Causal Association Between Exposure Factors And Coronary Artery Disease In Patients With Type 2 Diabetes Mellitus:A Two-sample Mendelian Randomization Study

ObjectivesThere were limited randomized trials for lifestyle and health behavior management in the care of patients with type 2 diabetes mellitus(T2DM)with comorbid coronary artery disease(CAD)which led to lack of direct causal evidence.Besides,observational studies suggested that psychosocial factors and some blood parameters were associated with CAD in the patients with T2 DM.Whether these relationships are causal remains unclear,due to confounding bias and reverse causality.We aim to investigate the causal association between exposure factors including lifestyle factors,anthropometric parameters,psychosocial factors,serum mineral elements and blood parameters,and CAD in patients with T2 DM using large-sample gene sequencing data.MethodsStatistical analyses were performed using two-sample Mendelian randomization(MR)method.The instrumental variables for each exposure factor were constructed based on publically available data from several large-scale gene-sequencing consortiums,such as genome-wide association studies(GWAS)conducted by Sleep Disorder KP consortium,GIANT consortium,GIS consortium,GSCAN consortium and based on the data of Cohorts for Heart and Aging Research in Genomic Epidemiology(CHARGE)and UK Biobank(UKBB).The association between selected genetic variants and CAD in individuals with T2 DM were obtained from a recent UKBB-based GWAS,which included 14,453 individuals of European descent with T2DM(3,700 cases of CAD).In addition,considering the large number of blood parameters,we first conducted a single-center retrospective study.A total of 2,107 patients with T2 DM who were hospitalized in the department of cardiology and received coronary angiography from September 2017 to September 2020 were successively included,and relevant medical record data and examination results were extracted.According to the results of coronary angiography,they were divided into CAD group(1,772 patients)and non-CAD group(385 patients).Multivariate Logistic regression analyses were used to screen out the blood parameters that were related to incidence of CAD in the patients with T2 DM.Then MR method was used to investigate whether these blood parameters were causal risk factors.We combined the causal effect estimate of each single nucleotide polymorphism(SNP)using random-effect inverse variance weighted(IVW)method as main results.Maximum likelihood,fixed-effect IVW,simple median,weighted median method and leave-one-out analysis were performed to test the robustness of main results.Funnel plots and MR-Egger intercept tests were used to detect potential pleiotropy.Then,considering the effect of T2 DM itself on CAD,various statistical models were used to explore the causal association between causal risk factors and T2 DM,and scatter plots were drawn for visual display.In addition,MR mediating effect analyses were conducted to further explore the mediating effects of T2 DM on causal risk factors and CAD,and the corresponding effect-mediated proportions were calculated using the coefficient difference method.Then,genetic instrument for CAD was constructed to investigate the reverse causal effect of CAD on sedentary behavior,sleep status and depression.Finally,a specialized web tool called m Rnd was used to calculate the statistical power of each MR analysis.ResultsIn terms of lifestyle factors,sedentary behavior,insomnia and smoking were causal risk factors for CAD in patients with T2 DM,while long sleep duration was a protective factor.Specifically,in individuals with T2 DM,CAD risk was reduced by 37% for each60-minute increase in genetically predicted daily sleep time.The risk of CAD increased by about 40% for every 1.5 hours of daily sedentary time.Individuals with insomnia had an 18% increased risk of developing CAD,and those who once smoked regularly had an approximately 40% increased risk.We found no evidence for the causal association between moderate to vigorous physical activity,alcohol consumption,or coffee consumption and CAD in patients with T2 DM.Subgroup analyses suggested that sleep duration less than 7 hours per day was associated with CAD in patients with T2 DM,while sleep duration longer than 9 hours was not associated with CAD.The effects of different types of sedentary behaviors were not consistent.Sedentary television watching and driving,rather than computer use,were causally associated with CAD in patients with T2 DM.Further analyses showed that