Grant Number:	5R21CA287457-03 Interpret this number
Primary Investigator:	Harrell, Paul
Organization:	Old Dominion University
Project Title:	Does Tobacco Social Media Marketing Alter Adolescent Risk Perceptions and Use? Longitudinal Data-Adaptive Estimators and Causal Inference to Enhance Understanding
Fiscal Year:	2024

Project Summary Youth trajectories of tobacco use are of critical importance in the future of public health. Most tobacco use begins in adolescence and adolescents are particularly vulnerable to long-term tobacco use. While cigarette smoking significantly declined among adolescents over the past decade, overall tobacco/nicotine use actually increased by some measures, primarily due to increased rates of use of electronic nicotine delivery systems (ENDS). These changes in population levels of adolescent tobacco use are of critical importance in predicting tobacco control challenges and opportunities in the next decade. As the scientific debate unfolds regarding ENDS health effects, adolescents nationwide are learning about tobacco products and attempting to rectify conflicting information to create and refine their own beliefs regarding these products. At the same time, youth are increasingly living their lives filtered through social media account participation. Marketers are well aware of this fact and act to influence youth to purchase their products. Tobacco marketers are certainly no exception. Social media marketing is particularly important to examine as it is a prevalent source of adolescent exposure that is amenable to policy change. Social media marketing differs from other forms of marketing in that it exploits participatory facilitation, algorithmic coordination, context- specific promotions, real-world enmeshment, and seamless integration with purchasing opportunities. Although some research has examined this issue, there are important methodological concerns that may lead to flawed or incomplete understanding of the relationships involved and hinder the development of policies or other approaches to reduce the influence of social media marketing on youth ENDS use. We will address limitations in prior research by using both longitudinal data-adaptive, semiparametric estimators and a causal inference approach. We will focus on two risk perceptions: harmfulness and addictiveness. The degree to which risk beliefs act as mediators of the effects of social media marketing on use is unclear. Effective ENDS control strategies for adolescents require a more comprehensive understanding of how attitudes change over time and the potential impact of social media marketing on these changes The Population Assessment of Tobacco and Health (PATH) dataset provides a unique opportunity to examine youth beliefs about tobacco products, marketing exposure, and patterns of tobacco use among adolescents in the United States longitudinally. The current study will examine underutilized, recent Waves to examine the impact of online ENDS marketing on use with longitudinal data-adaptive, semiparametric estimators and to examine risk perceptions as potential mediators explaining the effects of adolescent marketing exposure on ENDS use with a causal inference approach.