The impact of outdoor advertising on adolescent tobacco use: Exposures at home and school
Loading...
Date
2018-03
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Purpose: Prevalence of tobacco use among adolescents has remained stagnant since 2011.1 Major predictors of adolescent tobacco initiation are density of tobacco retailers near adolescents’ homes and schools. Further, adolescents who visit tobacco retailers most often, and are therefore exposed to tobacco product advertisements within the stores, are more likely to initiate tobacco use. An important gap in the literature is that we do not know whether the amount of tobacco product advertising on store exteriors affects adolescent tobacco use. We hypothesize that greater levels of exterior advertising at tobacco retailers around adolescents’ homes and schools will be associated with higher odds of tobacco use.
Research methods: A cohort of adolescent males ages 11-16 (N=1220) from urban and Appalachian Ohio were recruited in 2015 and 2016 and followed for one year. Ever use of any tobacco product was assessed at baseline and follow-up via self-report. Participants’ home and school addresses were also geocoded and spatially joined to census tracts. For the counties where the participants resided, fieldworkers conducted audits on a random sample of tobacco retailers (N=212) during the summer of 2016. These fieldworkers collected comprehensive information about retailers’ tobacco marketing, including the count and size of exterior tobacco advertisements. Exterior advertising scores based on count and size of exterior advertisements were calculated for each audited store. Median advertising scores were calculated by store type and then applied to all retailers of the same type in participants’ home and school census tracts. These median scores were summed to compute a total exterior advertising score per census tract. Odds of cigarette, electronic cigarette, or smokeless tobacco initiation at follow-up by density of tobacco retailers (i.e., sum of retailers per census tract) and amount of exterior advertising (i.e., sum of exterior advertising scores per census tract) in home and school census tracts were estimated using survey-weighted logistic regression. Due to linearity violations, density of retailers and exterior advertising score per census tract were modeled using weighted tertiles (low, medium, and high exposure levels). Adjusted, complete case analyses controlled for race, age, parental tobacco use, urban/Appalachian status, parental education, and census tract-level poverty. Due to small cluster sizes and low intraclass correlation coefficients, multi-level models were not used.5
Findings: Crude analyses revealed that odds of cigarette initiation were higher among participants living in census tracts with the highest retailer density compared to participants living in census tracts with the lowest retailer density (OR = 2.39; 95% confidence interval [CI]: 1.14, 5.02). Similarly, participants living in census tracts with high exterior advertising scores had greater odds of cigarette initiation than participants living in census tracts with low exterior advertising scores (OR = 2.71; 95% CI: 1.31, 5.59). In the adjusted analyses, participants living in census tracts with high exterior advertising scores still had greater odds of cigarette initiation compared to participants living in census tracts with low exterior advertising scores (OR=2.27; 95% CI: 1.11, 4.62). Neither retailer density nor exterior advertising score in school census tracts were associated with incident use of tobacco products.
Implications: This study is the first to demonstrate that quantity of exterior advertising in home census tracts is associated with greater odds of cigarette initiation among adolescents, even after controlling for several potential confounders of the association. Community- and individual-level tobacco control interventions should consider this as an important risk factor for adolescent tobacco use. Finally, policies to reduce the burden of exterior advertising at the census-tract level are worthy of exploration.
Description
Poster Division: Arts, Humanities, and Social Sciences: 2nd Place (The Ohio State University Edward F. Hayes Graduate Research Forum)
Keywords
Adolescents, Tobacco, Epidemiology, Advertising