【作者】 杨梅；

【导师】 陈磊；

【作者基本信息】 四川大学 ， 内科学（呼吸）（专业学位）， 2022， 硕士

【摘要】 目的:慢性阻塞性肺疾病(chronic obstructive pulmonary disease,COPD)是最常见的呼吸系统慢性疾病之一,主要特征为持续性的呼吸道症状和气流受限,其发病机制尚未完全阐明。临床证据表明COPD伴外周血嗜酸性粒细胞增高的患者具有较高急性加重风险和良好的糖皮质激素治疗反应,提示嗜酸性粒细胞可能在COPD发生发展中具有重要作用。探讨COPD伴外周血嗜酸性粒细胞增高的临床预测因素,有助于优化COPD管理策略。近年来,已有部分研究报道了与嗜酸性COPD相关的临床因素,但多为单因素分析结果,缺乏多因素校正的统计学证据支持,且判定高嗜酸性粒细胞的界限值不同,导致结论各异。更重要的是,目前尚缺乏基于中国COPD人群的相关研究。因此,本研究旨在描述中国COPD人群外周血嗜酸性粒细胞的总体水平,并通过构建多重线性回归模型,解析中国COPD人群伴外周血嗜酸性粒细胞增高的临床预测因素。材料和方法:本研究是一项回顾性分析,样本来源于中国成人肺部健康(China Pulmonary Health,CPH)研究。CPH研究是关于我国常见呼吸道疾病流行病学的大型横断面研究。按照既定的纳入和排除标准,我们从CPH研究中筛选并提取出中国四川地区COPD患者的相关数据,包括:人口学特征、个人史、家族史、生活环境和习惯、肺功能和血液检测指标。根据2022年《慢性阻塞性肺疾病全球倡议》(Global Initiative for Chronic Obstructive Lung Disease,GOLD)的相关推荐,以外周血嗜酸性粒细胞绝对计数100 cells/μl和300 cells/μl作为分组依据,进行单因素组间对比(<100 cells/μl、100-300 cells/μl、≥300 cells/μl),同时进行简单线性回归分析。单因素分析中P值≤0.2的变量被进一步纳入多重线性回归模型,以推断高嗜酸性粒细胞的临床预测因素,P值<0.05被认为有统计学差异。结果:本研究共纳入375名COPD患者,外周血嗜酸性粒细胞绝对计数中位数为138.8 cells/μl(四分位间距:82.3-226.6 cells/μl),其中嗜酸粒细胞绝对计数高于100 cells/μl者约66.7%,高于300 cells/μl者约14.7%。单因素组间对比提示男性、低身体质量指数(body mass index,BMI)和家庭收入、饲养宠物、使用生物燃料、高白细胞、低中性粒细胞百分比与外周血嗜酸性粒细胞增高显著相关(P<0.05)。在简单线性回归分析中,男性、低BMI、饲养宠物、使用生物燃料的阳性结果仍有体现(P<0.05),同时还揭示了高密度脂蛋白(high-density lipoprotein,HDL)与外周血嗜酸性粒细胞水平呈负相关(P=0.037)。基于单因素分析结果,共有10个变量被纳入多因素模型,包括年龄、男性、BMI、既往吸烟、家庭收入、饲养宠物、使用生物燃料、血清总胆固醇(total cholesterol,TC)、HDL和低密度脂蛋白(low-density lipoprotein,LDL)。多因素模型调整后的结果(多重线性回归分析)提示,年龄(β=2.819,95%CI:0.639 to 5.000,P=0.012)、男性(β=66.125,95%CI:16.350 to 115.900,P=0.009)与外周血嗜酸性粒细胞水平呈正相关,而HDL与外周血嗜酸性粒细胞呈负相关(β=-64.682,95%CI:-123.451 to-5.914,P=0.031)。上述回归模型解释了外周血嗜酸性粒细胞水平约11.4%的变异(R~2=0.114,校正后R~2=0.067,F=2.431,P=0.010)。结论:1.在中国四川地区的COPD患者群中,外周血嗜酸性粒细胞总体水平不高,多数介于100-300 cells/μl之间。2.通过调整BMI、吸烟情况、家庭收入、饲养宠物、使用生物燃料、TC、LDL后,男性、老年和低HDL与外周血嗜酸性粒细胞增高显著相关,是其可能的临床预测因素。更多还原

