Evaluating and Applying Google Timeline Location Data for Built Environment and Physical Activity Research

项目总结 建筑环境直接影响健康，例如通过空气污染和噪音暴露，以及 通过改变与健康相关的行为间接影响健康，对此体力活动是主要的 贡献者。几乎所有的流行病学研究都考察了长期建筑环境与 相关的暴露和健康仅依赖于住址。这具有与以下方面相关的严重限制 暴露于住宅选择的错误分类和偏见，并且没有捕捉到个人如何互动 与建筑环境的关系。一些研究进行了GPS监测，但这些研究包括少量人口 (&lt；500人)和较短的时间(一到四周)，因为成本和后勤限制。我们一直在 评估谷歌位置历史时间线数据(称为谷歌时间线，或GTL)作为可行的来源 用于推进建筑环境和健康研究的长期被动GPS时间活动数据。GTL数据为 从Android和Apple智能手机(安装了谷歌地图或其他应用程序)收集以供选择加入 参与者。到目前为止，还没有健康研究试图利用GTL来改善暴露评估或 流行病学研究。我们将开发和评估从以下方面得出建筑环境暴露的方法 GTL数据，使用保护研究参与者隐私的方法，并在 华盛顿双胞胎注册研究的子集，他们已经参加了为期两周的GPS和 加速度学研究。我们还将开发从GTL数据评估体力活动的方法，提供 全新的、客观的体力活动水平衡量标准。如果我们成功地开发和开发了 使用GTL数据进行流行病学研究的评估方法这将彻底改变我们如何 研究建筑环境对人体健康的影响。

PROJECT SUMMARY Built environments directly influence health, for example through air pollution and noise exposures, and indirectly influence health through changing health-related behaviors, for which physical activity is a major contributor. Nearly all epidemiology studies examining the relationships between long-term built environment related exposures and health have relied solely on residential address. This has severe limitations related to exposure misclassification and bias from residential selection, and does not capture how individuals interact with the built environment. Some studies have conducted GPS monitoring, but these include small populations (<500 people) and short time-periods (one to four weeks) due to cost and logistical constraints. We have been evaluating Google Location History Timeline data (referred to as Google TimeLine, or GTL) as a viable source of long-term passive GPS time activity data for advancing built environment and health research. GTL data is collected from both android and apple smartphones (that have Google maps or other apps installed) for opt-in participants. To-date, no health study has attempted to leverage GTL for improving exposure assessment or epidemiological research. We will develop and evaluate methods to derive built environment exposures from GTL data, using approaches that protect the privacy of study participants and evaluate these methods within a subset of the Washington Twin Registry Study who have already participated in a two-week GPS and accelerometry study. We will also develop approaches to assess physical activity from GTL data, providing a completely novel and objective measure of physical activity levels. If we are successful in developing and evaluating methods to use GTL data for epidemiological research this will completely revolutionize how we study built environment influences on human health.