基于两水平Logistic回归分析模型分析我国成人尿石症影响因素  

Analysis on the associated factors of adult urolithiasis in China based on two-level Logistic regression model

作　　者：李宇斯 曾国华[2] 麦赞林[2] 吴文起[2] 陶涛 郭卓源 李舒新 马金香 LI Yu-si;ZENG Guo-hua;MAI Zan-lin;WU Wen-qi;TAO Tao;GUO Zhuo-yuan;LI Shu-xin;MA Jin-xiang(School of Public Health,Guangzhou Medical University,Guangzhou 511439,China;The First Affiliated Hospital of Guangzhou Medical University,Guangzhou 510120,China)

机构地区：[1]广州医科大学公共卫生学院,广州511439 [2]广州医科大学附属第一医院,广州510120

出　　处：《中华疾病控制杂志》2019年第7期866-870,共5页Chinese Journal of Disease Control & Prevention

基　　金：国家自然科学基金(81670643);广州市教育局协同创新重大项目(201620011);广州市科技计划项目(201604020001,201704020193);广东省省级科技计划项目(2014A0 20209085,2017B030314108)~~

摘　　要：目的探讨我国成人尿石症患病的影响因素。方法于2013年5月-2014年7月,采用多阶段分层整群随机抽样,在全国随机抽7个省/直辖市,14个区/县的11个社区和19个自然村进行面对面问卷调查,包括泌尿系超声检查、血常规、尿常规和血液生化检查等。结果有效问卷9 310例中,尿石症患者1 447例,患病率为15. 5%(1 447/9 310);在14个地区之间的患病率有差异(χ~2=711. 523,P <0. 001),最低为山西农村(0. 76%),最高为广东农村(35. 99%)。零模型结果显示各地区尿石症患病有统计学聚集性(t=2. 48,P=0. 027),且组内相关系数ICC=48. 74%。随机效应模型结果显示男性(OR=1. 235,95%CI:1. 082~1. 411,P=0. 005)、年龄增长(OR=1. 101,95%CI:1. 047~1. 158,P=0. 001)、糖尿病史(OR=1. 411,95%CI:1. 192~1. 670,P=0. 001)、结石家族史(OR=1. 867,95%CI:1. 500~2. 323,P <0. 001)、低密度脂蛋白(low density lipoprotein,LDL)(OR=1. 150,95%CI:1. 050~1. 260,P=0. 006)、饮咖啡(OR=1. 352,95%CI:1. 065~1. 716,P=0. 017)、饮碳酸饮料(OR=1. 547,95%CI:1. 203~1. 990,P=0. 002)等为尿石症发生的危险因素,食醋酸(OR=0. 567,95%CI:0. 498~0. 645,P <0. 001)和适量食用豆类(OR=0. 726,95%CI:0. 628~0. 839,P <0. 001)等为尿石症发生的保护因素。结论尿石症患病具有地方聚集性,生活环境和饮食习惯影响尿石症的形成。Objective To investigate the risk factors of adult urolithiasis in China. Methods 14 areas including 11 communities and 19 villages were randomly selected from 7 provinces of China by multi-stage stratified cluster sampling method during the period of May 2013 to July 2014. Individuals were investigated by a face-to-face questionnaire and a physical examination including urinary tract ultrasonographic examinations,routine blood and urine tests and blood biochemical examination ect. Results In total,1 447 participants were found with the urolithiasis among 9 310 individuals and the overall prevalence was 15. 5%( 1 447/9 310). The prevalence of urolithiasis was significantly different among14 areas(χ^2= 711. 523,P < 0. 001),the lowest was the village in Shanxi( 0. 76%) and the highest was the village in Guangdong( 35. 99%). The intercept-only model further indicated the reginal aggregation for the individuals of urolithiasis( t = 2. 48,P = 0. 027) and the ICC was 48. 74%. The two-level Logistic regression model showed that the gender( OR = 1. 235,95% CI: 1. 082-1. 411,P = 0. 005),age( OR = 1. 101,95% CI: 1. 047-1. 158,P = 0. 001),diabetes mellitus( OR = 1. 411,95% CI: 1. 192-1. 670,P = 0. 001),family history of urinary calculi( OR = 1. 867,95% CI: 1. 500-2. 323,P <0. 001),LDL( OR = 1. 150,95% CI: 1. 050-1. 260,P = 0. 006),drinking coffee( OR = 1. 352,95%CI: 1. 065-1. 716,P = 0. 017) and drinking sodas( OR = 1. 547,95% CI: 1. 203-1. 990,P = 0. 002)were the risk factors for urolithiasis. By contrast,consumed more fermented vinegar( OR = 0. 567,95%CI: 0. 498-0. 645,P < 0. 001) and had a amount of legume( OR = 0. 726,95% CI: 0. 628-0. 839,P< 0. 001) were protective factors of urolithiasis. Conclusion The prevalence of urolithiasis among adults reveal an aggregation in area-level,influenced by life environment and dietary habits of individual.

关 键 词：Logistic回归分析模型 尿石症 影响因素 

分 类 号：R181.2[医药卫生—流行病学]