Mobile Ecological Momentary Diet Assessment: A Low Burden, Ecologically-Valid Approach to Measuring Dietary Intake in Near-Real Time

摘要 饱和脂肪和添加糖的过量摄入已被确定为早产的主要原因。 美国成年人的死亡率每年导致大约70万人死亡。2015-2020年 美国人膳食指南建议将这些营养素限制在总能量摄入量的10%以下，以防止 疾病实现这些公共健康建议将需要了解饱和脂肪的模式 和增加糖的摄入量，以便开发更有效的饮食干预措施。传统上， 饱和脂肪和添加糖的摄入量是通过食物频率问卷或24小时饮食回忆来测量的 （24小时）。这些方法对用户来说是时间密集型的，对用户来说是认知负担，对研究人员来说是昂贵的。他们是 也很容易回忆偏见和误报有关，部分原因是依赖于一个人的记忆， 长的回忆间隔和部分大小估计中的错误。建议的膳食评估方法旨在 通过生态瞬时评估（EMA）解决这些限制。EMA使用最新技术， 可以更新和改进传统评估方法的抽样方法。EMA研究经常使用 移动的手机应用程序，通过全天定期提供的简短自动调查来评估事件。 因此，EMA可以缩短召回间隔，以改善报告错误，减少用户和研究人员的负担 同时最大化生态效益。迄今为止，用于饮食评估的移动的EMA方法（mEMDA）用于 研究是具体的。它们没有得到系统的开发，也没有得到广泛的优化。 用于研究。该项目将代表第一个研究质量和全自动化，基于EMA的移动的 饮食评估研究工具。在最近的试点工作中，我们证明了mEMDA的潜在效用。简要 进行移动的调查以及网络辅助的24小时调查，以估计预定义的零食摄入量。在这里， 拟议项目的目标是系统地开发和测试mEMDA应用程序和采样方法， 准确估计不同人群中饱和脂肪和添加糖的摄入量。为此，我们将导出 一份在文化和人口统计学上具有代表性的食品和饮料清单， 使用最近的NHANES数据（目标1）来研究美国饮食中的饱和脂肪和添加糖;制定 以用户为中心设计mEMDA应用程序和分析平台，并使用视觉食物图像进行份量估计 和营养分析能力（目标2）;确定最佳的mEMDA采样方法（事件相关与 区间视情况抽样）（目标3）;并比较从饱和脂肪估计能量摄入的准确性 使用优化的mEMDA应用程序和采样方法与采访者辅助的24小时， 控制喂养研究。mEMDA应用程序的未来应用包括：（1）可靠地评估瞬时 其他食物或营养素的摄入（例如，水果和蔬菜摄入量，钠），（2）与移动的整合 为参与者提供实时饮食反馈的干预平台（3）同时捕获膳食背景 变量（例如，社会、环境和心理社会变量），以便将来及时进行饮食干预。

ABSTRACT The excessive intake of saturated fat and added sugars has been identified as a leading cause of premature mortality among adults in the U.S. contributing to approximately 700,000 deaths each year. The 2015-2020 Dietary Guidelines for Americans recommend limiting these nutrients to <10% total energy intake to prevent disease. Achieving these public health recommendations will require understanding the patterns of saturated fat and added sugar intake so more effective dietary interventions can be developed. Traditionally, estimates of saturated fat and added sugar intake are measured using food frequency questionnaires or 24-hr dietary recalls (24HR). These methods are time-intensive and cognitively taxing for users and costly for researchers. They are also highly prone to recall bias and misreporting related due, in part, to the reliance on a person’s memory over long recall intervals and errors in portion size estimation. The proposed dietary assessment method aims to address these limitations with ecological momentary assessment (EMA). EMA uses updated technology and sampling methods that can update and improve upon traditional assessment methods. EMA studies often use mobile phone apps to assess events with brief, automated surveys delivered periodically throughout the day. EMA can, thereby, shorten recall intervals to improve reporting errors and reduce user and researcher burden while maximizing the ecological validity. To date, mobile EMA methods for diet assessment (mEMDA) used in research have been study specific. They have not been systematically developed nor optimized for widespread use in research. This project would represent the first research-quality and fully automated, EMA-based mobile dietary assessment research tool. In recent pilot work, we demonstrated the potential utility of mEMDA. A brief mobile survey performed as well as web-assisted 24HR to estimate the intake of predefined snack foods. Here, the goal of the proposed project is to systematically develop and test a mEMDA app and sampling approach to accurately estimate the intake of saturated fat and added sugars in a diverse population. To do this we will derive a culturally- and demographically representative list of foods and beverages that contribute a majority (>70%) of the saturated fat and added sugars in the American diet using recent NHANES data (Aim 1); develop with a user-centered design the mEMDA app and analysis platform with visual food images for portion size estimation and nutrient analysis capabilities (Aim 2); determine the best mEMDA sampling approach (event-contingent vs. interval-contingent sampling) (Aim 3); and compare the accuracy of estimating energy intake from saturated fat and added sugars using the optimized mEMDA app and sampling approach vs. interviewer-assisted 24HR in a controlled-feeding study. Future applications of the mEMDA app include: (1) reliably assessing momentary intakes of other foods or nutrients (e.g., fruit and vegetable intake, sodium), (2) integration with mobile intervention platforms to give real-time, dietary feedback to participants (3) concurrent-capturing meal context variables (e.g., social, environmental, and psycho-social variables) for future, just-in-time dietary interventions.