【Abstract】 Objective:Chronic obstructive pulmonary disease(COPD)is one of the most common respiratory diseases,characterized by persistent respiratory symptoms and airflow limitation.Currently its mechanisms have been understudied.Clinical evidence has demonstrated that COPD patients with increased blood or sputum eosinophil level have higher risk of acute exacerbation and favorable response to inhaled corticosteroid therapy,implying that eosinophils play a potential role in the development and progression of COPD.Exploring clinical predictors for COPD with elevated eosinophils could further advance the strategy for disease management.Recent studies have revealed a number of possible factors associated with eosinophilic COPD.However,the majority has only conducted univariate analyses,lacking further statistical evidence on multivariable adjustment,and applied different cut-off values for defining high eosinophil level.These contribute to inconsistent conclusions.Moreover,investigation on Chinese patients with COPD has been lacking.Therefore,this study aims to describe typical blood eosinophil level in a Chinese COPD cohort,and explore clinical predictors for high blood eosinophils in COPD using multiple linear regression analysis.Materials and Methods:This is a retrospective analysis,with study samples from the China Pulmonary Health(CPH)study.The CPH study is a large cross-sectional study on the epidemiology of the common respiratory diseases in China.Based on the eligible criteria,we chose and extracted data on COPD patients in Sichuan province,China,including demographics,personal and family history,living condition,spirometry and blood test.According to the recommendation in the 2022 Global Initiative for Chronic Obstructive Lung Disease strategy,all included patients were categorized into three groups by blood eosinophil thresholds of 100 cells/μl and 300 cells/μl.We performed comparison among groups(<100 cells/μl,100-300 cells/μl,≥300cells/μl)and univariate linear regression analysis.Variables with the p value≤0.2 in univariate analyses were further assessed for inclusion in the multiple linear regression model,to explore and conclude clinical predictors for high blood eosinophils.P value<0.05 was defined as statistical significance.Results:This study included 375 patients with COPD,in which the median absolute count of blood eosinophil was 138.8 cells/μl(interquartile range:82.3-226.6cells/μl).66.7%of patients had a blood eosinophil count≥100 cells/μl and 14.7%≥300 cells/μl.Comparison among three groups revealed that male gender,lower body mass index(BMI)and family income,raising pet,biomass use,higher blood leukocyte level and lower blood neutrophil level were associated with increased blood eosinophil count(p<0.05).In univariate linear regression analysis,results concerning male sex,low BMI,raising pet and biomass use were also significant(p<0.05),and high-density lipoprotein(HDL)was further found negatively correlated with blood eosinophil count(p=0.037).On basis of univariate analyses,10variables were included in the multivariable model,comprising age,male sex,BMI,former smoking,family income,raising pet,biomass use,total cholesterol(TC),HDL and low-density lipoprotein(LDL).Multivariable analyses(multiple linear regression analysis)indicated that age(β=2.819,95%CI:0.639 to 5.000,P=0.012)and male gender(β=66.125,95%CI:16.350 to 115.900,P=0.009)were positively associated with blood eosinophil count,whereas HDL(B=-64.682,95%CI:-123.451 to-5.914,p=0.031)was a negative predictor for high blood eosinophils.The regression model above explained approximately 11.4%variances of the blood eosinophil count(R~2=0.114,adjusted R~2=0.067,F=2.431,P=0.010).Conclusion:1.In COPD patients from Sichuan province of China,the peripheral blood eosinophil levels are relatively low,and mostly between 100-300 cells/μl.2.After adjusting for BMI,former smoking,family income,raising pet,biomass use,TC and LDL,elevated blood eosinophil level was correlated with male gender,oldness and low HDL,which were the potential predictors.更多还原