Establishment and validation of a risk prediction model for hypertension complicated with cardiovascular disease in middle-aged and elderly patients in Jing'an District,Shanghai

LUO Yuanxin; LI Fang; TAN Jun; PENG Derong

doi:10.3969/j.issn.1002-266X.2025.01.016

Abstract：Objective To establish a risk prediction model for cardiovascular disease(CVD)complication in mid-dle-aged and elderly hypertensive patients in Jing'an District,Shanghai,and to provide a scientific basis for community health services in this area.Methods Totally 1,127 hypertensive patients in Jing'an District,Shanghai were selected and divided into the modelling group of 751 cases and the validation group of 376 cases according to the ratio of 2:1.Data were collected through questionnaires and laboratory tests;questionnaires included general information such as gender,age,height,body mass,waist circumference,smoking,alcohol consumption,diabetes mellitus prevalence,medication use,and family history,while laboratory tests included total cholesterol(TC),triacylglycerol(TG),low-density lipopro-tein(LDL)cholesterol,high-density lipoprotein(HDL)cholesterol,blood uric acid,urinary proteins,and renal func-tion,and other indicators.The patients were followed up for 5 years and the number of cases of CVD or death due to CVD was recorded.Logistic regression was used to analyse the independent risk factors for hypertension complicated with CVD and to draw a nomogram model,and the differentiation,degree of fitting,calibration and clinical utility of the predictive model were assessed by using the area under the curve(AUC)of the receiver operating characteristic(ROC),the Hosmer-Lemeshow goodness-of-fit test,calibration curve and decision curve(DCA).Results No statistically significant differ-ences were found in the general information or laboratory indicators between the modelling and validation groups(all P>0.05).A total of 163(21.65％)of 751 study subjects in the modelling group progressed to CVD.Logistic regression anal-ysis showed that advanced age,male,not taking antihypertensive medication,overweight or obesity,history of smoking,hypertriglyceridemia,hypercholesterolemia,hyperuricaemia,family history of hypertension,positive urinary protein,and moderate to high risk of lipid risk were independent risk factors for CVD among middle-aged and elderly hypertensive pa-tients.The nomogram prediction model of CVD in middle-aged and elderly hypertensive patients was constructed according to the independent influencing factors,and the ROC curve analysis showed that the AUC of CVD in middle-aged and elder-ly hypertensive patients of the modelling group and validation group was 0.763 and 0.806,respectively;the calibration curves showed that the calibration curves in the modelling group and the validation group basically fitted the ideal curves;and the results of Hosmer-Lemeshow Hosmer-Lemeshow goodness-of-fit test showed that the difference between the nomo-gram model and the ideal model in the modelling and validation groups was not statistically significant(P>0.05).The DCA curve results showed that the nomogram prediction model showed high clinical practical value in both modelling and validation groups.Conclusion Advanced age,male,not taking antihypertensive drugs,overweight or obesity,smoking history,hypertriglyceridemia,hypercholesterolemia,hyperuricemia,family history of hypertension,positive urinary pro-teins,and medium-high risk of lipid risk are the independent risk factors for CVD in middle-aged and elderly patients with hypertension,and the predictive efficacy of nomogram prediction model based on the above risk factors is good.