Bidirectional relationship between type 2 diabetes mellitus and coronary artery disease: Prospective cohort study and genetic analyses

Abstract

Abstract Background: While type 2 diabetes mellitus (T2DM) is considered a putative causal risk factor for coronary artery disease (CAD), the intrinsic link underlying T2DM and CAD is not fully understood. We aimed to highlight the importance of integrated care targeting both diseases by investigating the phenotypic and genetic relationships between T2DM and CAD. Methods: We evaluated phenotypic associations using data from the United Kingdom Biobank (N = 472,050 ). We investigated genetic relationships by leveraging genomic data conducted in European ancestry for T2DM, with and without adjustment for body mass index (BMI) (T2DM: N case/N control = 74,124 /824,006; T2DM adjusted for BMI [T2DMadjBMI]: N case/N control = 50,409 /523,897 ) and for CAD (N case/N control = 181,522 /984,168 ). We performed additional analyses using genomic data conducted in multiancestry individuals for T2DM (N case/N control = 180,834 /1,159,055). Results: Observational analysis suggested a bidirectional relationship between T2DM and CAD (T2DM→CAD: hazard ratio [HR] = 2.12, 95% confidence interval [CI]: 2.01–2.24; CAD→T2DM: HR = 1.72, 95% CI: 1.63–1.81). A positive overall genetic correlation between T2DM and CAD was observed (r g = 0.39, P = 1.43 × 10–75), which was largely independent of BMI (T2DMadjBMI–CAD: r g = 0.31, P = 1.20 × 10–36). This was corroborated by six local signals, among which 9p21.3 showed the strongest genetic correlation. Cross-trait meta-analysis replicated 101 previously reported loci and discovered six novel pleiotropic loci. Mendelian randomization analysis supported a bidirectional causal relationship (T2DM→CAD: odds ratio [OR] = 1.13, 95% CI: 1.11–1.16; CAD→T2DM: OR = 1.12, 95% CI: 1.07–1.18), which was confirmed in multiancestry individuals (T2DM→CAD: OR = 1.13, 95% CI: 1.10–1.16; CAD→T2DM: OR = 1.08, 95% CI: 1.04–1.13). This bidirectional relationship was significantly mediated by systolic blood pressure and intake of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors, with mediation proportions of 54.1% (95% CI: 24.9%–83.4%) and 90.4% (95% CI: 29.3%–151.5%), respectively. Conclusion: Our observational and genetic analyses demonstrated an intrinsic bidirectional relationship between T2DM and CAD and clarified the biological mechanisms underlying this relationship.