| 摘要: |
| 肥胖可能与建成环境有关。本文通过
建立包含建成环境5D变量的密度、多样性、可
达性、设计、换乘点距离等15小类要素在内的环
境量化数据集；使其与医疗机构居民肥胖状况
登记数据叠加，得到的“居民健康——城市建成
环境”综合数据集并建立多维度居民肥胖情况
与建成环境指标判断矩阵，采用机器学习随机
森林算法对数据集进行分类，判断出各个环境
要素对居民肥胖程度影响的相关性和贡献度。
发现北京市石景山区中，可达性特征中街道中间
性和街道弯曲度对居民肥胖影响的贡献度最
大，生活圈公园覆盖数和生活圈水系数量贡献
度最小。通过环境要素的改善对居民肥胖状况
进行干预是具有可行性的。 |
| 关键词: 居民肥胖 生活圈 建成环境 机器学
习 环境要素对肥胖贡献度 |
| DOI：10.13791/j.cnki.hsfwest.20220311 |
| 分类号: |
| 基金项目:国家自然科学基金青年项目（51508378） |
|
| The Impact of Environmental Factors in the Daily Life Circle on Residents’ Health: A CaseStudy of the Obesity Status of Residents in Shijingshan District, Beijing |
|
YUN Yingxia,WANG Shanchao,MA Chao,WANG Jie,REN Lijian
|
| Abstract: |
| Among all current public health problems, obesity becomes one of the major
problems in China. The obesity rate in large urbanized areas has been rising year after year,
and these places tend to have the most obese populations. After decades of rapid urbanization in
China, large urban agglomerations have become the region with the most intensive construction
activities and the most drastic environmental changes in China. Currently, more than 60% of
China’s population lives in large urban communities. In a city, the environment that has the
closest contact with residents’ life is the life circle composed of their daily living activities.
Therefore, changes in the elements of the built environment of the community life circle will
have the most direct impact on the health of residents. Based on international research, the
relationship between the built environment of community life circle and obesity has been
established. The community environment can affect the degree of obesity of residents. The
environment may affect overweight through conductive or inhibitory effects on physical activity
and healthy diet and obesity. Due to differences in many factors such as ethnicity, culture, social
development stage, and economic development level, urban construction in Chinese cities may
differ from others in different countries. This study focuses on China’s national conditions and
discusses the relationship between the built environment and obesity levels in China’s large
cities.
The study selects Shijingshan District, Beijing as the research area. In order to quantify the
relationship between the built environment of the community living circle and obesity degree,
firstly it describes the built environment of the community living circle quantitatively, and then
establishes the “5D” evaluation system. This system contains five categories of environmental
factor indicators such as density, diversity, accessibility, design, and distance to transfer points,
and subdivided into 15 sub-categories based on past researches. In more detail the “5D” is
density, the population density of the street where the living circle is located, diversity (quantified
by the number of POI points in the living circle, the mixed entropy of the living circle POI,
and the density of the POI points in the living circle), accessibility (using the number of park
coverage in the living circle, the living circle Number of water systems, density of walkable
street intersections, number of walkable street intersections, street curvature, street neutrality),
design (quantified by street pedestrian score, residents rate street cycling environment), andtransfer point distance (quantified by Metro coverage to quantify). According to the general travel survey data, starting from the geometric center
of the residential area and the 700 meters network distance as radius, the range of the built environment most frequently used by the respondents
was determined, and then the entire study area was divided into several map sheets. The above methods were used to quantify the built environment
elements of the community living circle, and the data set was superimposed with the registration data of residents’ obesity status provided by medical
institutions in Beijing. A comprehensive judgment data matrix that links the BMI index of residents and the built environment of the living circle
was formulated. In order to bring higher calculation accuracy to the multivariate classification calculation, the study adopts machine learning (random
forest algorithm) to judge the correlation and contribution of various elements in the built environment of community living circle to the degree of
obesity.
In Shijingshan District, Beijing, all of the “5D” environmental factor variables in the built environment of community living circles have an
impact on the degree of obesity of residents living in it. Among them, the residents’ rating of bicycle facilities, subway coverage, the density of
walkable street intersections, and the number of walkable street intersections have a positive correlation with the residents’ BMI index, and other
environmental factors are negatively correlated. Among the accessibility features of “5D” elements, street neutrality and street curvature contribute
the most, while the number of park coverage in the living circle and the number of water systems in the living circle contribute the least.
It is possible to identify the obesity-causing factors that may lead to the obesity of residents in cities through machine learning algorithms
in large quantities. The built environment elements of the living circle in Shijingshan area of Beijing are slightly different from the community
environment elements that can lead to obesity in the current research in European and American countries. A growing body of international research
focuses on reducing obesity rates by changing the physical environment of the community. Based on the research results, it suggests that large cities
in China can ensure the mixing of functions in the living circle by increasing the intermediary and connectivity of streets, expanding the types
and numbers of various facilities in the construction of community living circles, in order to achieve the goal of reducing residents’ obesity and
maintaining residents’ health. |
| Key words: Resident Obesity Living Circle Built Environment Machine Learning Contribution of Environmental Factors to Obesity |
