首页|Logistic回归与决策树模型在探讨中心性肥胖影响因素中的应用

Logistic回归与决策树模型在探讨中心性肥胖影响因素中的应用

扫码查看
目的 掌握赤峰市松山区成年居民中心性肥胖的流行情况,采用Logistic回归与决策树模型探讨成年居民中心性肥胖的相关影响因素,为制定防控成入中心性肥胖的策略和措施提供参考依据.方法 2021年采用多阶段随机抽样方法,对赤峰市松山区18岁及以上常住居民的超重肥胖流行情况进行调查,包括问卷调查和体格检查,并采用Logistic回归和决策树模型对成年居民中心性肥胖的影响因素进行深入挖掘与探讨,利用SPSS 20.0软件进行统计分析,检验水准α=0.05.结果 赤峰市松山区成年居民中心性肥胖率为68.55%.不同性别、年龄、文化程度、婚姻状态、职业、睡眠质量、一日三餐是否规律、是否直接参与农牧活动、是否有慢性病家族史的成年居民中心性肥胖率的差异均有统计学意义(P均<0.05).Logistic回归分析结果显示:女性中年/老年、有配偶、一日三餐规律、有慢性病家族史均为患中心性肥胖的危险因素.而文化程度高则是患中心性肥胖的保护因素.中心性肥胖影响因素的决策树模型共3层,筛选出5个风险变量,包括年龄、性别、婚姻状况、文化程度、是否直接参与农牧活动,其中年龄是最主要的影响因素.结论 赤峰市松山区成年居民中心性肥胖率处于较高水平.性别、年龄、文化程度、婚姻状况、是否直接参与农牧活动、是否有慢性病家族史是其主要影响因素.相关部门应给予高度重视并尽快制定相应措施进行干预.
Application of logistic regression and decision tree model in exploring the influencing factors of central obesity
Objective To master the prevalence of central obesity among adult residents in Songshan district of Chifeng City,and by using logistic regression and decision tree models to explore the relevant influencing factors of central obesity among adult residents,to provide reference basis for formulating strategies and measures to prevent and control adult central obesity.Methods In 2021,multistage random sampling method were used to investigate the prevalence of overweight and obesity among permanent residents aged 18 and above in Songshan District of Chifeng City,which including questionnaire survey and physical examination.Logistic regression and decision tree model were used to deeply explore and explore the influencing factors of adult residents'central obesity,and SPSS 20.0 software were used for statistical analysis,the test level α=0.05.Results The central obesity rate of adult residents in Songshan District of Chifeng City was 68.55%.There were statistically significant differences in the central obesity rate among adult residents of different gender,age,education level,marital status,occupation,sleep quality,regularity of three meals a day,direct participation in farming and animal husbandry activities,and family history of chronic diseases(All P<0.05).Logistic regression analysis showed that female,having a spouse,three meals a day,and family history of chronic diseases were all risk factors for central obesity.A high level of education was a protective factor for central obesity.The decision tree model for the influencing factors of central obesity consisted of three layers,and five risk variables were selected,including age,gender,marital status,education level,and direct participation in agricultural and pastoral activities.Among them,age was the most important influencing factor.Conclusions The central obesity rate of adult residents in Songshan district of Chifeng City is at a high level.Gender,age,educational level,marital status,whether to directly participate in farming and animal husbandry activities,and whether to have a family history of chronic diseases are the main influencing factors.Relevant departments should pay high attention to this issue and formulate corresponding measures for intervention as soon as possible.

central obesityinfluencing factorsdecision treeadult residents

龚亮亮、王琪、白钢、边博浩、何永哲、高萌萌、刘培艳、刘铮然

展开 >

包头医学院,内蒙古 包头 014040

包头医学院第一附属医院,内蒙古 包头 014017

中心性肥胖 影响因素 决策树 成年居民

2024

预防医学情报杂志
中华预防医学会,四川省疾病预防控制中心

预防医学情报杂志

CSTPCD
影响因子:0.681
ISSN:1006-4028
年,卷(期):2024.40(2)
  • 19