Fig. 1
From: Diabetes mellitus and idiopathic pulmonary fibrosis: a Mendelian randomization study
Study design. (A) Univariable MR to identify the causal association between T1D or T2D and IPF, which involves IVs and requires three assumptions: (I) strong association between IVs and the exposure of interest; (II) no direct relationship between IVs and outcome, with their effect only indicated through exposure; (III) independence of IVs from any confounding factors affecting both exposure and outcome. (B) Multivariable MR to evaluate if T1D remains causal given T2D. MR, Mendelian randomization; T1D, type 1 diabetes; IPF, idiopathic pulmonary fibrosis; T2D, type 2 diabetes; IVs, instrumental variables