Relationship Between Longitudinal Trajectories Of Metabolic Syndrome-related Indexes And Coronary Heart Disease In Health Management Population

Background:Coronary heart disease(CHD)has become an urgent disease problem in public health with the highest proportion of deaths worldwide,leading to 8.93 million deaths and accounting for 16.0%of all causes of death in 2017.Global disease burden showed that years of life lost(YLLs)for CHD ranked first among all diseases,and the disease burden of CHD in developing countries accounts for 60%of that in the world There are a large number of cardiovascular disease patients in China,the prevalence rate of the total population remains high.Among them,the number of CHD patients has exceeded 10 million,in addition,the mortality rate of CHD continued unabated.One common sense is that CHD has caused severe socioeconomic losses and shortened lives and needs to be early prevented.As the most common type of chronic cardiovascular disease,CHD is caused by coronary atherosclerosis through various pathogenic pathways.Chest pain is a common symptom for CHD,however,most patients have no typical clinical symptoms.Indeed,early atherosclerosis can occur at childhood,with long-term accumulation and synergy of multiple risk factors,plaque in the arterial wall accumulates and leads to angiostenosis.In order to reduce potential hazards and adverse outcomes caused by CHD,it is imperative to early identify and evaluate the multi-source causes of CHD to prevent the development of CHDAs the metabolic syndrome-related indexes,overweight and obesity,hypertension,low-density lipoprotein cholesterol(LDL-C),triglycerides(TG),high-density lipoprotein cholesterol(HDL-C)and fasting blood glucose(FPG)were commonly examined in health management population,and their associations with CHD were widely studied,but the analysis objects of traditional research methods were cross-sectional data or longitudinal data of the difference between two points,which can only simply characterize or represent the level difference of this index.However,due to the long-term and chronic progression of CHD,risk factors at a certain time point or a short time period cannot fully capture the longitudinal changing patterns of risk factors and hard to reflect the development of CHD,while the longitudinal trajectories of metabolic syndrome-related index may be closely associated with the pathological process of CHD.Object:GBTM was adopted to explore the relationship between longitudinal trajectories of overweight and obesity,hypertension,LDL-C,HDL-C,TG,FPG and CHD risk in health management cohorts,which will provide epidemiological evidence of multiple longitudinal MetS related factors for CHD risk assessment,and provide scientific proof for etiology prevention of CHD and long-term multi-factorial health maintenance.Methods:Based on the large-scale routine medical examination cohort of Affiliated Hospital of Jining Medical University,under the condition of quality control of baseline variables,we included individuals with at least 4 physical examination records of overweight and obesity,hypertension,LDL-C,HDL-C,TG,and FPG,and excluded physical examination observations during and after diagnosis of CHD for CHD patients Using group-based trajectory modeling(GBTM)method,we investigated the univariate trajectories of overweight and obesity and log FPG indexes,as well as multivariate trajectories of high SBP and high DBP,and multivariate trajectories of LDL-C,HDL-C and log TG.We further described in detail the statistical distribution and CHD prevalence according to the trajectory group status,and then plotted the proportion of trajectory group membership in each age category.CHD incidence densities among different trajectory group membership were compared,CHD cumulative risk curves for MetS related index trajectories were plotted.Using two-dimensional grids to group four trajectory variables,the distribution heat maps and the incidence density heat maps of CHD patients were drawn.Cox proportional hazard regression model was used to explore the association between trajectory membership of various variables and CHD risk,compared with that between traditional forms of variables and CHD risk.Finally,we defined the multiple trajectory sum score and further investigated the relationship between score groups and CHD risk.Results:1.Trajectories for overweight and obesity were divided into low stable group(group 1),increasing group(group 2),moderate fluctuation group(group 3),and high stable group(group 4).Differences in important baseline variables of distinct overweight and obesity trajectory groups were statistically significant(P<0.001).The CHD incidence density in group 4 was higher than in group 1(P<0.001).Cox risk assessment results showed that hazard ratios(HRs)of increasing group and moderate fluctuation group were not statistically different compared with low stable group(P>0.05),while high stable group had a significantly higher CHD risk(P<0.05),but the effect was not independent of baseline BMI(P>0.05).The effect of overweight and obesity trajectory on CHD did not show a prominent advantage over baseline BMI,mean BMI,and BMI standard deviation during the follow-up.2.Trajectories