Study On Prediction Model Of Carotid Atherosclerosis In Overweight Or Obese People

Objective: Through establishing the risk nomogram model of carotid atherosclerosis in overweight or obese people to realize early recognition and intervention,thus delaying the occurrence of cardiovascular events.Methods:(1)We selected people who underwent physical examination in the Physical Examination Department of the Affiliated Hospital of Guilin Medical College from July to October 2017 and July to October 2019 as our research subjects,collected the results of their name,gender,Age,disease history,height,weight,biochemical indicators and Color Doppler Ultrasound of carotid artery,and then calculated their body mass index(BMI),triglyceride to high-density lipoprotein cholesterol ratio(TG/HDL＿c),triglyceride glucose index(TyG)and triglyceride glucose-body mass index(TyG-BMI).After that,we took the overweight or obese people in 2017 as the modeling group and the overweight or obese people in 2019 as the validation group,divided them into case group and control group based on whether they had carotid atherosclerosis,and then compared and analyzed the data characteristics of the modeling group and validation group.(2)We performed a logistic regression analysis on the research variables of the modeling group to determine risk factors,and costructed a nomograph model,with its predictive evaluation performed by the receiver operating characteristic(ROC)curve,calibration curve,decision curve analysis(DCA)and clinical impact curve(CIC).At the same time,we gave a validation to the model.(3)We used a single-noninvasive index and biochemical index to constrcut a nomograph model respectively,and then evaluated and compared the models.Results:(1)There are 1850 cases in the modeling group and 1522 cases in the validation group.The results of data comparison between the two groups showed an increasing level of the age,uric acid(UA),creatinine(Cr),fasting plasma glucose(FPG),total cholesterol(TC),low-density lipoprotein cholesterol(LDL＿c),TyG,TyG-BMI and the constituent ratio of hypertension in the case group(P<0.05),and there was no significant difference in gender composition ratio,alanine aminotransferase(ALT),aspartate aminotransferase(AST),triglyceride(TG),high-density lipoprotein cholesterol(HDL＿c),TG/HDL＿c between groups(P>0.05).(2)We conducted a logistic regression analysis on the research variables of the modeling group,and the results of multivariate analysis showed that there was a statistical significance in their age,hypertension,UA,FPG,LDL＿c and TyG(P<0.05),with which we built a nomograph model(recorded as the conventional nomograph model).(3)The ROC area under the curve(AUC)of the conventional nomogram model was0.775,with the Youden index being 0.4290,the sensitivity being 72.66%,and the specificity being 70.24%;the calibration curve of the model fitted well with the ideal curve;DCA showed a threshold probability(Pt)between 0.09 and 0.88,with a clinical net benefit greater than 0;CIC displayed a red curve(Number high risk curve)containing a blue curve(Number high risk with event curve).(4)The AUC of the validation group was 0.797,with the Youden index being0.4659,the sensitivity being 78.16%,and the specificity being 68.43%;the calibration curve of the model basically coincided with the ideal curve;DCA showed that Pt ranged from 0.07 to 0.80,with a clinical net benefit greater than0;CIC displayed the Number high risk curve containing the Number high risk with event curve.(5)In the modeling group,we constructed nomogram model with age and hypertension(recorded as Noninvasive nomogram model),also with UA,FPG,LDL＿c and TyG(recorded as Biochemical nomogram model).Compared with the Conventional nomogram model,the Noninvasive nomogram model also has a moderate predictive ability,but with a statistically significant difference in the AUC of the two models(P<0.05).Except that,the clinically practical value of the Conventional nomogram model is also superior to the Noninvasive nomogram model;the predictive ability and clinically practical value of the Conventional and Noninvasive nomogram models are superior to those of the Biochemical nomogram model.Conclusion: This study is based on age,hypertension,UA,FPG,LDL＿c.The risk nomogram model of carotid atherosclerosis in overweight or obese people established by TyG has good predictive ability and clinical practicability Moreover,the nomogram model constructed based on age and hypertension also has certain predictive power,and may also have certain guiding significance for overweight or obese populations in areas with limited medical conditions.