Causal associations between type 1 diabetes mellitus and cardiovascular diseases: A Mendelian randomization study

Background The presence of type 1 diabetes mellitus (T1DM) has been demonstrated to pose an increased risk for developing cardiovascular diseases (CVDs). However, the causal relationships between T1DM and CVDs remain unclear due to the uncontrolled confounding factors and reverse causation bias of the observational studies. Aim To investigate the causal relationships between T1DM and seven major CVDs, including myocardial infarction (MI), heart failure (HF), coronary artery disease (CAD), atrial fibrillation (AF), coronary atherosclerosis, peripheral atherosclerosis, and stroke, using a two-sample bidirectional Mendelian randomization (MR) method. Method We selected genetic instruments for T1DM and the seven CVDs from the largest available genome-wide association studies (GWAS) of European ancestry for the MR analysis. Three complementary methods: inverse variance weighted (IVW), weighted median, and MR-Egger were used for the MR estimates. The potential pleiotropic effects were assessed by MR-Egger intercept and MR-PRESSO global test. Additionally, multivariable MR (MVMR) analysis was performed to examine whether T1DM has independent effects on CVDs with adjustment of potential confounding factors. Moreover, a two-step MR approach was used to assess the potential mediating effects of these factors on the causal effects between T1DM and CVDs. Results Causal effects of T1DM on peripheral atherosclerosis (odds ratio [OR]=1.06, 95% confidence interval [CI]: 1.02–1.10; p = 0.002)] and coronary atherosclerosis (OR=1.03, 95% CI: 1.01–1.05; p = 0.001) were found. The results were less likely to be biased by the horizontal pleiotropic effects (both p values of MR-Egger intercept and MR-PRESSO Global test > 0.05). In the following MVMR analysis, we found the causal effects of T1DM on peripheral atherosclerosis and coronary atherosclerosis remain significant after adjusting for a series of potential confounding factors. Moreover, we found that hypertension partly mediated the causal effects of T1DM on peripheral atherosclerosis (proportion of mediation effect in total effect: 11.47%, 95% CI: 3.23%–19.71%) and coronary atherosclerosis (16.84%, 95% CI: 5.35%–28.33%). We didn’t find significant causal relationships between T1DM and other CVDs, including MI, CAD, HF, AF, or stroke. For the reverse MR from CVD to T1DM, no significant causal relationships were identified. Conclusion This MR study provided evidence supporting the causal effect of T1DM on peripheral atherosclerosis and coronary atherosclerosis, with hypertension partly mediating this effect.