sedentary behavior,insomnia,and smoking were all causally associated with T2 DM,and T2 DM mediated about 50% and 30% effects of sedentary behavior and smoking on CAD,respectively.Reverse MR analyses found no causal effect of CAD on sedentary behavior,sleep duration,or insomnia.In terms of anthropometric parameters,gene-predicted adult height was a causal protective factor for CAD in patients with T2 DM,while body mass index(BMI),body fat rate(BFR),and visceral adipose tissue(VAT)were causal risk factors.Specifically,for every 1 Standard Deviation(SD)increase in height in patients with T2 DM,the risk of CAD decreased by 5%;For each SD increase in BMI,BFR and VAT,the risk increased by 17%,58% and 35%,respectively.Further analyses suggested that BMI,BFR and VAT were all causally associated with T2 DM,and T2 DM mediated about 60%,60% and 20%effects of these indexes on CAD,respectively.As for psychosocial factors,depression is a causal risk factor for CAD in patients with T2 DM.In addition,this association was affected by the severity of depression.The risk of CAD in patients with T2 DM increased by about 30% for those with major depression and T2 DM played a mediating role between major depression and CAD.However,no causal effect of mild depression was found.Although both high intelligence level and long years of education seemed to be correlated with decreased risk of CAD in patients with T2 DM,they did not reach statistical significance.Similarly,no causal association was found between anxiety disorder and neuroticism,and CAD in patients with T2 DM.In addition,reverse MR analysis found no causal effect of CAD on depression.In terms of serum mineral elements,serum iron concentration was a causal protective factor for CAD in patients with T2 DM,while serum selenium concentration was a causal risk factor.Specifically,the risk of CAD decreased by 18% for every 1 SD increase in serum iron concentration in patients with T2 DM,and the risk increased by 25% for every1 SD increase in serum selenium concentration.We found no causal association between serum calcium,magnesium,zinc and copper levels,and CAD in patients with T2 DM.Further analyses suggested that ferritin concentration and transferrin saturation in patients with T2 DM were also causally associated with CAD.Furthermore,we found a causal effect of serum selenium on T2 DM,but not a mediating effect of T2 DM on serum selenium and CAD.For the blood parameters,multivariate Logstic regression analysis of observational study suggested that Hb A1 c,albumin,and serum uric acid levels were independent risk factors for CAD in patients with T2 DM,while hemoglobin concentration was a protective factor.Further subgroup analyses showed that the effect of serum uric acid on CAD was only found in the older patients(≥65 years old)with T2 DM.MR study suggested that gene-determined low density lipoprotein cholesterol(LDLC),total cholesterol(TC),triglyceride(TG),serum uric acid and growth differentiation factor 15(GDF-15)levels were causal risk factors for CAD in patients with T2 DM.High density lipoprotein cholesterol(HDL-C)and hemoglobin concentrations exerted protective effects on CAD.Specifically,the risk of CAD increased by 51%,52% and 38%for every 1 SD increase in TC,LDL-C,and TG levels in patients with T2 DM,respectively.The risk of CAD decreased by 17% for every 1 SD increase in HDL-C.The risk of CAD increased by 13% for each 1mg/d L increase in serum uric acid,increased by 38% for every 1 SD increase in GDF-15,and decreased by 16% for each SD increase in hemoglobin concentration in patients with T2 DM.In addition,we found the causal effect of HDL-C and serum uric acid on T2 DM,and the effect of HDL-C on CAD was partially mediated by T2 DM.No mediating effect of T2 DM on serum uric acid and CAD was found.We found no convincing evidence for the causal effect of blood glucose,Hb A1 c,albumin and homocysteine levels on CAD in patients with T2 DM.ConclusionFor the lifestyle intervention and mental health management in patients with T2 DM,in addition to smoking cessation and weight control,more attention should be paid to reducing sedentary behavior(especially sedentary television watching and driving),improving sleep(especially sleep duration less than 7 hours and insomnia)and regular mental health screening(especially depression disorder)to prevent CAD and reduce coronary events.Serum selenium,serum iron,hemoglobin,serum uric acid and GDF-15 levels might be potential therapeutic targets.Large-sample multi-center randomized controlled trials are needed for further validation.