潘怡,辛鹏,郑文龙,李威,王德征.天津市中年人群睡眠状况与代谢综合征关系的决策树分析[J].Chinese journal of Epidemiology,2020,41(8):1266-1271 |
天津市中年人群睡眠状况与代谢综合征关系的决策树分析 |
Relationship between both duration and quality of sleep, and metabolic syndrome identified by classification tree model in midlife population in Tianjin |
Received:December 31, 2019 |
DOI:10.3760/cma.j.cn112338-20191231-00924 |
KeyWord: 睡眠 代谢综合征 影响因素 决策树 |
English Key Word: Sleep Metabolic syndrome Influence factors Classification tree |
FundProject: |
Author Name | Affiliation | E-mail | Pan Yi | Tianjin Centers for Disease Control and Prevention, Tianjin 300011, China | | Xin Peng | Tianjin Centers for Disease Control and Prevention, Tianjin 300011, China | | Zheng Wenlong | Tianjin Centers for Disease Control and Prevention, Tianjin 300011, China | | Li Wei | Tianjin Centers for Disease Control and Prevention, Tianjin 300011, China | | Wang Dezheng | Tianjin Centers for Disease Control and Prevention, Tianjin 300011, China | wangdezheng1001@163.com |
|
Hits: 3078 |
Download times: 1157 |
Abstract: |
目的 探讨睡眠质量、睡眠时间与MS的关系。方法 本研究以2015年中国成人慢性病与营养监测天津地区监测人群为研究对象,采用多阶段整群随机抽样方法抽取天津市7个区42个监测点45~59岁中年人群1 388人为研究对象,对其进行睡眠质量、睡眠时间与MS关系的描述性流行病学分析,采用χ2检验、单因素logistic回归和决策树模型进行分析。结果 研究对象中睡眠质量良好、一般、较差3组MS患病率分别为30.4%、37.4%和43.1%;睡眠时间充足(6.0~7.5 h)、睡眠时间不足和睡眠时间较长(<6.0 h、>7.5 h)的人群MS患病率分别为30.6%和36.5%。调整了年龄和性别的单因素logistic回归显示睡眠质量差、睡眠时间不足和睡眠时间较长都是MS的危险因素,差异均有统计学意义(P<0.05)。决策树模型筛选出性别(重要性为0.004,标准化重要性为100%)、睡眠质量(重要性为0.004,标准化重要性为99.5%)、睡眠时间(重要性为0.002,标准化重要性为38.6%)、文化程度(重要性为0.001,标准化重要性为22.3%)、盐摄入量(重要性为0.001,标准化重要性为22.2%)为MS的重要影响因素,并存在交互作用。结论 睡眠质量和睡眠时间是天津市中年人群MS的重要影响因素,应关注中年人群的睡眠及健康状况,并加强对中年人的健康教育,提高其自我健康管理能力,预防控制MS及其他慢性疾病。 |
English Abstract: |
Objective To study the association of duration and quality of sleep with metabolic syndrome. Methods Multi-stage cluster random sampling method was used to extract data from the ‘2015 adult chronic disease and nutrition monitoring population project’ from 42 communities of 7 districts in Tianjin. A total of 1 388 residents aged 45 to 59 years old were selected to analyze the relationship between both duration and quality of sleep and the metabolic syndrome. χ2 test,non-conditional logistic regression and classification tree models were used for data analysis. Result The prevalence of metabolic syndrome appeared as 30.4%, 37.4% and 43.1% in the poor, common or well sleep groups respectively. The prevalence rates of metabolic syndrome in the sleep time <6.0 h and >7.5 h, 6.0-7.5 h groups were 36.5% and 30.6% respectively. After adjusting for gender and age, results from the logistic regression analysis showed that short or long sleep duration, as well as poor sleep quality all appeared as risk factors on metabolic syndrome (P<0.05). According to the classification tree model, factors as gender (importance:0.004, standardized importance:100%), quality of sleep (importance:0.004, standardized importance:99.5%), duration of sleep (importance:0.002, standardized importance:38.6%), education level (importance:0.001, standardized importance:22.3%) and salt intake (importance:0.001, standardized importance:22.2%) were all important on metabolic syndrome and with interactive effects. Conclusions Both quality and duration of sleep were important influencing factors on metabolic syndrome among midlife population in Tianjin. More attention should be paid to sleep and health status among the midlife population. Ability on self-management of health should also be strengthened through health education, to prevent metabolic syndrome and other chronic diseases in this population. |
View Fulltext
Html FullText
View/Add Comment Download reader |
Close |
|
|
|