1、基金项目:河南省医学科技攻关计划省部共建重点项目(SBGJ202002058)作者简介:陈杰(1995 ),女,硕士在读,主要从事老年消化系统疾病基础与临床研究。E-mail:1986901873 qq com通信作者:李晓丽(1977 ),女,博士,主任医师,主要从事老年消化系统疾病基础与临床研究。E-mail:zdyfylixiaoli163 com代谢因素与结直肠息肉发病相关性及风险预测模型建立陈杰,赵倩茹,董玲,李冰,赵向娅,杨轶,杨胜楠,田蕊,李晓丽(郑州大学第一附属医院老年综合二科,河南 郑州 450052)摘要 目的分析代谢相关因素与结直肠息肉(CP)发病之间的相关性,探讨 CP
2、 发病的独立危险因素,并建立风险预测模型。方法选取 2020 年 9 月至 2022 年 7 月在郑州大学第一附属医院老年病科就诊且首次行消化内镜检查的患者675 例,根据内镜下有无 CP 分为 CP 组380 例、非结直肠息肉(NCP)组295 例。单因素分析2 组患者之间一般特征资料、代谢相关疾病、代谢相关临床指标间的差异,通过多因素二元 Logistic 回归确定 CP发病的独立危险因素并构建风险预测模型,受试者工作特征曲线检验该模型预测效能。结果单因素分析结果显示,年龄、性别、体重指数、长期吸烟、长期饮酒、恶性肿瘤家族史、高血压、糖尿病、冠心病、脂肪肝、血脂异常、高尿酸血症、外周动脉粥
3、样硬化、临床实验室指标(谷丙转氨酶、谷氨酰转肽酶、前白蛋白、总胆汁酸、血尿素氮、血肌酐、血尿酸、空腹血糖、糖化血红蛋白、甘油三酯、高密度脂蛋白)在 CP 组与 NCP 组之间比较差异均有统计学意义(2=3 476,P=0 001;2=55 927,P 0 001;2=13 732,P 0 001;2=28 283,P 0 001;2=36 936,P 0 001;2=7 882,P=0 005;2=25 818,P 0 001;2=42 334,P 0 001;2=17 750,P 0 001;2=34 806,P 0 001;2=36 732,P 0 001;2=36 111,P 0 001
4、;2=16 254,P 0 001;Z=2 917,P=0 004;Z=4 508,P 0 001;t=3 771,P 0 001;Z=1 993,P=0 046;Z=2 993,P=0 003;t=4 886,P 0 001;t=4 501,P 0 001;Z=4 376,P 0 001;Z=5 868,P 0 001;Z=4 099,P 0 001;t=2 921,P=0 004)。多因素回归结果显示,年龄60 岁(O=2 684,95%CI:1 200 6 001,P=0 016)、男性(O=1 895,95%CI:1 291 2 782,P=0 001)、长期吸烟(O=1 585,95
5、%CI:1 039 2 418,P=0 033)、长期饮酒(O=1 937,95%CI:1 303 2 880,P=0 001)、恶性肿瘤家族史(O=1 585,95%CI:1 072 2 334,P=0 021)、糖尿病(O=1 963,95%CI:1 243 3 102,P=0 004)、高尿酸血症(O=2 297,95%CI:1 424 3 706,P=0 001),血脂异常(O=1 809,95%CI:1 249 2 620,P=0 002)是结直肠息肉发病的独立危险因素。结直肠息肉发病风险预测模型为:P=1+e2 055+0 987 (60岁)+0 639 (男性)+0 461 (长
6、期吸烟)+0 661 (长期饮酒)+0 461 (恶性肿瘤家族史)+0 675 (糖尿病)+0 832 (高尿酸血症)+0 593 (血脂异常),曲线下面积为 0 773(95%CI:0 737 0 808),敏感性 73 20%,特异性 70 60%。结论老年、男性、长期吸烟、长期饮酒、糖尿病、高尿酸血症、血脂异常是 CP 发病的独立危险因素,构建 Logistic 回归模型可简便有效预测结直肠息肉发病风险。高风险人群,尤其是老年人,应及时行结直肠镜筛查。关键词 结直肠息肉;代谢;风险预测模型DOI:10 3969/j issn 1673-5412 2022 06 004 中图分类号 735
7、 3 文献标识码 A 文章编号 1673 5412(2022)06 0478 07Correlation between metabolic factors and colorectal polyps and establishment of riskprediction modelCHEN Jie,ZHAO Qianru,DONG Ling,LI Bing,ZHAO Xiangya,YANG Yi,YANG Shengnan,TIAN ui,LI Xiaoli(Department of Integrated Geriatrics,the First Affiliated Hospital
8、of Zhengzhou University,Zhengzhou 450052,China)Abstract ObjectiveTo analyze the correlation between metabolic related factors and colorectal polyp(CP),andto explore the independent risk factors of CP and establish a risk prediction model MethodsFrom September 2020 toJuly 2022,All the 675 patients wh
