68: Effects of social status and stress on adolescent smoking

Abstract

Purpose: Smoking in adolescence has been associated with increased perceived stress and lower socioeconomic status (SES). Furthermore, stress is a plausible pathway that may explain the relationship between SES and smoking. Recent work suggests that subjective perceptions of social status (SSS) within schools may also be important predictors of health outcomes independent of SES measures. This study investigated how SES, school SSS, and perceived stress were related to risk of smoking in adolescents. Methods: Longitudinal cohort study of 841 non-Hispanic black and white 7-12 graders from a Midwestern public school district. In the 2002-2003 academic year, students completed validated survey measures assessing school SSS, cigarette smoking behaviors, and perceived stress. Parents reported their own and their partner’s educational attainment (high school graduate or less, some college, college graduate or higher reference) as a measure of SES. The highest level of parent education (PE) was used in analyses. Students who reported having smoked in the last 30 days were classified as “current” smokers. Those who reported they had never tried smoking were considered “nonsmokers.” Among baseline nonsmokers, those who became current smokers at 1 year follow-up (FU) were classified as “initiators.” Those who had smoked in the past but were not current smokers at baseline were excluded. Hierarchical logistic regression first estimated the effect of PE, school SSS, and stress on risk of baseline current smoking and smoking initiation at 1 year FU and then assessed whether stress mediated any demonstrated social status-smoking relationships. Results: 44.8% were black and 51.6% female. 77.6% did not have a college educated parent. Mean stress levels were higher among those with lower PE ( high school or less 26.6; some college 24.8; college graduate or higher 24.3, p .0001) and lower school SSS (r -.22, p .0001). At baseline, 8.8% were current smokers. Both PE and SSS were associated with baseline current smoking. For PE, having a college educated parent was protective (OR some college 2.59, 95% CI 1.44-4.66, and OR high school degree or less 2.34, 95% CI 1.35-4.07). Higher school SSS was also protective (OR .76, 95% CI .66-.88). Higher stress increased risk of smoking (OR 1.06, 95% CI 1.02-1.09). Addition of the stress variable to the regression model did not suggest mediation. At FU, 8% of nonsmokers had become current smokers. Although PE was not associated with smoking initiation, lower school SSS did predict smoking initiation (OR 1.16, 95% CI 1.01-1.35), as did higher baseline stress (OR 1.07, 95% CI 1.03-1.12). The effect of school SSS was not mediated by stress. Conclusions: Youth with lower social status, whether measured objectively by SES or subjectively by SSS, are at increased risk of current smoking, but only lower SSS in the more local school environment predicted smoking initiation. Although higher stress is associated with both baseline smoking and smoking initiation and with lower social status, there was no evidence that stress accounts for the association between social status and smoking at either time point. Support: NICHD, William T. Grant Foundation.