围绝经期女性发生动脉粥样硬化性心血管疾病的危险因素及其预测模型构建

Risk Factors for Atherosclerotic Cardiovascular Disease in Perimenopausal Women and Construction of Its Prediction Model

目的探讨围绝经期女性发生动脉粥样硬化性心血管疾病(ASCVD)的危险因素,并构建其预测模型。方法回顾性选取2021年7—10月于河北工程大学附属医院参加健康体检的围绝经期女性1000例为研究对象,根据其是否发生ASCVD分为发生组(n=344)和未发生组(n=656)。比较两组临床资料〔年龄、BMI、有无糖尿病、有无高血压、有无高脂血症、饮酒史、吸烟史、SBP、DBP、空腹血糖(FBG)、餐后2 h血糖(2 h PBG)、血钙、促卵泡刺激素(FSH)、雌二醇(E_(2))、骨密度〕。采用ROC曲线确定FSH、E_(2)、骨密度预测围绝经期女性发生ASCVD的最佳截断值,采用多因素Logistic回归分析探讨围绝经期女性发生ASCVD的影响因素,并构建其预测模型。采用ROC曲线分析预测模型预测围绝经期女性发生ASCVD的价值。结果发生组有糖尿病、高血压、高脂血症、吸烟史者所占比例高于未发生组,FSH、E_(2)、骨密度低于未发生组(P<0.05)。ROC曲线分析结果显示,FSH、E_(2)、骨密度预测围绝经期女性发生ASCVD的最佳截断值分别为4.32μg/L、128.42 pmol/L、-1.48。多因素Logistic回归分析结果显示,有糖尿病、高血压、高脂血症、吸烟史及FSH<4.32μg/L、E_(2)<128.42 pmol/L、骨密度<-1.48是围绝经期女性发生ASCVD的危险因素(P<0.05)。构建预测模型:Logit(P)=-35.466+1.294×糖尿病+0.860×高血压+0.936×高脂血症+0.546×吸烟史+1.425×FSH+1.332×E_(2)+1.490×骨密度。ROC曲线分析结果显示,该预测模型预测围绝经期女性发生ASCVD的AUC为0.932〔95%CI(0.892,0.973)〕,最佳截断值为78.090,灵敏度为0.903,特异度为0.860。结论有糖尿病、高血压、高脂血症、吸烟史及FSH<4.32μg/L、E_(2)<128.42 pmol/L、骨密度<-1.48是围绝经期女性发生ASCVD的危险因素,本研究构建的预测模型对围绝经期女性发生ASCVD的预测价值较高。

Objective To investigate the risk factors for atherosclerotic cardiovascular disease(ASCVD)in perimenopausal women,and construct its prediction model.Methods A total of 1000 perimenopausal women who underwent physical examination in Affiliated Hospital of Hebei Engineering University from July to October 2021 were retrospectively selected as the study subjects.They were divided into the occurrence group(n=344)and the non-occurrence group(n=656)according to whether they had ASCVD.The clinical data[age,BMI,diabetes mellitus,hypertension,hyperlipidemia,drinking history,smoking history,SBP,DBP,fasting blood glucose(FBG),2 h postprandial blood glucose(2 h PBG),blood calcium,follicle stimulating hormone(FSH),estradiol(E_(2)),bone mineral density]of the two groups were compared.ROC curve analysis was used to determine the optimal cut-off values of FSH,E_(2) and bone mineral density for predicting ASCVD in perimenopausal women,the influencing factors of ASCVD in perimenopausal women were analyzed by multivariate Logistic regression analysis,and the prediction model was established.The predictive value of prediction model for predicting ASCVD in perimenopausal women was analyzed by ROC curve analysis.Results The proportion of patients with diabetes,hypertension,hyperlipidemia and smoking history in the occurrence group were higher than those in the non-occurrence group,while FSH,E_(2) and bone mineral density were lower than those in the nonoccurrence group(P<0.05).ROC curve analysis results showed that the optimal cut-off values of FSH,E_(2) and bone mineral density for predicting the occurrence of ASCVD in perimenopausal women were 4.32μg/L,128.42 pmol/L and-1.48,respectively.Multivariate Logistic regression analysis showed that diabetes mellitus,hypertension,hyperlipidemia,smoking history,FSH<4.32μg/L,E_(2)<128.42 pmol/L,bone mineral density<-1.48 were risk factors for ASCVD in perimenopausal women(P<0.05).Construction of prediction model:Logit(P)=-35.466+1.294×diabetes+0.860×hypertension+0.936×hyperlipidemia+0.546×smoking+1.425×FSH+1.332×E_(2)+1.490×bone mineral density,ROC curve analysis results showed that,the AUC of prediction model for predicting ASCVD in perimenopausal women was 0.932[95%CI(0.892,0.973)],the optimal cut-off value was 78.090,the sensitivity and specificity were 0.903 and 0.860 respectively.Conclusion Diabetes mellitus,hypertension,hyperlipidemia,smoking history,FSH<4.32μg/L,E_(2)<128.42 pmol/L,bone mineral density<-1.48 are the risk factors for ASCVD in perimenopausal women.The prediction model constructed in this study has high predictive value for ASCVD in perimenopausal women.