Substance familiarity in middle childhood and adolescent substance use

Abstract

Background: Childhood familiarity with (knowledge of) substances is a potentially important, currently understudied adolescent substance use risk factor. We aimed to describe changes in childhood familiarity with substances and to test whether baseline familiarity predicts early adolescent substance use.

Methods: Utilizing the Substance Use Module of the longitudinal cohort study, Adolescent Brain Cognitive Development (ABCD; US youth aged 9-10 years followed for 10 years) through Data Release 4 (n=7896; individuals who completed all six assessments in the first three years), we conducted longitudinal mixed models and survival analyses to describe changes in familiarity and to determine the adjusted odds of substance use by age 13 based on number of familiar substances at baseline.

Results: The sample consisted of 3754 females and 4142 males, aged 9-10 at baseline, with majority White individuals (68.9%). Unconditional time models indicated age significantly predicted familiarity (B=0.08, p<0.001; R²=0.288) with ~3.59 familiar substances at 9 years increasing to ~7.43 substances at 13 years. Family history, home use, peer use, and neighborhood availability predicted familiarity, accounting for 1% of additional variance (R²=0.299; ∆R²=0.011). For each additional familiar substance at baseline, adjusted odds of future use increased 1.28 times (95% CI 1.22, 1.34).

Conclusions: This is the first study to characterize substance familiarity in this age range as a predictor of future substance use. Familiarity increases with age (age being the most predictive indicator). Familiarity at age 9-10 predicts early adolescent substance use. As such, childhood familiarity may represent an easily implemented screening tool for at-risk youth.

Keywords: Adolescent; Cohort; Familiarity; Longitudinal; Risk factor.