心血管疾病低风险群体颈动脉斑块危险因素及预测模型构建

Risk factors and predictive model of carotid plaque in population at low risk of cardiovascular disease

目的探讨心血管疾病低风险有颈动脉斑块人群的危险因素,构建列线图预测模型。方法分别于2019年11月、2021年11月在广西壮族自治区(广西)柳州市两镇开展现况调查,现场招募481例成年人作为研究对象。使用传统危险因素负担分类评估心血管疾病发生风险;运用颈动脉超声评估研究对象的颈动脉斑块。采用Lasso回归模型进行特征筛选,利用logistic回归模型分析危险因素,最后建立列线图预测模型。采用受试者工作特征(receiver operator characteristic,ROC)曲线下面积(area under curve,AUC)评估预测模型的预测能力。利用Bootstrap自抽样法对模型进行内部验证,利用一致性指数(concordance index,C-index)评估其区分度,利用Hosmer-Lemeshow拟合优度检验评估其校准度。结果研究对象共481人,有302例低风险群体,20.53%(62/302)的低风险群体存在颈动脉斑块。调整因素后的多因素logistic回归分析结果显示,年龄(OR=1.16,95%CI:1.09~1.24)、高血压(OR=3.28,95%CI:1.41~7.64)、既往饮酒(OR=2.98,95%CI:1.13~7.85)、血清细胞间黏附分子-1(serum intercellular adhesion molecule-1,ICAM-1)(OR=1.77,95%CI:1.13~2.78)和天冬氨酸氨基转移酶(aspartate aminotransferase,AST)(OR=9.01,95%CI:1.18~68.80)增加低风险群体发生颈动脉斑块的风险,而体质量指数(body mass index,BMI)(OR=0.86,95%CI:0.75~0.98)降低其风险。构建的列线图预测模型AUC为0.82(95%CI:0.76~0.88),该模型的C-index为0.82(95%CI:0.77~0.88),具有良好的校准度(χ^(2)=11.972,P=0.1525)。结论低风险群体存在较高的颈动脉斑块现患率,其危险因素有别于传统的心血管疾病危险因素,构建的列线图预测模型具有较好的区分度和校准度,对于低风险有颈动脉斑块异质群体的早期识别与个体化心血管疾病精准防治措施制定具有一定作用。

Objective To explore the risk factors of carotid plaque in the population at low risk of cardiovascular disease and to develop a risk-prediction nomogram for carotid plaque in this population.Methods A survey was conducted in November 2019 and November 2021 in two towns of Liuzhou,Guangxi Zhuang Autonomous Region,in which 481 adults were recruited.Traditional risk factor burden classification was used to evaluate the risk of cardiovascular disease,and carotid artery ultrasound was performed to evaluate carotid plaque.Lasso regression model was used for risk factor selection,logistic regression model was fitted to analyze risk factors,and finally a nomogram was plotted based on the developed risk-prediction model for carotid plaque.The area under the receiver operating characteristic(ROC)curve(AUC)was estimated to evaluate the prediction ability of the prediction model.Bootstrap self-sampling method was used for internal validation of the model,the concordance index(C-index)to evaluate its discrimination ability,and Hosmer-Lemeshow goodness-of-fit test to evaluate its calibration.Results Among the 481 participants,302 were at low risk of cardiovascular disease,of whom 20.53%(62/302)had carotid plaques.Multivariable logistic regression after adjusting for relevant factors showed that age(OR=1.16,95%CI:1.09-1.24),hypertension(OR=3.28,95%CI:1.41-7.64),previous alcohol consumption history(OR=2.98,95%CI:1.13-7.85),serum intercellular adhesion molecule-1(ICAM-1)(OR=1.77,95%CI:1.13-2.78)and aspartate aminotransferase(AST)(OR=9.01,95%CI:1.18-68.80)level were associated with increased risk of carotid plaque in people at low risk of cardiovascular disease,while body mass index(BMI)(OR=0.86,95%CI:0.75-0.98)was associated with reduced risk.The AUC of the constructed nomogram was 0.82(95%CI:0.76-0.88),the concordance index was 0.82(95%CI:0.77-0.88),and the calibration was satisfactory(χ^(2)=11.972,P=0.1525).Conclusions Carotid plaque may be prevalent in the population at low risk of cardiovascular disease,and its risk factors are different from traditional cardiovascular risk factors.The nomogram prediction model developed in this study shows a good degree of discrimination and calibration.It may facilitate early identification of population at low risk of cardiovascular but with heterogeneous risk of carotid plaque,and subsequent individualized prevention and treatment measures for cardiovascular disease.