Patterns of Anxiety, Stress, and Substance Use during Adolescence

The goal of the current study (N = 1263) was to investigate the relationship between type of substance use (cannabis, alcohol, cigarettes), level of substance use (non-user, regular user), and anxiety (somatic anxiety, perceived stress) over time in an adolescent population. This study examined independent and joint contributions of anxiety in predicting adolescent drug use over two years. It was hypothesized that higher levels of baseline anxiety would predict higher levels of substance use two years later. Results showed that while somatic anxiety and perceived stress independently predicted higher levels of cannabis and cigarette use, only perceived stress predicted future use when examined together; neither perceived stress nor somatic anxiety predicted alcohol use. Our second goal was to investigate baseline levels and changes in anxiety over time as a function of adolescent drug use and level of substance use groups (non-users, regular users). The study used hierarchical linear modeling to test the hypothesis that regular users would show higher levels of baseline anxiety and decrease anxiety faster over time than non-users for all substances. Results showed that both somatic anxiety and perceived stress decreased over time for everyone. Adolescents who smoked cigarettes and marijuana regularly showed higher levels of baseline perceived stress and somatic anxiety than non-users. However, perceived stress decreased at the same rate for all three substances and all levels of use. Importantly, all differences in perceived stress at baseline and changes over time are unlikely to be clinically significant given the small statistical estimates. Somatic anxiety was heightened at baseline for any adolescents smoking cigarettes or cannabis regularly. Additionally, somatic anxiety decreased over time faster for individuals who were smoking cigarettes and cannabis regularly, suggesting that the anxiolytic effects of cannabis and cigarettes may be effective at decreasing some aspects of anxiety. Level of alcohol use did not influence the baseline perception of stress but regular alcohol drinkers did show higher baseline somatic anxiety than non-users; regular drinking did not affect the decrease of perceived stress nor somatic anxiety over time. Potential explanations for the findings, study limitations, and suggestions for future research are discussed.