A longitudinal study of the association between internalizing problems during early adolescence and alcohol and cannabis use disorder at age 18 in a community sample

Introduction: Substance use disorders, such as alcohol use disorder (AUD) and cannabis use disorder (CUD), in young adults are an important public health concern. AUD and CUD may have deleterious impacts on the ongoing maturation of the brain and lead to impaired thinking, memory, emotional, and learning functions. Research on how internalizing problems among adolescents are associated with subsequent substance use disorders is limited and understanding these potential links may provide important insights as the prevalence of internalizing disorders has increased in the young adult population. This study assessed the longitudinal association between general internalizing problems in eighth grade and AUD and CUD at 18 years of age, as well as the unique associations of eighth grade depressive symptoms, physical symptoms of anxiety, and social anxiety symptoms with AUD and CUD at 18 years of age.
Methods: This analysis used data from a community sample of 521 adolescents recruited from Seattle public middle schools. We characterized general internalizing problems in eighth grade as a latent factor indicated by depressive symptoms, physical symptoms of anxiety, and social anxiety symptoms. Using structural equation modeling with a probit link, we examined the association between the latent factor and AUD and CUD at age 18. Unique associations of the three internalizing indicators with AUD and CUD independent of the internalizing problems latent factor were examined in separate models. All models adjusted for sex, highest level of parent/caregiver education, and previous substance use in the past six months at the 8th grade visit.
Results: We did not find evidence of an association between the internalizing problems latent factor and AUD or CUD in unadjusted (AUD: B = -0.049, 95% CI: -0.21, 0.12; CUD: B = 0.073, 95% CI: -0.095, 0.24) or adjusted (AUD: B = -0.017, 95% CI: -0.19, 0.16; CUD: B = 0.078, 95% CI: -0.095, 0.25) models. Nor did we find evidence of a unique association between any of the three individual internalizing indicators and AUD or CUD in adjusted models independent of the internalizing problems latent factor.
Conclusions: Our study found that internalizing problems and individual indicators of internalizing problems in eighth grade were not associated with AUD and CUD in early adulthood. Despite the lack of associations, a contribution of this study is the latent characterization of internalizing problems comprised of depressive symptoms and anxiety
symptoms indicators. Future studies can investigate if there are associations between internalizing problems during high school and substance use disorder outcomes to better understand the effect of internalizing problems when it is observed closer in time to the outcome.