A Spatial Agent-Based Model of Walking Behavior in Cities

描述(由申请人提供)：人们对社区的社会和建成环境特征如何影响居民的身体活动产生了极大的兴趣。由于步行作为体力活动的一个组成部分的相关性，以及通过在日常生活中增加步行来增加人群体力活动水平的可能性，步行的环境决定因素的研究受到了特别关注。研究记录了步行与环境特征的关系，如土地利用组合和目的地的邻近程度、街道连通性和人行道的存在、环境的美学和设计特征以及安全和暴力水平。然而，关于这项工作的政策影响，重要的问题仍然存在。一个主要挑战是需要考虑到个人之间的多重动态关系(例如，一个人的行为影响他或她周围其他人的行为)，个人与其环境之间的关系(例如，环境随着个人的行为而变化，反之亦然)，以及环境之间的关系(例如，影响安全水平的街道连通性)。考虑这些动态关系是理解环境因素如何运作和确定各种政策可能产生的影响的基础。系统科学方法论(包括基于主体的模型或ABM)作为更好地捕捉人口健康问题中固有的一组复杂的动态关系的方式受到了越来越多的关注，但在具体研究问题上的应用却很少。该提案的总体目标是开发一个基于主体的通用模型(ABM)来描述人们在一个假想城市中的行走行为。这个模型将被用来加强对环境因素影响步行人口水平和步行社会经济不平等的方式的科学理解。具体目标是：(1)建立一个通用的基于主体的模型来模拟人们在虚拟城市中的步行行为；(2)使用所提出的模型来具体研究混合土地使用和安全如何影响人口水平和步行中的社会经济不平等；以及(3)探索使模型反映真实城市的可行性(包括数据可用性和新的数据需求)。这项提案将应用新的系统方法来解决政策阻力的人口健康问题，并将产生这一方法在人口健康中应用的首批范例之一。它将开发一个通用模型，有助于理解跨城市运行的动态，并将为进一步开发针对真实城市的更现实的模型奠定基础。我们提出的系统方法在提高我们对步行的环境决定因素的理解方面具有巨大的潜力，这将是确定相关政策并预测其在未来研究中的影响的关键。这些模型的发展代表了证据用于选择和支持干预措施和政策以预防包括心血管疾病和癌症在内的重大慢性病的方式的范式变化。 与公共健康相关：该项目将开发一个基于空间代理的通用模型，以确定城市人口步行行为的决定因素。该模型可以用来更好地理解环境特征影响步行的动态方式，并确定不同类型干预措施的合理影响。这一基本模型将为开发特定于给定城市的更现实的模型奠定基础。这些建模方法的使用可能会对用于作出政策决策的证据类型以及评估公共卫生干预措施和政策的可能影响的方式产生重大影响。

DESCRIPTION (provided by applicant): There has been an explosion of interest in how features of the social and built environments of communities affect the physical activity of residents. Because of the relevance of walking as a component of physical activity, and the possibility of increasing population-levels of physical activity by increasing walking in daily life, the study of the environmental determinants of walking has received special attention. Studies have documented associations of walking with environmental features such as land use mix and proximity of destinations, street connectivity and presence of sidewalks, aesthetic and design features of the environment, and safety and violence levels. However, important questions remain regarding the policy implications of this work. A major challenge is the need to account for the multiple dynamic relations between individuals (e.g. the behavior of one individual affecting those of others around him or her), between individuals and their environments (e.g. the environment changing in response to the behaviors of individuals and vice versa), and between environments (e.g. street connectivity affecting levels of safety). Accounting for these dynamic relationships is fundamental to understanding how environmental factors operate and to identifying plausible effects of various policies. Systems science methodologies (including agent-based models or ABMs) have received increasing attention as a way to better capture the complex set of dynamic relationships inherent in population health problems, but few applications to specific research problems exist. The overall goal of this proposal is to develop a generic agent-based model (ABM) of people's walking behavior within a hypothetical city. This model will be used to enhance scientific understanding of the ways in which environmental factors affect population-levels of walking and socioeconomic inequalities in walking. The Specific Aims are: (1) To build a generic agent-based model to simulate people's walking behavior within a hypothetical city; (2) To use the proposed model to specifically investigate how mixed land use and safety could contribute to population- levels of walking and to socioeconomic inequalities in walking; and (3) To explore the feasibility (including data availability and new data needs) of making the model reflect a real city. This proposal will apply novel systems methodologies to a policy resistant population health problem and will generate one of the first exemplars of applications of this approach in population health. It will develop a generic model useful for understanding dynamics that operate across cities and will lay the ground work for further development of more realistic models specific to real cities. The systems approach we propose has great potential for improving our understanding of the environmental determinants of walking which will be crucial to identifying relevant policies and anticipating their effects in future research. The development of these models represents a paradigm change in the way evidence is used to select and support interventions and policies to prevent major chronic conditions, including cardiovascular disease and cancer. PUBLIC HEALTH RELEVANCE: This project will develop a generic spatial agent-based model of the determinants of population-levels of walking behavior in a city. The model can be used to better understand the dynamic ways in which environmental features affect walking and identify the plausible impacts of different types of interventions. This basic model will lay the groundwork for the development of more realistic models specific to a given city. The use of these modeling approaches could have major implications for the types of evidence that are used to make policy decisions and for the ways in which the possible impact of public health interventions and policies is assessed.