Mendelian randomization to assess causal effects of blood lipids on coronary heart disease: lessons from the past and applications to the future

Purpose of review Mendelian randomization is a technique for judging the causal impact of a risk factor on an outcome from observational data using genetic variants. Although evidence from Mendelian randomization for the effects of major lipids and lipoproteins on coronary heart disease (CHD) risk has been around for a relatively long time, new data resources and new methodological approaches have given fresh insight into these relationships. The lessons from these analyses are likely to be highly relevant when it comes to lipidomics and the analyses of lipid subspecies whose biology is less well understood. Recent findings Although analyses of low-density lipoprotein cholesterol and lipoprotein(a) are unambiguous as there are genetic variants that associate exclusively with these risk factors and have well understood biology, analyses for triglycerides, and high-density lipoprotein cholesterol (HDL-c) are less clear. For example, a subset of genetic variants having specific associations with HDL-c are not associated with CHD risk, but an allele score including all variants associated with HDL-c does associate with CHD risk. Recently developed methods, such as multivariable Mendelian randomization, Mendelian randomization-Egger, and a weighted median method, suggest that the relationship between HDL-c and CHD risk is null, thus confirming experimental evidence. Robust methods for Mendelian randomization have important utility for understanding the causal relationships between major lipids and CHD risk, and are likely to play an important role in judging the causal relevance of lipid subspecies and other metabolites measured on high-dimensional phenotyping platforms.