Investigation To The Causality Between Age At First Birth,Number Of Live Births And Prevalent Coronary Heart Disease Risk

PART1 Association between age at first birth,number of live births and coronary heart disease risk:a meta-analysisObjective:We conducted a meta-analysis to explore the association between age at first birth,number of live births and prevalent risk of coronary heart disease(CHD).Methods:The keywords“age at first birth”/"number of live births"+"coronary heart disease"were searched in the PubMed and Web of Science.We collected prospective studies published before December 31rd,2018,which were focused on women.Duplicate,irrelevant or data missing articles were excluded.Cochrane and I~2 were used for heterogeneity test,and when the test result was statistically significant then we used subgroup analysis or meta-regression to clarify the sources of heterogeneity.Furthermore,fixed or random pooling model was selected for main effects according to the result of I~2.Besides,funnel plot and egger test were used for publication bias.Moreover,sensitivity analysis was performed by excluding any literature and observing the main effects of the remaining literatures.Results:We analyzed 7 prospective studies(total n=1448710)focused on the relationship between AFB and CHD risk.Compared to the higher age at first birth(AFB)group,women in the lower AFB group had higher risk for CHD(HR=1.199,95%CI:1.130–1.268).Heterogeneity test was not significant(I~2=1.3%,P=0.415)and there was no publication bias(P=0.864).We conducted a meta-analysis of 6 prospective studies(total n=988367)focused on the relationship between number of live births and CHD.Compared to the women with fewer live births,the women with more live births had higher risk for CHD(HR=1.059,95%CI:1.020–1.098).Heterogeneity test was not significant(I~2=0.0%,P=0.956)and there was no publication bias(P=0.423).Conclusions:The lower AFB and more live births were risk factors for CHD.PART2 The network mendelian randomization analysis for the association between age at first live birth,number of live births and coronary heart disease riskObjective:We conducted the network mendelian randomization(MR)analysis to explore the causal relationship between reproductive behavior and coronary heart disease(CHD),and to find out potential mediator between reproductive behaviors and CHD.Methods:We conducted two-sample mendelian randomization analysis based on genome-wide association studies(GWAS),and then explored possible pathway from age at first live birth and number of live births to CHD by network mendelian randomization.The data from GWAS study include:CARDIoGRAMplusC4D database for CHD data(case=22233,control=64762);Sociogenome database for reproductive behaviors data(age at first birth n=251151,the number of live births n=343072);GIANT database for body mass index(BMI),hip circumference,waist circumference and waist-hip ratio data(n=339224).ENGAGE database for telomere length data(n=37684),MAGIC database for diabetes mellitus and fasting blood-glucose data(case=766558,control=46186),and Global Lipids Genetics database(n=94595)for low density lipoprotein,total cholesterol and triglyceride data.Results:1.The MR analysis between reproductive behaviors and CHD risk using10 SNPs instrument variables(IVs)showed,higher AFB contributed to lower risk for CHD(OR=0.868,95%CI:0.809–0.930)and MR-egger analysis was not significant(P=0.212).The relationship between number of live births and CHD was not significant(P=0.363).2.The MR analysis between AFB and risk factor of CHD using 10 SNPs as IVs showed,one year earlier in AFB contributed to 0.06 kg/m~2 decrease in BMI(?=-0.060,95%CI:-0.087–-0.033,P<0.001)and MR-egger analysis was not significant(P=0.071).However,there was no significant correlation between AFB and other risk factors for CHD(hip circumference P=0.180,waist circumference P=0.465,waist-hip ratio P=0.062,telomere length P=0.582,fasting blood-glucose P=0.611,type 2 diabetes P=0.234,low density lipoprotein P=0.973,total cholesterol P=0.356,and triglyceride P=0.701).3.The MR analysis between BMI and CHD using 25 SNPs as IVs showed,higher BMI was a risk factor for CHD(OR=1.471,95%CI:1.285–1.684,P<0.001),and MR-egger analysis was not significant(P=0.081).4.There was no reverse causal relationship between age at first birth and BMI(P=0.182).Conclusions:Lower AFB was suggested to be a risk factor for CHD,and BMI could be a potential mediator in the pathway between AFB and CHD.