9、o had been hospitalized in the Department of Geriatrics of the First Affiliated Hospital ofZhengzhou University and had undergone digestive endoscopy for the first time were selected They were divided into theCP group of 380 patients and the non-colorectal polyp(NCP)group of 295 patients Univariate
10、analysis was used to an-874JOUNAL OF BASIC AND CLINICAL ONCOLOGY Vol.35 No.6 Dec.2022alyze the differences of general characteristics,metabolism related diseases,and metabolism related clinical indicators Theindependent risk factors of colorectal polyps were determined by multivariate binary logisti
11、c regression,and a risk predictionmodel was constructed The prediction efficiency of the model was tested by the subject work characteristic curve esultsThe results of univariate analysis showed that age,gender,body mass index,long-term smoking,long-term drinking,family history of malignant tumor,hy
12、pertension,diabetes,coronary heart disease,fatty liver,dyslipidemia,hyperurice-mia,peripheral atherosclerosis,and clinical laboratory indicators(alanine aminotransferase,glutamyl transpeptidase,pre-albumin,total bile acid,blood urea nitrogen,blood creatinine,blood uric acid,fasting blood glucose,gly
13、cosylated red eggwhite,triglyceride,high density lipoprotein)were significantly different between the CP group and the NCP group(2=3 476,P=0 001;2=55 927,P 0 001;2=13 732,P 0 001;2=28 283,P 0 001;2=36 936,P 0 001;2=7 882,P=0 005;2=25 818,P 0 001;2=42 334,P 0 001;2=17 750,P 0 001;2=34806,P 0001;2=367
14、32,P 0001;2=36111,P 0001;2=16 254,P 0 001;Z=2 917,P=0 004;Z=4 508,P 0 001;t=3 771,P 0 001;Z=1 993,P=0 046;Z=2 993,P=0 003;t=4 886,P 0 001;t=4 501,P 0 001;Z=4 376,P 0 001;Z=5 868,P 0 001;Z=4 099,P 0 001;t=2 921,P=0 004)Theresults of multivariate binary logistic regression showed that age 60 years old
15、(O=2 684,95%CI:1 200-6 001,P=0 016),male(O=1 895,95%CI:1 291-2 782,P=0 001),long term smoking(O=1 585,95%CI:1 039-2 418,P=0 033),long term drinking(O=1 937,95%CI:1 303-2 880,P=0 001),family history of malignanttumor(O=1 585,95%CI:1 072-2 334,P=0 021),diabetes(O=1 963,95%CI:1 243-3 102,P=0 004),hyper
16、uricemia(O=2 297,95%CI:1 424-3 706,P=0 001)and dyslipidemia(O=1 809,95%CI:1 249-2 620,P=0 002)were independent risk factors for colorectal polyps The risk prediction model of colorectal polyps was estab-lished:P=1+e-2055+0987 (60 years old)+0639 (male)+0461 (long term smoking)+0661 (long term drinking)+0461 (family history of malignant tumor)+0 675 (diabetes)+0 832 (hyperuricemia)+0 593 (dyslipidemia)The area under the curve was 0 773(95%CI:0 711-0 785),the sen-sitivity was 73 20%,the specificit