高血压合并阻塞性睡眠呼吸暂停低通气综合征患者发生动脉硬化风险预测模型的建立与验证

Establishment and validation of risk prediction model for atherosclerosis in patients with hypertension complicated with obstructive sleep apnea-hypopnea syndrome

摘要:

目的 探讨高血压合并阻塞性睡眠呼吸暂停低通气综合征（OSAHS）患者发生动脉硬化的危险因素，复制风险预测模型并验证其有效性。方法 选取2019年12月—2021年12月新疆医科大学第一附属医院高血压科收治的1 014例高血压合并OSAHS患者的临床资料。根据行动态血压监测计算动态动脉硬化指数（AASI），严格按照纳排标准筛选，最终纳入380例患者作为研究对象。将患者以7∶3随机分为建模组和验证组，分别有266和114例；根据建模组患者是否存在动脉硬化分为研究组（AASI > 0.55）和对照组（AASI ≤ 0.55）。运用单因素分析和多因素逐步Logistic回归分析筛选发生动脉硬化的独立影响因素，依此构建列线图。分别通过受试者工作特征（ROC）曲线下面积（AUC）、校准曲线、Hosmer-Lemeshow检验、决策曲线分析和临床影响曲线评价模型的预测能力、校准度及临床适用性。结果 建模组与验证组患者性别构成、年龄、高血压病史、饮酒史、家族史、体质量指数（BMI）、尿素氮（BUN）、肌酐（Cr）、估算肾小球滤过率（eGFR）、尿酸（UA）、空腹血糖（FBG）、甘油三酯（TG）、总胆固醇（TC）、高密度脂蛋白胆固醇（HDL-C）、低密度脂蛋白胆固醇（LDL-C）、天冬氨酸氨基转移酶（AST）、丙氨酸氨基转移酶（ALT）、颈动脉内膜中膜厚度（IMT）、24 h平均收缩压（SBP）、24 h平均舒张压（DBP）、白天平均SBP、白天平均DBP、夜间平均SBP、夜间平均DBP、平均血氧饱和度（MSaO₂）、最低血氧饱和度（LSaO₂）、呼吸暂停低通气指数（AHI）、AASI比较，差异均无统计学意义（P >0.05）。研究组与对照组患者性别构成、年龄、饮酒史、家族史、BMI、BUN、Cr、eGFR、UA、TG、TC、HDL-C、LDL-C、AST、ALT、24 h平均DBP、dSBP、dDBP、nSBP、MSaO₂比较，差异均无统计学意义（P >0.05）。两组患者FBG、24 h平均SBP、nDBP、LSaO₂、AHI、IMT、高血压病史比较，差异均有统计学意义（P <0.05）。多因素逐步Logistic回归分析显示：FBG ［O^R=1.375（95% CI：1.058，1.786）］、24 h平均SBP［O^R=1.019（95% CI：1.001，1.038）］、高血压病程［O^R=2.721（95% CI：1.302，5.687）］、IMT［OR=2.197（95% CI：1.104，4.372）］、AHI ［O^R=2.657（95% CI：1.441，4.899）］是高血压合并OSAHS发生动脉硬化的危险因素（P <0.05）。ROC曲线分析结果显示，建模组AUC为0.741（95% CI：0.675，0.807），敏感性68.1%（95% CI：0.573，0.788），特异性71.6%（95% CI：0.653，0.780）；验证组AUC为0.771（95% CI：0.681，0.861），敏感性78.1%（95% CI：0.638，0.924），特异性63.4%（95% CI：0.530，0.738）。校准曲线和Hosmer-Lemeshow检验显示该模型预测结果与实际结果的一致性较好（P >0.05）；决策曲线及临床影响曲线验证列线图具有临床适用性。结论 高血压合并OSAHS患者发生动脉硬化的危险因素包括FBG、24 h平均SBP、高血压病程、IMT、AHI，以此为基础复制风险预测模型，甄别高血压合并OSAHS患者发生动脉硬化的高危人群，为其早期防治提供临床依据。

关键词:阻塞性睡眠呼吸暂停低通气综合征;高血压;动态动脉硬化指数;列线图;风险预测模型

Abstract:

