TOBACCO OUTLET DENSITY, PROXIMITY, AND TOBACCO USE AMONG AFRICAN AMERICAN YOUNG ADULTS IN BALTIMORE CITY: MECHANISMS OF ACTION

Abstract Objectives: Research exploring the relationship between tobacco use and exposure to tobacco outlets via density and proximity is inconclusive, and no studies focus exclusively on African American young adults – although this population is often the target of tobacco advertisements and exposed to higher levels of tobacco outlet density. To better understand the relationship between tobacco outlet exposure and tobacco use among African American young adults research is needed on the mechanisms driving this relationship. This study explored three specific mechanisms – moderation, mediation, and growth. The specific aims of this investigation were 1. to assess the moderating role of gender in the relationship between tobacco outlet exposure (i.e., density and proximity to outlets) and past month tobacco use among African American young adults living in Baltimore City, Maryland. 2. to assess the mediating role of perceived harmfulness and disapproval of cigarette use in the relationship between tobacco outlet exposure and past month tobacco use among African American young adults living in Baltimore City, Maryland, and 3. to assess the correlation between the growth of tobacco outlet density and past month tobacco use among African American young adults living in Baltimore City, Maryland. Methods: Each aim used data from the Johns Hopkins Second Generation Baltimore Prevention Program (BPP) cohort. The BPP is a field trial of two universal first grade interventions, respondents are predominately African American, and were followed through adulthood. Assessments of tobacco and other drug use began in the sixth grade. Geospatial methods were used to determine the density of tobacco outlets surrounding participants’ homes’, as well as how close participants’ lived to the nearest tobacco outlet. Aim 1 was assessed using logistic regression models via generalized estimating equations. Aim 2 was assessed using path analyses. Aim 3 used parallel process growth curve modeling to assess the simultaneous growth of tobacco outlet density and past month tobacco use. Results: In Aim 1, sex modified the relationship between tobacco outlet density and past month tobacco use, and this relationship was only significant among women, such that after adjusting for confounding, the relationship between past month tobacco use and tobacco outlets that sold tobacco but not alcohol was positive and significant (aOR = 1.02; p < 0.05). Additionally, among women, the relationship between past month tobacco use and tobacco outlets that sold both tobacco and alcohol was positive and significant (OR = 1.04; p < 0.05), however, after adjusting for confounding the relationship was no longer statistically significant (aOR = 1.04; p = 0.12). Proximity to the nearest tobacco outlet was not significantly associated with past month tobacco use for men or women. On average, people lived closer to and the density per quarter mile was higher for outlets that sold tobacco, but not alcohol, relative to outlets that sold both tobacco and alcohol. For example, on average, people lived within 552 meters (standard deviation [SD] = 459 meters) of an outlet that sold tobacco, but not alcohol, and within 658 meters (SD = 500 meters) of an outlet that sold both tobacco and alcohol. Furthermore, there were on average 16.4 outlets per quarter mile (SD = 26.4) that sold tobacco only, whereas the average density for outlets that sold both tobacco and alcohol was 7 outlets per quarter mile (SD = 11.3). In Aim 2, disapproval of cigarette use, but not perceived harmfulness of cigarette use was a significant mediator in the relationship between tobacco outlet density and past month tobacco use, and results varied by gender, such that disapproval of cigarette use significantly mediated the relationship between tobacco outlet density and past month tobacco use for men but not women. The relationship between proximity to the nearest tobacco outlet and past month tobacco use was not mediated by disapproval or perceived harmfulness of cigarette use for men or women. In Aim 3, correlations between several growth factors were evaluated, of primary interest was the correlation between the random effects on the slopes of past month tobacco use and tobacco outlet density. This relationship was not significant (r = 0.31; p-value = 0.08). However, the correlation between the random effects on the intercepts of past month tobacco use and tobacco outlet density was statistically significant (r = 0.19; p-value < 0.05). Conclusions: This study has implications for research, practice and policy. For example, public health practitioners aiming to reduce tobacco use among African American young adults living in areas where tobacco outlet density is high should consider gender-specific interventions as this study implies that the relationship between tobacco outlet density and tobacco use, to include mediating factors, differ by sex. Furthermore, policy should focus on zoning restrictions that prevent the establishment of tobacco outlets in residential neighborhoods, recognizing that neighborhoods where tobacco outlet density is initially high may be associated with initial decisions to use tobacco among African American young adults. Future research should continue evaluating mechanism driving the relationship between exposure to tobacco outlets and tobacco use among this population. Such research will potentially reveal other mediators and moderators that are malleable to intervention, and this will lead to comprehensive preventive efforts that impose restrictions on tobacco outlet density and zoning, while simultaneously addressing other correlates of tobacco use