Using Mendelian randomisation to assess causality in observational studies

Abstract

ObjectiveMendelian randomisation (MR) is a technique that aims to assess causal effects of exposures on disease outcomes. The paper aims to present the main assumptions that underlie MR, the statistical methods used to estimate causal effects and how to account for potential violations of the key assumptions.MethodsWe discuss the key assumptions that should be satisfied in an MR setting. We list the statistical methodologies used in two-sample MR when summary data are available to estimate causal effects (ie, Wald ratio estimator, inverse-variance weighted and maximum likelihood method) and identify/adjust for potential violations of MR assumptions (ie, MR-Egger regression and weighted Median approach). We also present statistical methods and graphical tools used to evaluate the presence of heterogeneity.ResultsWe use as an illustrative example of a published two-sample MR study, investigating the causal association of body mass index with three psychiatric disorders (ie, bipolar disorder, schizophrenia and major depressive disorder). We highlight the importance of assessing the results of all available methods rather than each method alone. We also demonstrate the impact of heterogeneity in the estimation of the causal effects.ConclusionsMR is a useful tool to assess causality of risk factors in medical research. Assessment of the key assumptions underlying MR is crucial for a valid interpretation of the results.

Intake of processed meat, but not sodium, is associated with risk of colorectal cancer: Evidence from a large prospective cohort and two-sample Mendelian randomization

“…Third, we observed wide CIs under MR-Egger method in MR analyses, which may suggest underpower. However, MR-Egger method is often underpowered in studies and thus used as secondary sensitivity analysis [62], but the results were qualitatively consistent with the primary analysis of inverse-variance weighted method. Fourth, we did not specifically adjust for macronutrients (e.g., fat, protein, carbohydrates); instead, we adjusted for intakes of food (such as fruit, vegetables, red meat, fish, cereal), as they are the major sources of macronutrients as well as micronutrients.…”

LDL-C plays a causal role on T2DM: a Mendelian randomization analysis

“…Therefore, incorporating the natural randomization inherent in the generation of genetic individuality, as can be accomplished with use of the Mendelian randomization (MR) method, provides a useful complement to traditional epidemiological studies [23,24]. MR uses genetic variants as an instrumental variable (IV) to estimate and assess casual relationships between exposure of interest and outcomes [25][26][27][28] (Figure 1).…”

Body Shape Phenotypes And Breast Cancer Risk: A Mendelian Randomization Analysis.

“…[17][18][19] Second, we used the Cochran's Q test to evaluate heterogeneity of causal effects from each variant, with a p-value < 0.05 indicating statistical signi cance. 20 In case of heterogeneity, likely indicating pleiotropy 21 , a random-effects IVW MR analysis was used. 22 This method relaxes the third assumption as the total pleiotropic effect of a single genetic variant no longer needs to be null but assumes zero mean between all the genetic variants (i.e.…”

“…violation of one or more MR assumptions). 21,24 Both tests provide valid MR estimates in the presence of overall directional pleiotropy but suffer from reduced power. 23,24 Finally, we also used the MR Pleiotropy RESsidual Sum and Outlier (MR-PRESSO) method to identify and remove any outlying variants 26 and leave-one genetic variant-out analysis were used to assess whether any association was driven by speci c genetic variants.…”