基于LASSO回归模型的动脉粥样硬化易损斑块风险因子分析

Analysis of Risk Factors in Atherosclerosis Vulnerable Plaque Based on LASSO Regression

目的筛选动脉粥样硬化(AS)易损斑块相关风险因素,并建立AS易损斑块预测模型。方法回顾性分析2019年1月至2021年5月收治的急性冠脉综合征(ACS)患者241例,其中稳定型斑块患者85例,不稳定型斑块患者156例。收集患者入院时的一般临床资料及入院24 h循环血相关指标水平。使用LASSO回归模型进行变量筛选,建立预测模型;绘制模型列线图并绘制受试者操作特征(ROC)曲线,计算曲线下面积(AUC)。结果稳定型斑块组与不稳定型斑块组患者血小板计数、肌酐、血尿素氮、高密度脂蛋白、低密度脂蛋白、脑钠肽、肌红蛋白、C反应蛋白、载脂蛋白A1、载脂蛋白B1水平差异均有统计学意义(P<0.05)。采用LASSO回归模型筛选并构建了两个AS易损斑块风险因子模型:模型1包括年龄、性别、吸烟史、心绞痛史、低密度脂蛋白水平、血小板计数、载脂蛋白A1水平、左心室射血分数;模型2包括年龄、性别、吸烟史、左心室射血分数和载脂蛋白A1水平。其AUC分别为0.799(95%CI:0.725-0.872)和0.776(95%CI:0.698-0.855),两者对斑块易损性的预测能力相当。结论基于LASSO回归模型筛选出的两个AS易损斑块风险模型的预测效率均相对较高,可在AS易损斑块的诊治中发挥重要作用。

Objective To identify the risk factors related to vulnerable plaque in atherosclerosis(AS),and establish risk prediction models for vulnerable plaque in AS.Methods A retrospective observational cohort study was carried out on the patients with acute coronary syndrome(ACS)from January 2019 to May 2021.A total of 241 patients were divided into stable plaque(n=85)and unstable plaque(n=156)groups.Clinical information and laboratory examination of the patients were collected.LASSO regression was used to establish the risk prediction models based on risk factors.Nomograms were formulated,receiver operating characteristic(ROC)curves were used to assess the predictive power and the area under the curve(AUC)was calculated.Results There were significant differences in platelet(PLT),creatinine(Cre),high-density lipoprotein(HDL),low-density lipoprotein(LDL),brain natriuretic peptide(BNP),myoglobin(MYO),apolipoprotein A1(APOA1),apolipoprotein B1(APOB1)and C-reactive protein(CRP)between the two groups(P<0.05).Two AS vulnerable plaque risk prediction models were constructed by LASSO regression with AUC 0.799(95%CI 0.725-0.872)and 0.776(95%CI 0.698-0.855)respectively.Model 1 included age,gender,smoking history,angina pectoris history,LDL,PLT,APOA1 and ejection fraction;Model 2 included age,gender,smoking history,ejection fraction and APOA1.There was no significant difference of AUC between the two models.Conclusion The two AS vulnerable plaque risk factor models may be used in the diagnosis and prevention of AS plaque stability.