Objective To investigate the risk factors for atherosclerosis in patients with hypertension complicated with obstructive sleep apnea-hypopnea syndrome (OSAHS), and to construct a risk prediction model and to verify its effectiveness.Methods The clinical data of 1014 patients with hypertension complicated with OSAHS admitted to the Department of Hypertension of the First Affiliated Hospital of Xinjiang Medical University from December 2019 to December 2021 were selected. The ambulatory arterial stiffness index (AASI) was calculated by ambulatory blood pressure monitoring. In strict accordance with the inclusion and exclusion criteria, 380 patients were finally included. The patients were randomly divided into the modeling group (n = 266) and the validation group (n = 114), and those in the modeling group were further divided into the study group (AASI > 0.55) and the control group (AASI ≤ 0.55) according to whether the patients had atherosclerosis. The independent factors affecting the occurrence of arteriosclerosis were screened by univariable analysis and multivariable Logistic regression analysis, based on which the nomogram was constructed. The predictive performance, calibration, and clinical applicability of the model were evaluated using the area under the receiver operating characteristic (ROC) curve (AUC), calibration curve, Hosmer-Lemeshow test, decision curve analysis and clinical impact curve, respectively.Results There was no difference in gender composition, age, history of hypertension, history of alcohol consumption, family history, body mass index (BMI), levels of blood urea nitrogen (BUN) and creatinine (Cr), estimated glomerular filtration rate (eGFR), levels of uric acid (UA), fasting blood glucose (FBG), triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), aspartate aminotransferase (AST) and alanine aminotransferase (ALT), carotid intima-media thickness (IMT), 24-hour average systolic blood pressure (SBP), 24-hour average diastolic blood pressure (DBP), daytime average systolic blood pressure, daytime average diastolic blood pressure, nighttime average systolic blood pressure, nighttime average diastolic blood pressure, mean blood oxygen saturation (MSaO₂), lowest blood oxygen saturation (LSaO₂), apnea-hypopnea index (AHI), and AASI between the modeling group and the validation group (P > 0.05). The gender composition, age, history of alcohol consumption, family history, BMI, levels of BUN and Cr, eGFR, levels of UA, TG, TC, HDL-C, LDL-C, AST and ALT, 24-hour average DBP, dSBP, dDBP, nSBP and MSaO₂ were not different between the study group and the control group (P > 0.05). There were differences in the level of FBG, 24-hour average SBP, nDBP, LSaO₂, AHI, IMT, and duration of hypertension between the two groups of patients (P < 0.05). The multivariable Logistic regression analysis denoted that the level of FBG [O^R =1.375 (95% CI: 1.058, 1.786) ], 24-hour average SBP [O^R = 1.019 (95% CI: 1.001, 1.038)], duration of hypertension [O^R = 2.721 (95% CI: 1.302, 5.687) ], IMT [O^R = 2.197 (95% CI: 1.104, 4.372) ] and AHI [O^R = 2.657 (95% CI: 1.441, 4.899) ] were independent factors affecting atherosclerosis in patients with hypertension complicated with OSAHS (P < 0.05). The ROC curve analysis indicated that the AUC of the modeling group was 0.741 (95% CI: 0.675, 0.807), with a sensitivity of 68.1% (95% CI: 0.573, 0.788) and a specificity of 71.6% (95% CI: 0.653, 0.780), and that the AUC of the validation group was 0.771 (95% CI: 0.681, 0.861), with a sensitivity of 78.1% (95% CI: 0.638, 0.924) and a specificity of 63.4% (95% CI: 0.530, 0.738). The calibration curve and the Hosmer-Lemeshow test demonstrated that the predicted results of the model were in good agreement with the actual results. Decision curve analysis and clinical impact curve verified the clinical applicability of the nomogram.Conclusions The factors affecting atherosclerosis in patients with hypertension complicated with OSAHS include the level of FBG, 24-hour average SBP, duration of hypertension, IMT and AHI. The risk prediction model based on these factors could identify high-risk groups for developing atherosclerosis in patients with hypertension complicated with OSAHS, and provide clinical guidance for early prevention and treatment.

Key words:obstructive sleep apnea-hypopnea syndrome;hypertension;ambulatory arterial stiffness index;nomogram;risk prediction model