Systems Science and the Dieting Paradox: Investigating the Complex Mechanisms through which Dieting Influences Obesity and Eating Disorder Risk

肥胖和饮食失调（例如，暴饮暴食，不健康的体重控制行为）是复杂的， 相关的公共卫生问题。节食在肥胖和饮食失调病因学中的作用是 争议虽然在行为减肥干预的背景下节食可以导致临床上 虽然节食可以显著减轻体重和暴饮暴食，但它也是肥胖和紊乱的一个强有力的风险因素。 吃这种“节食悖论”对制定和传播公共卫生信息提出了挑战 以及预防和治疗肥胖和饮食失调的干预措施。考虑到潜在的不良后果 事实上，如果不实施解决肥胖问题的战略， 不良肥胖相关健康后果的流行（例如，糖尿病），关键是要超越一个 “一刀切”的做法，并确定应促进哪些战略，为谁，根据什么 情节解开“节食悖论”并制定有效的量身定制和/或有针对性的干预措施 需要更好地理解节食的异质性。例如，许多支持节食的人 不要减少他们的热量摄入。然而，即使在没有减少能量摄入的情况下， 与不良情感后果和暴饮暴食有关。迄今为止，针对这一问题的研究 异质性还没有充分利用数学和计算工具，解释复杂性 这些与体重相关的行为、认知、情感体验和健康之间的相互作用是固有的。 此外，很少有纵向研究有足够的深度和广度来全面研究如何 不同的节食模式与体重和紊乱的饮食轨迹变化有关。解决 这些关于节食、饮食失调和肥胖的矛盾问题，并为学校的干预提供了信息， 医疗保健和社区环境，该项目将使用来自EAT项目研究的数据以及 通过结合两种强大的计算方法，拓扑数据分析（TDA）和代理 基于模型（ABM）。EAT项目非常适合本研究，因为其长期随访超过了关键的 生命过程的各个阶段，以及对肥胖和饮食失调相关变量的综合评估。TDA 将被用来表征不同的“节食悖论”亚组使用自我报告的节食，以及 与体重相关的行为和心理变量，并检查它们与饮食失调的关系 和体重状况的变化。TDA结果将用于开发、测试 并验证ABM，ABM将作为虚拟实验室，以测试病因假设和潜在的干预措施 解决方案利用这些前沿方法，这项工作的总体目标是更好地表征 体重和饮食失调的复杂性和具有临床意义的异质性 行为和产生超越“一刀切”的真实的世界预防和治疗办法 可能无意中对弱势亚群体造成不利后果的干预战略。

Obesity and disordered eating (e.g., binge eating, unhealthy weight control behaviors) are complex, inter- related public health problems. The role of dieting in the etiology of obesity and disordered eating is controversial. Although dieting in the context of behavioral weight loss interventions can lead to clinically significant weight and binge eating reductions, dieting is also a robust risk factor for obesity and disordered eating. This “dieting paradox” presents a challenge for developing and disseminating public health messages and interventions that prevent and treat obesity and disordered eating. Given potential adverse consequences of dieting and the fact that a failure to implement strategies to address obesity could lead to increased prevalence of adverse obesity-associated health consequences (e.g., diabetes), it is critical to move beyond a “one size fits all” approach and identify which strategies should be promoted, for whom, and under what circumstances. Untangling the “dieting paradox” and developing effective tailored and/or targeted interventions requires a better understanding of the heterogeneity of dieting. For example, many people who endorse dieting do not reduce their caloric intake. Even in the absence of energy intake reductions, however, dieting can be associated with adverse affective consequences and binge eating. To date, research addressing this heterogeneity has not fully leveraged mathematical and computational tools that account for the complexity inherent in the interplay between these weight-related behaviors, cognitions, affective experiences, and health. Moreover, few longitudinal studies have sufficient depth and breadth to comprehensively examine how different dieting patterns are associated with weight and disordered eating trajectory variations. To address these paradoxical issues around dieting, disordered eating, and obesity and inform interventions in school, healthcare, and community settings, this project will use data from the Project EAT studies and the advantages offered by combining two powerful computational approaches, topological data analysis (TDA) and agent based modeling (ABM). Project EAT is uniquely suited for this study given its long-term follow-up over critical life course periods and comprehensive assessment of variables relevant to obesity and disordered eating. TDA will be utilized to characterize different “dieting paradox” sub-groups using self-reported dieting, as well as weight-related behavioral and psychological variables, and to examine their association with disordered eating and weight status during the adolescent-to-adult transition. The TDA results will be utilized to develop, test, and validate ABMs that will serve as virtual laboratories to test etiologic hypotheses and potential intervention solutions. Utilizing these cutting edge approaches, the overarching goal of this work is to better characterize the complexity and clinically meaningful heterogeneity across the spectrum of weight and disordered eating behaviors and generate real world prevention and treatment approaches that move beyond “one size fits all” intervention strategies that could inadvertently lead to adverse consequences for vulnerable sub-populations.