Overview of Mendelian Randomization Analysis
10.4078/jrd.2020.27.4.241
- Author:
Young Ho LEE
1
Author Information
1. Division of Rheumatology, Department of Internal Medicine, Korea University College of Medicine, Seoul, Korea
- Publication Type:Review Article
- From:Journal of Rheumatic Diseases
2020;27(4):241-246
- CountryRepublic of Korea
- Language:English
-
Abstract:
he inference of causality from observational evidence may be problematic, as observational studies frequently include confounding factors or reverse causation for the identification of associations between exposure and outcome. Thus, in observational studies, the association between a risk factor and a disease of interest may not be causal. A randomized controlled trial (RCT) is considered the gold standard, because it has the best possibility to establish a relationship between a risk factor and an outcome. However, RCTs cannot always be performed, because they can be costly, impractical, or even unethical. One of the alternatives is to perform Mendelian randomization (MR) experiments that are similar to RCTs in terms of study design.The MR technique uses genetic variants related to modifiable traits/exposures as tools to detect causal associations with outcomes. MR can provide more credible estimates of the causal effect of a risk factor on an outcome than those obtained in observational studies by overcoming the limitations of observational studies. Therefore, MR can make a substantial contribution to our understanding of complex disease etiology. MR approaches are increasingly being used to evaluate the causality of associations with risk factors, because well-performed MR studies can be a powerful method for exploring causality in complex diseases. However, there are some limitations in MR analyses, and an awareness of these limitations is essential to interpret the results. The validity of results from MR studies depends on three assumptions that should be carefully checked and interpreted in the context of prior biological information.
