Using Social Network Analysis to Understand Peer Influences on ENDS Use

摘要 可燃烟草（香烟）和电子尼古丁输送系统中的尼古丁消耗量 （ENDS）或电子烟已成为美国的一个严重健康威胁，特别是在青少年中。 几十年的研究已经证明了青少年的社交网络，特别是友谊 网络，影响烟草产品的开始，持续使用和品牌选择。然而，迄今为止， 没有关于友谊网络如何影响ENDS使用的数据或研究。本研究通过以下方式解决了这一差距： 建议在两个新发起的约2,500人的队列中增加友谊网络问题， 每个学校的学生都在测量ENDS的使用情况。已经为数据收集获得资金 这些队列将于2019年11月和2020年10月开始。IRB批准添加社交网络 在与南加州大学保护办公室主任协商后，问题已提交给IRB。 人类实验对象。要增加的友谊网络问题要求学生说出他们最亲密的人 在学校的年级朋友。所有同意的学生的名字都预先输入名册，以便 输入几个字母后姓名自动填充。这一过程已在其他非尼古丁 学习和工作都很好。这些数据将被用来测试青少年是否受到他们的影响。 朋友发起并继续使用ENDS，以及朋友是否影响品牌和口味的选择， 吸食大麻此外，网络选择过程将被测试，当人们做出 网络的变化与他们的行为一致。待检验的其他假设包括： 同伴的影响和选择是否：（1）在同性恋者和/或更强的联系中更强;（2）延伸到 双重或多用途;（3）发生的品牌和口味的选择;（4）发生的双重结束和大麻的使用。 鉴于数据的纵向性质，我们将构建ENDS和烟草使用轨迹，并确定 网络变化与不同的轨迹相关联。总而言之，这一建议是一个及时的 有机会为两个新发起的资助的以学校为基础的青少年群体增加一个重要的数据 ENDS研究;即友谊网络数据。初步数据来自1，616个横截面样本 南达科他州一个学区的学生显示， 朋友vaping。需要纵向数据来确定这种关系的方向并更多地应用 复杂的分析技术，如随机行为者导向模型。

Abstract Nicotine consumption in both combustible tobacco (cigarettes) and with electronic nicotine delivery systems (ENDS), or vaping, has become a serious health threat in the United States, particularly among adolescents. Decades of research have documented the many ways adolescent social networks, particularly friendship networks, influence the initiation, continued use, and brand choice of tobacco products. To date, however, there is no data or research on how friendship networks influence ENDS use. This study addresses this gap by proposing to add friendship network questions to two newly initiated cohorts of approximately 2,500 high school students each that are measuring ENDS use. Funding has already been acquired for the data collection for these cohorts and will start in November 2019 and October 2020. IRB approval to add the social network questions has been submitted to the IRB after consultation with the director of the USC office for the protection of human subjects. The friendship network questions to be added ask the students to name their closest friends in their grade at school. The names of all consented students are pre-entered into a roster so that the names auto-fill after a few letters are entered. This process has been implemented in other non-nicotine studies and works well. These data will then be used to test whether adolescents are influenced by their friends to initiate and continue ENDS use, as well as whether friends influence brand and flavor choices and marijuana uptake. In addition, network selection processes will be tested which occurs when people make network changes to be consistent with their behavior. Additional hypotheses to be tested include determining whether peer influence and selection: (1) are stronger among homophilous and/or stronger ties; (2) extends to dual- or poly-use; (3) occurs for brands and flavors choices; (4) occurs for dual ENDS and marijuana use. Given the longitudinal nature of the data we will construct ENDS and tobacco use trajectories and determine if network changes are associated with different trajectories. In sum, this proposal represents a timely opportunity to add a crucial piece of data to two newly initiated funded cohorts of school-based adolescent ENDS studies; namely, friendship network data. Preliminary data from a cross-sectional sample of 1,616 students in one school district in South Dakota showed a strong correlation between individual vaping and friend vaping. Longitudinal data are needed to determine the direction of this relationship and apply more sophisticated analytic techniques such as stochastic actor-oriented models.