Linking dynamic environments with childhood obesity

描述（由申请人提供）：提议的是一个培训和研究项目，利用正在进行的纵向研究来调查动态环境对儿童肥胖的影响。肥胖和久坐行为对儿童健康的威胁越来越普遍，这可能反映了儿童及其家庭所经历的身体活动机会、食物环境和社会背景的变化。建筑环境的特征与儿童和成人的健康行为和体重有关，但这种联系的因果性质尚不清楚。一个关键的挑战是区分对观察到的关联的竞争解释：社区对个人身体活动和健康的影响与积极健康的个人选择支持他们生活方式的社区的趋势。拟议的研究有可能证实或缓和先前关于该主题的研究结论。鉴于以前的横断面研究在很大程度上假设居住环境是静态的，拟议的项目将使用关于居住流动性和动态环境的数据来更好地建立所谓的环境原因和健康影响的时间序列。可用的社会背景措施将被用来显示建筑环境协会如何与社会背景相比较或相互作用。我将选择分析方法来充分利用这些数据，并评估相互竞争的因果假设。我将从两个纵向研究中分析建筑环境、社会背景、身体活动和体重方面的数据。第一个是“脆弱家庭儿童幸福研究”，这是一项对4898名美国大城市儿童的出生队列研究，对单亲家庭进行了抽样调查，目前正在收集9岁儿童的数据。第二项是健康场所试验，对新建的智能增长开发项目进行自然实验评估，预计建成的环境特征将对当地的社交网络和亲子对的身体活动产生有益影响。这些队列的分析将解决特定的假设，重点关注建筑环境、社会背景及其相互作用的致肥效应。除了使用纵向和自然实验设计外，我还将探索一系列分析策略，以测试观察到的关联的因果性质。特别是，我将通过分层分析、结构方程建模、潜在轨迹分析、有向无环图、倾向评分匹配和工具变量分析来检查测序、特异性和选择。拟议的研究将辅以以下方面的经验和教学培训：(1)减少健康差距的环境方法，(2)社会背景与健康的理论和测量，(3)轨迹分析和生命过程流行病学，以及(4)研究设计和分析方法，以提供因果推理。这项研究和培训将使我能够启动一个独立研究项目，研究城市建筑和社会环境对儿童体育活动和健康的影响。

DESCRIPTION (provided by applicant): Proposed is a program of training and research using ongoing longitudinal studies to investigate the effects of dynamic environments on childhood obesity. Obesity and sedentary behavior are increasingly common threats to childhood health, likely reflecting changes to the physical activity opportunities, food environment, and social contexts experienced by children and their families. Characteristics of the built environment have been linked to health behaviors and weight of both children and adults, but the causal nature of the link is not clear. A key challenge is to distinguish between competing explanations for the observed associations: the effect of neighborhoods on individual physical activity and health versus the tendency for active and healthy individuals to select neighborhoods that support their lifestyle. The proposed research has the potential to corroborate or temper the conclusions of previous research on this topic. Whereas previous cross-sectional studies have largely assumed that residential environments are static, the proposed project will use data on residential mobility and dynamic environments to better establish the temporal sequence of purported environmental causes and health effects. Available social context measures will be leveraged to show how built environment associations compare to or interact with those of social context. I will select analytical approaches to take full advantage of these data and to assess competing causal hypotheses. I will analyze data on built environments, social context, physical activity, and weight from two longitudinal studies. The first is the Fragile Families Study of Childhood Wellbeing, a birth cohort study of 4,898 children in large US cities that oversampled single parent families and is currently collecting data at age 9. The second is the Healthy PLACES trial a natural experimental evaluation of a newly built Smart Growth development, with built environment characteristics expected to have beneficial effects on local social networks and physical activity from parent-child pairs. Analyses in these cohorts will address specific hypotheses focusing on the obesogenic effects of the built environment, social context, and their interaction. In addition to the use of longitudinal and natural experimental designs, I will explore a range of analytical strategies selected to test the causal nature of observed associations. In particular, I will examine sequencing, specificity, and selection through stratified analyses, structural equation modeling, latent trajectory analysis, directed acyclic graphs, propensity score matching, and instrumental variable analysis. The proposed research will be complemented by experiential and didactic training in (1) environmental approaches to decrease health disparities, (2) theory and measurement of social context and health, (3) trajectory analysis and lifecourse epidemiology, and (4) study designs and analytical approaches to inform causal inference. This research and training will enable me to launch a program of independent research on the urban built and social environments shaping childhood physical activity and health.