiometabolic Disease, Dyslipidemia, Atherosclerotic Disease (CAD/PAD), Lipid Metabolism, Nonstatins, Novel Agents, Statins

Keywords: Odds Ratio, Coronary Artery Disease, Risk Reduction Behavior, Hydroxymethylglutaryl-CoA Reductase Inhibitors, Lipids, Risk Factors, Lipoproteins, LDL, Dyslipidemias, Polymorphism, Single Nucleotide, Biological Markers, Phenotype, Troponin I, Cardiovascular Diseases, Cholesterol, HDL, Triglycerides

< Back to Listings