Metabolic Syndrome Exploratory Factor Analysis In Xinjiang Kazakhs And Cohort Study On Construction Of A Risk Prediction Model For Cardiovascular Disease Risk

Objective: 1.The study was a retrospective and prospective two-way cohort study to explore the major risk factors affecting the high incidence of MS in Xinjiang Kazakh.Draw the MS factor pattern diagram to preliminarily grasp the aggregation characteristics of the Kazakh MS factor in Xinjiang;2.To verify the impact of the main factors on the incidence of CVD in the baseline population factor analysis,sort the main factors,and provide reference for the development of effective interventions in the future;3.To construct a simple and convenient Kazakhs CVD prediction model with MS main factor as predictor,and to provide reference for the future development of CVD and other chronic disease prevention strategies.Methods: 1.The study population included Kazakh residents who received routine health checks in Nalati County,Yili,Xinjiang from 2010 to 2017.Three follow-up studies were conducted in 2013,2016 and 2017,respectively,through local social security information and local hospital medical records,and through the Autonomous Region Center for Disease Control to collect CVD death data for the 2010-2017 study population;2.Factor analysis was performed in 706 MS subjects to extract CVD-related factors with specific clinical significance from 18 biomarkers tested in routine health check-up.Logistic regression model,based on an extreme case-control design with 191 CVD patients and 384 controls,was performed to evaluate the extracted factors used to identify CVD.3.Then,Cox model,based on a cohort design with 2007 subjects who completed the 5-year follow-up were used as internal populations,was conducted to validate their predictive effects.Another Kazakh research site,Kalabulak village,was the source of 219 external verification controls,was performed to evaluate the extracted factors used to identify CVD.4.The feasibility of the CVD prediction model constructed by MS main factors and original variables was compared,and the area under the ROC curve was compared with internal population data for evaluation,and to construct a simple and convenient model risk of CVD risk.Results: 1.The incidence of CVD in the Kazakh population was 15.10%(95%CI: 13.48%-16.61%),15.89% for males and 14.74% for females.The incidence of CVD in external population was 13.70%(95%CI: 9.11%-18.29%),12.40% for males and 14.62% for females.There were no significant differences in age,gender,SBP,DBP,smoking,alcohol consumption,prevalence of MS and incidence of CVD among the internal and external populations.According to JIS and IDF standards,the crude prevalence rates of Xinjiang Kazakhs MS were 30.88%(male: 29.41%,female: 31.86%)and 25.63%(male: 21.03%,female: 28.66%).According to the IDF criteria,the prevalence of MS was statistically different between men and women,and the prevalence of MS in women was higher than that in men(?2=16.713,P<0.001).The cardiovascular risk factors in the MS Metabolism Network Center and the low HDL-C and hypertension components in the MS diagnostic component were statistically different in the male and female MS populations.2.Factor analysis was used for Xinjiang Kazakh male and female Met S patients,and the main factors related to CVD were extracted from 18 routine physical examination indicators.Explore the structural patterns of MS components in Kazakh male and female populations.The results show that MS can be explained by seven independent metabolic component main factors.For the male group,the first factor was named obesity factor and contributed by weight & waistline & BAI,liver function factor by TBIL & IBIL,Lipid factor by HDL-C& APOA & ALB,renal metabolic factor by UA& CREA & BUN,liverenzyme factor by ALT& AST & ?-HBDH,blood pressure factor by SBP & DBP,blood glucose factor by FPG & FMN.For the female group,the first was named liver function factor & contributed by TBIL& IBIL & ALB,obesity factor by weight& waistline & BAI,lipid factor by HDL-C & APOA,blood pressure factor by SBP & DBP,liver enzyme factor by ALT& AST & ?-HBDH,renal metabolic factor by UA& CREA & BUN,blood glucose factor by FPG & FMN.3.Logistic discriminant model based on case-control study,the area under the ROC curve(AUC)for discriminating CVD in the Kazakh male population was 0.857(95% CI: 0.807-0.898),the cut-off value was 0.30,the sensitivity was 80.49%,and the specificity was 81.71%.The AUC for discriminating CVD in the Kazakh female population was 0.852(95% CI: 0.809-0.889),the cut-off value was 0.23,the sensitivity was 88.07%,and the specificity was 66.36%.4.Model validation was performed on 2007 internal populations based on the Cox predictive model of the cohort study.The AUC for predicting 5-year CVD in the male population was 0.752(95% CI: 0.720-0.781),the predicted cut-off value was 0.50,the sensitivity was 62.10%,and the specificity was 78.22%.The AUC for predicting 5-year CVD in the female population was 0.807(95% CI: 0.783-0.829),the predicted cut-off value was 0.53,the sensitivity was 83.71%,and the specificity was 65.60%.5.Model validation was performed on 219 external populations based on the Cox predictive model of the cohort study.The AUC for predicting 5-year CVD in the male population was 0.752(95% CI: 0.720-0.781),the predicted cut-off value was 0.50,the sensitivity was 62.10%,and the specificity was 78.22%.The AUC for predicting 5-year CVD in the female population was 0.866(95% CI: 0.795-0.919),the predicted cut-off value was 0.52,the sensitivity was 84.21%,and the specificity was 76.58%.6.The Cox prediction model was constructed by MS main factor and original variable,and the internal population data was processed by stepwise regression method.For male population,the CVD model constructed main factor AUC of 0.748(95% CI: 0.717-0.778)higher than the original model variables AUC was constructed CVD 0.692(95% CI: 0.658-0.723),and statistically significant(Z=3.993,P<0.05).The CVD model AUC of the female population main factor was 0.806(95% CI: 0.783-0.829),and the AUC of the CVD model constructed higher than the original variable was 0.780(95% CI: 0.756-0.803),and there were statistical differences(Z = 2.553,P = 0.010).Conclusions : 1.The incidence of CVD and the prevalence of MS in Xinjiang Kazaks are higher than the national level.2.The metabolic components of MS are abnormally aggregated,which is the result of a combination of multiple causes.Through factor analysis,it was found that seven identical main factors were extracted from Kazakh MS and MS,namely obesity factor,liver function factor,lipid factor,kidney metabolism factor,liver enzyme factor,blood pressure factor and blood glucose factor.Although the main components of MS are different in the composition and order of the male and female MS population,they are mainly based on abnormal aggregation of obesity factors,liver function factors and lipid factors,and liver function factors may be the main factor unique to the Kazakhs MS population.3.In this study,a 5-year Kazakh CVD onset prediction model was constructed using seven MS main factors.Through the verification of internal and external populations,the fitted CVD incidence prediction model can well discriminate and predict the occurrence of Xinjiang Kazakh CVD.4.The nomogram of the risk of developing CVD for 5 years can more directly evaluate the risk of Kazakhs CVD,and provide clues for individuals to prevent and treat CVD.