Skip to main content
Fig. 4 | BMC Medical Research Methodology

Fig. 4

From: Bayesian Mendelian randomization with an interval causal null hypothesis: ternary decision rules and loss function calibration

Fig. 4

MR model with 360 instrumental variables, (\(\mathrm IV_{1}\), \(\mathrm IV_{2}\), ..., \(\mathrm IV_{360}\)), used in our illustrative study to assess the causal effect of BMI on juvenile myocardial infarction (JMI). Observed values of 5 potential confounders (sex, smoking status, alcohol consumption, cocaine consumption and age), were included in the model (see main text). The model assumes there are no associations between IVs and JMI other than those mediated by BMI

Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us