Limits to Causal Inference based on Mendelian Randomization: A Comparison with Randomized Controlled Trials

Abstract

"Mendelian randomization" refers to the random assortment of genes transferred from parent to offspring at the time of gamete formation. This process has been compared to a randomized controlled trial of genetic variants. This could greatly aid observational epidemiology by potentially allowing an unbiased estimate of the effects of gene products on disease outcomes. However, studies utilizing Mendelian randomization to estimate effects of gene products on outcomes should be interpreted with caution. In this paper, the authors discuss some of the challenges facing epidemiologists in the analysis and interpretation of Mendelian randomization studies, particularly those that become apparent when the analogy with randomized controlled trials is closely examined. The authors conclude that Mendelian randomization is a powerful addition to etiologic research tools. However, care must be taken, because drawing valid causal inferences from its application depends upon more extensive assumptions than are required in randomized controlled trials.

Response to Comment on Ye et al. The Association Between Circulating Lipoprotein(a) and Type 2 Diabetes: Is It Causal? Diabetes 2014;63:332–342

“…We acknowledge that more generally the MR approach may pose challenges and limitations in addressing causal inference (6). However, we appropriately applied MR in our work, and our findings are in keeping with other prevailing evidence thus far suggesting that a causal association with Lp(a) levels is unlikely.…”

Identifying the odds ratio estimated by a two‐stage instrumental variable analysis with a logistic regression model

“…The outcome (Y) is a binary variable, taking value 1 with probability π 1 , which is a logistic function of the risk factor and covariate V. Although G will be thought of later as an IV, it could here be any covariate dividing the population independently of V into strata with different mean risk factor levels. (11) The conditional log odds ratio (CLOR) conditional on V is:…”

“…In model (11), this would occur if . However, when this is not true, the adjusted two-stage estimate is attenuated [42].…”

Mendelian randomisation studies of type 2 diabetes: future prospects

“…In this research framework, the inter-relationship and consistency of associations among genetic variants that encode or regulate a biomarker, the expression or circulating levels of the biomarker, and disease risk may characterise the true magnitude of the relationship between the biomarker and risk of disease ( Fig. 1) [5][6][7]. Reproducible evidence for a statistical association between a genetic variant and biomarker provides the basis for a Mendelian randomisation study.…”

“…In this context, Herder et al [3], based on specificity and directional consistency of associations among genetic variants, biomarker and disease risk, provide some evidence for a causal link between circulating levels of macrophage migration inhibitory factor and risk of type 2 diabetes. However, much more robust statistical evidence is required to make causal inferences using this approach [6,[10][11][12].…”