Dynamics of Health and Adolescent Social Networks

描述（由申请人提供）：本项目的总体目标是开发和测试同伴网络、健康行为和心理健康共同进化的动态模型，该模型可用于更好地针对健康干预措施。长期以来，研究人员一直在研究社交网络与健康各个方面（包括抑郁和健康行为）之间的联系。然而，这方面的大部分工作都将网络视为固定的和外生的。在这个项目中，我们使用动态网络数据统计建模的最新进展，研究健康因素可能积极影响网络创建和演变的方式。我们使用来自国家青少年健康纵向研究（AddHealth）的数据和基于行为者的模型来调查社会影响和同伴选择过程在理解同伴网络、心理健康和健康行为之间相互关系方面的相对贡献。该项目的研究结果将对新兴的社会网络动态研究以及长期存在的关于同伴对健康影响的文献作出重大贡献。具体来说，这个项目做出了三个创新贡献：(1)我们明确地建立了基于个人特征（包括健康行为（选择））的关系形成可能性模型，并将其与个人与朋友同化时的行为变化（影响）区分开来。在这样做的过程中，我们调查了哪种类型的青少年更有可能选择具有特定健康行为的同伴，或者相反，更有可能改变自己的行为以符合朋友的行为。(2)通过使用AddHealth，我们能够在代表广泛背景的多所学校中检查这些问题。(3)我们利用一系列计算机模拟，其参数由经验数据估计，作为研究多种干预策略的手段，并评估其减少吸烟，酒精使用和抑郁症患病率的潜力。公共卫生相关性：该项目通过加强我们对青少年健康行为和心理健康发展所涉及的社会网络动态的理解，为公共卫生做出贡献。这些发现对减少青少年吸烟和饮酒以及改善心理健康的干预措施的发展具有重要意义。

DESCRIPTION (provided by applicant): The overarching goal of this project is to develop and test a dynamic model of the co-evolution of peer networks, health behavior, and psychological well-being, which can be used to better target health interventions. Researchers have long investigated the link between social networks and various aspects of health including depression and health behaviors. However, most of this work has viewed networks as fixed and exogenous. In this project we examine the ways in which health factors may actively influence network creation and evolution over time using recent advances in statistical modeling of dynamic network data. We use data from the National Longitudinal Study of Adolescent Health (AddHealth) and an actor-based model to investigate the relative contributions of social influence and peer selection processes in understanding the interrelationships between peer networks, psychological well-being and health behavior. The findings from this project will serve as a major contribution to the emerging research on social network dynamics as well as to long standing literature on peer effects on health. Specifically, this project makes three innovative contributions: (1) We explicitly model the likelihood of relationships forming based upon individual traits, including health behavior (selection), and differentiate this from changes in behavior as individuals assimilate to their friends (influence). In so doing, we investigate what types of adolescents are more at risk of selecting peers with particular health behavior or conversely are at risk of changing their behavior to conform to friends' behavior. (2) By using AddHealth, we are able to examine these questions in multiple schools that represent a wide range of contexts. (3) We utilize a series of computer simulations, whose parameters are estimated from empirical data, as a means to investigate multiple intervention strategies and evaluate their potential for reducing the prevalence of smoking, alcohol use, and depression. PUBLIC HEALTH RELEVANCE: This project contributes to the public health by enhancing our understanding of the social network dynamics involved in the development of adolescent health behaviors and psychological well-being. The findings have important implications for the development of interventions to reduce adolescent smoking and drinking and improve psychological health.