2型糖尿病合并阻塞性睡眠呼吸暂停低通气综合征的列线图预测模型构建

Construction of a nomogram prediction model for type 2 diabetes mellitus combined with obstructive sleep apnea-hypopnea syndrome

目的 构建2型糖尿病(T2DM)合并阻塞性睡眠呼吸暂停低通气综合征(OSAHS)的列线图预测模型。方法 选取2018年9月至2020年4月自贡市第四人民医院完成多导睡眠监测的526例T2DM患者为研究对象,并按照7∶3的比例随机分为训练组和验证组。通过对T2DM患者进行单变量和多变量的logistic回归分析,本研究确定了OSAHS的风险因素,并据此构建了一个预测模型。模型的有效性通过计算受试者工作特征曲线下的面积(AUC)、使用校正曲线和决策曲线(DCA)进行评估。结果 多因素logistic分析结果显示,体重指数(OR=1.09,95%CI:1.01~1.17)、高血压病史(OR=3.80,95%CI:3.43~4.26)、血清低密度脂蛋白(OR=1.35,95%CI:1.03~1.56)及25(OH)维生素D(OR=0.94,95%CI:0.91~0.96)水平是T2DM合并OSAHS发生的独立预测因素(P <0.05)。根据上述变量建立列线图预测模型,并在训练组和验证组中预测T2DM合并OSAHS发生的AUC分别为0.742和0.789。Hosmer-Lemeshow拟合优度检验显示模型具有较好的拟合度(P> 0.05)。DCA显示预测模型能够获得净收益的风险阈值大于0.9。结论 本研究成功建立并验证了一种性能良好的列线图预测模型,有助于提高T2DM患者合并OSAHS的早期识别和筛选能力。

Objective To construct a nomogram prediction model for type 2 diabetes mellitus(T2DM) combined with obstructive sleep apnea-hypopnea syndrome(OSAHS).Methods A total of 526 T2DM patients who completed polysomnography in Zigong Fourth People's Hospital from September 2018 to April 2020 were selected as the study subjects and randomly divided into a training group and a validation group in a 7 ∶ 3 ratio.The risk factors of T2DM patients with OSAHS were analyzed through univariate and multivariate logistic regression,and a nomogram prediction model was established.The effectiveness of the model was evaluated using the area under the curve(AUC),calibration curve and decision curve analysis(DCA).Results The multivariate logistic analysis showed that body mass index(OR=1.09,95%CI:1.01-1.17),history of hypertension(OR=3.80,95%CI:3.43-4.26),serum low-density lipoprotein(OR=1.35,95%CI:1.03-1.56),and 25(OH) vitamin D(OR=0.94,95%CI:0.91-0.96) levels were independent predictors of T2DM combined with OSAHS(P < 0.05).The nomogram prediction model was established based on the above variables and the AUC of T2DM combined with OSAHS in the training and validation groups was predicted,which was 0.742 and 0.789,respectively.The Hosmer-Lemeshow goodness of fit test showed that the model had a good fit(P > 0.05).DCA showed that the risk threshold for the prediction model to achieve net benefits was greater than 0.9.Conclusion This study successfully establishes and validates a high-performance nomogram prediction model,helping improve the early recognition and screening ability of T2DM patients combined with OSAHS.