摘要
目的构建代谢型肥胖正常体重(MONW)发病风险预测模型,为该人群的早期防控提供科学依据。方法从2018年1月至2019年1月在武警某部医院体检的人群中,筛选符合MONW诊断标准的患者356例作为病例组,在体检结果正常者中随机选取356例作为对照组。从2019年1—12月的体检人群中筛选高危人群343例作为验证队列,进行外部验证。收集调查对象的一般情况、生物标志物水平以及所选定的易感基因的单核苷酸多态性位点信息。采用SPSS25.0软件进行t检验、χ^(2)检验和单因素logistic回归分析,采用多因素logistic回归模型构建预测模型,绘制受试者工作特征(ROC)曲线对模型进行区分度评价。结果多因素logistic回归模型中有统计学意义的指标按其作用强度从强到弱分别为空腹血糖(OR=4.264,95%CI:1.598~11.378)、FTO基因的单核苷酸多态性位点rs1451085(OR=1.615,95%CI:1.344~1.942)和位点rs9939609(OR=1.111,95%CI:1.048~1.178)、尿素(OR=1.129,95%CI:1.010~1.261),均有统计学意义(P<0.05,P<0.01)。验证队列中经模型预测阳性者290例,按照MONW诊断标准判断为患者的有318例。模型的灵敏度为88.67%,特异度为68.24%,阳性似然比为2.79,阴性似然比为0.16,阳性预测值为97.25%,阴性预测值为32.08%,预测的整体准确率为87.17%,ROC曲线的曲线下面积为0.811。结论该研究构建的预测模型有较好的辅助预判功能,有一定的临床应用价值。
Objective To construct a risk prediction model for people of metabolic obesity with normal weight(MONW),and provide the scientific basis for the early prevention and control of those population.Methods From January 2018 to January2019,356 patients who was from the population undergoing physical examination in a hospital of the Armed Police Force and met the diagnostic criteria of MONW were selected as the case group,and 356 participants with normal physical examination results were randomly selected as the control group.From January to December 2019,343 cases of high-risk groups were selected as a verification cohort for external verification.The general information,the level of biomarkers,and the single nucleotide polymorphism locus of selected susceptible genes in subjects were collected.The t test,χ^(2) test and single-factor logistic regression method were used to analyze the data;multi-factors logistic regression model was used to construct the prediction model;and receiver operating characteristic(ROC)curves were drawn to evaluate the discrimination of the model.Results The statistically significant indicators in the multivariate logistic regression model were fasting blood glucose(OR=4.264,95%CI:1.598-11.378)and the single nucleotide polymorphism locus of FTO gene rs1451085(OR=1.615,95%CI:1.344-1.942)and rs9939609(OR=1.111,95%CI:1.048-1.178),urea(OR=1.129,95%CI:1.010-1.261),P<0.05 or P<0.01.In verification cohort,290 cases were predicted positive by prediction model,318 cases were determined according to MONW diagnostic criteria.The sensitivity of the model was 88.67%,while the specificity was 68.24%,the positive likelihood ratio was 2.79,the negative likelihood ratio was 0.16,the positive predictive value was 97.25%,the negative predictive value was 32.08%,and the overall prediction accuracy rate was87.17%.The area under the ROC curve was 0.811.Conclusion The predictive model constructed in this study has a good auxiliary predictive function and has certain clinical application value.
作者
王瑾瑾
魏港
闫国立
WANG Jin-jin;WEI Gang;YAN Guo-li(Medical School,Henan University of Chinese Traditional Medicine,Zhengzhou,Henan Province 450046,China;不详)
出处
《中国慢性病预防与控制》
CAS
CSCD
北大核心
2021年第12期887-891,共5页
Chinese Journal of Prevention and Control of Chronic Diseases
基金
河南省高等学校重点科研项目(18A330003)
河南省高等学校青年骨干教师培养计划(2020GGJS109)
河南省自然科学基金项目(202300410253)。
关键词
代谢型肥胖体重正常
生物标志物
易感基因
预测模型
Metabolically obese with normal weight
Biomarkers
Susceptibility gene
Predictive model
作者简介
王瑾瑾,博士,副教授,从事慢性非传染性疾病病因学研究,E-mail:wangjinjin@hactcm.edu.cn;通信作者:闫国立,E-mail:yanguoli@hactcm.edu.cn。