for high SBP and high DBP were divided into low stable group(group 1),low increasing group(group 2),moderate decreasing group(group 3),moderate increasing-decreasing group(group 4),SBP high decreasing-DBP low decreasing group(group 5)and high decreasing group(group 6).Differences in baseline data of distinct hypertension multivariate trajectory groups were statistically significant(P<0.001).The CHD incidence densities in group 3,4,5,and 6 were higher than those in group 1(P<0.001).Cox risk assessment results showed that HRs of moderate decreasing group,moderate increasing-decreasing group,and high decreasing group were significantly higher than that of low stable group(P<0.05).After adjusting baseline blood pressure,people in the middle-increasing-decreasing group still have an independent CHD risk(P<0.05).Compared with baseline information of blood pressure and standard deviation of SBP and DBP during the whole follow-up,trajectory patterns of high SBP and high DBP can discover the effect of blood pressure on CHD and had higher CHD risk than mean DBP.3.6 groups for multivariate trajectories of LDL-C,HDL-C and TG were identified,group 1 had low increasing LDL-C,moderate high HDL-C and low slowly decreasing TG levels,group 2 had low-moderate increasing LDL-C,high increasing HDL-C and low decreasing TG,group 3 had moderate decreasing LDL-C,low fluctuations HDL-C and high slowly decreasing TG,group 4 had moderate slowly increasing LDL-C,low increasing HDL-C and moderate-high slowly decreasing TG,group 5 had moderate-high steadily increasing LDL-C,moderate-high slowly increasing HDL-C and moderate slowly decreasing TG,group 6 had high decreasing LDL-C,moderate fluctuation HDL-C and moderate-high decreasing TG.Differences in baseline important variables of distinct lipid trajectory groups were statistically significant(P<0.001).The differences between important baseline variables among trajectory groups were statistically significant(P<0.001).The CHD incidence densities of lipid trajectory groups 3,4,5,and 6 were higher compared with that in group 1(P<0.05).Cox regression showed that group 6 had significantly higher CHD risk than that in group 1(P<0.05),but it was not an independent risk factor for CHD(P>0.05)after adjustment for baseline lipid levels.The HRs of lipid trajectories were higher than that of baseline lipids and mean lipids.4.Trajectories of FPG were divided into low increasing group(group 1),low-moderate increasing group(group 2),moderate increasing group(group 3),moderate-high increasing group(group 4),high fluctuation group(group 5)and high increasing group(group 6).Differences in baseline important variables of distinct FPG trajectory groups were statistically significant(P<0.001).The CHD incidence density of FPG trajectory group 5 was higher compared with that in group1(P<0.001).Cox regression showed that high fluctuation group had a higher CHD risk than low increasing group(P<0.05),even after adjustment for baseline FPG,FPG high fluctuation group still have an independent CHD risk(P<0.05).Compared with baseline FPG,mean FPG,FPG standard deviation,the longitudinal trajectory of FPG can better excavate the risk effect of FPG on CHD.5.According to CHD incidence densities,risk scores were assigned to the trajectories of metabolic syndrome-related variables,then the study population was divided into group 1(0 to 3 points),group 2(4 to 7 points),group 3(8 to 11 points),group 4(12 to 15 points)and group 5(16 to 18 points)based on additive risk scores.Cox regression model was used to evaluate the relationship between different risk score groups and CHD risk.The results showed that after adjustment for baseline covariates,group 2,group 3,group 4,and group 5 were all independent risk factors for CHD compared with group 1(P<0.05)Conclusions:1.GBTM has a good application effect in the real-world health management cohort,it can accurately depict the longitudinal dynamic change patterns of MetS related indexes Each variable has trajectory patterns of increasing,stable fluctuations and decreasing,which have reflected the likely situation that indicators gradually deteriorated or declined due to intervention.GBTM can also identify hypertension trajectories of middle-aged and elderly people characterized by isolated systolic hypertension(1SH),and lipids trajectory groups characterized by high TG and low HDL-C,which are in line with clinical practice2.The longitudinal trajectories of MetS related indicators play certain roles in the mining and quantification of CHD risk factors.GBTM could identify trajectory membership with high CHD risk or with deteriorative trends of indicators.3.Compared with traditional indicators,the trajectory patterns of each risk factor and additive risk scores of trajectories play extensive advantages in CHD risk assessment4.The trajectory results show that the intervention of metabolic risk factors has a certain lag.Subsequent studies can further construct discriminant models for different trajectories to early identify trajectory membership with high CHD risk,break long-term effects of multi-category,multi-dimensional risk factors and achieve personalized prevention and treatment of CHD from the perspective of primary prevention.