Online Program

Finding the smoking gun: Could suspension high school increase smoking 12 years later?

*Janet E Rosenbaum, SUNY Downstate 

Keywords: adolescents, matched sampling, smoking, health disparities

Many adolescents experiment with smoking, but it’s unknown which factors predict long-term smoking among similar adolescents. School suspension has doubled in recent decades, and could impact smoking due to secondary deviance: the theory that youth punished harshly for experimental deviance will maintain/increase deviance. This study tests whether youth suspended from high school are more likely to smoke daily in young adulthood than similar non-suspended youth. We tested this hypothesis in the National Longitudinal Study of Adolescent Health, which includes 8638 participants never suspended at baseline (1995), of whom 548 were suspended in 1996. We compared smoking and marijuana use in 2008, at ages 25-31. To minimize confounding, we used Mahalanobis nearest-neighbor and exact matching methods to identify 1653 non-suspended young adults matched exactly on baseline smoking level and with similar levels of 19 other baseline socioeconomic, educational, and substance use factors. We estimated relative risk of smoking daily in 2008 using a Poisson working model in the matched sample, controlling for the same 20 baseline factors. Compared with non-suspended students, suspended students were twice as likely to smoke daily, use marijuana on the survey day, and report lifetime cannabis and drug dependence. After matching, suspended were 36% more likely to smoke daily (IRR 1.36 (1.02, 1.81), p=0.03) but did not vary in marijuana use or lifetime dependence, results consistent with secondary deviance. Subsequent analyses will attempt to distinguish between whether suspension caused or was just a marker for a propensity towards deviance.