Environments, Preferences and Childhood Obesity: Evidence From a Natural Experiment

项目摘要 儿童和青少年肥胖仍然是21世纪世纪严重的公共卫生挑战。最近， 建筑和政策（B&P）环境在促进饮食、活动和BMI方面的作用已经得到了 重要的注意。但是，评估这些关系是具有挑战性的，因为考虑到选择和 异质性个人根据对健康和其他因素的偏好选择社区， 可能会影响估计。由于随机分配是不可行的，专家们呼吁自然的和准自然的分配。 利用外源变异的实验方法。此外，环境可能不会影响所有 个体之间的关系是平等的，并且可以平均到小的或零的影响，产生误导性的结论。等 到目前为止，异质性仅通过人口统计特征进行了研究。然而，证据显示， 环境的影响甚至在这些亚组中也是混合的，这表明其他因素可能在起作用。一个 新兴的文献表明，时间和风险（T&R）偏好（例如，未来导向和风险规避） 可以发挥这样的作用。T&R偏好差异很大，并与健康行为有关。 因此，在未来取向和风险规避方面存在差异的个体， 受到B&P环境的影响。然而，据我们所知，这还没有经过实证检验。 青少年是研究这些问题的一个关键亚群;他们的肥胖率已经翻了两番 在过去的三十年里，他们的T&R偏好，健康行为，以及与B&P的互动 环境正在向独立发展，因此现在是采取干预措施的关键时刻。我们的目标是 利用一个难得的机会来估计B&P环境对肥胖的影响， 行为和青少年的结果，同时解决选择和异质性。我们 建议扩大我们的美国国立卫生研究院资助的军事青少年的环境，运动和营养研究（M- TEENS），由于军队家庭的强制（重新）安置，它利用了一个自然实验。M-TEENS 收集了1519名12-13岁军人家庭儿童的三波饮食、活动和BMI数据。 我们提议的扩展将收集另外两波数据（2017年和2018年），产生5年纵向数据。 在B&P环境中，很少有外源性的儿童内变化。一个新的贡献， 这项拟议中的研究是检查环境的影响是否因青少年的年龄和他们的年龄而异。 父母的T&R偏好这是一个时间敏感的应用程序;我们的队列将完成高中在 结束拟议的数据收集，使其搬迁，从而环境，不那么外生 然后这项研究可能会产生很大的影响，因为它解决了一个非常重要的健康问题， 结合了自然的实验研究设计与纵向数据，并高度创新，其重点是 青少年和父母的偏好在理解环境影响中的作用。

PROJECT SUMMARY Child and adolescent obesity continue to be serious public health challenges of the 21st century. Recently, the role of the built- and policy (B&P) environments in contributing to diet, activity and BMI has received significant attention. But, evaluating these relationships is challenging due to concerns about selection and heterogeneity. Individuals select neighborhoods based on preferences towards health and other factors, which may bias estimates. Because random assignment is infeasible, experts have called for natural- and quasi- experimental methods that leverage exogenous variation. Moreover, environments may not affect all individuals equally and can average out to small or null effects, yielding misleading conclusions. Such heterogeneity has thus far only been examined by demographic characteristics. However, evidence on environments’ effects is mixed even within these subgroups, suggesting that other factors may be at work. An emerging literature suggests that time-and risk (T&R) preferences (e.g. future-orientation and risk-aversion) may play such a role. T&R preferences vary considerably and have been linked with health behaviors. Therefore, it is possible that individuals who vary in their future-orientation and risk-aversion are differentially impacted by the B&P environment. However, to our knowledge, this has not been examined empirically. Adolescents represent a key subpopulation in which to study these issues; their obesity rates have quadrupled during the past thirty years, and their T&R preferences, health behaviors, and interactions with the B&P environment are evolving towards independence, making this a critical time for interventions. Our goal is to exploit a rare opportunity to estimate the effects of B&P environments on obesogenic behaviors and outcomes among adolescents while addressing selection and heterogeneity. We propose to expand our NIH-funded Military Teenagers’ Environment, Exercise, and Nutrition Study (M- TEENS), which leveraged a natural experiment due to the compulsory (re)location of Army families. M-TEENS collected up to 3 waves of data on diet, activity and BMI from 1519 12-13 year old children in Army families. Our proposed extension will collect 2 additional waves of data (2017 & 2018), yielding a 5-year longitudinal panel with rarely-available exogenous within-child changes in the B&P environment. A novel contribution of the proposed study is the examination of whether the environment’s effects vary by adolescents’ and their parents’ T&R preferences. This is a time-sensitive application; our cohort will complete high-school at the end of the proposed data collection, making their relocations and, therefore environments, less exogenous thereafter. This study is likely to have a high impact because it addresses a highly significant health issue, combines a natural experiment study design with longitudinal data, and is highly innovative in its focus on the role of adolescents’ and parents’ preferences in understanding environments’ effects.