Adolescent substance use and high school noncompletion: exploring the nature of the relationship using a discordant twin design

Background and Aims: Previous studies have demonstrated associations between substance use and reduced educational attainment; however, many were unable to account for potential confounding factors like genetics and the rearing environment. In the few studies that controlled for these factors, the substances assessed were limited to alcohol, cannabis, and tobacco. To address these limitations, we examined the relationship between adolescent use of seven kinds of substances, the number of additional substances used, and high school noncompletion within a large sample of Australian twins. Design: A series of two-level generalized mixed effects logistic regressions were conducted to examine associations between adolescent substance use and high school noncompletion. Setting: Australia. Participants: A total of 9579 adult Australian twins from two cohorts of the Australian Twin Registry. Measurements: Assessments of high school completion, childhood major depression, conduct disorder symptoms, substance use initiation, demographics, and parental educational attainment using the Australian version of the Semi-Structured Assessment for the Genetics of Alcoholism. Findings: There were unique within-twin-pair effects of use of sedatives (odds ratio [OR] = 22.39 [95% confidence interval (CI) = 1.18–423.48]) and inhalants/solvents (OR = 10.46 [95% CI = 1.30–84.16]) on high school noncompletion. The number of substances used in adolescence was strongly associated with high school noncompletion across all discordant twin models (ORs from 1.50–2.32, Ps < 0.03). Conclusions: In Australia, adolescent substance use appears to be associated with early school dropout, with the effects of any given substance largely because of the confounding factors of parental education, childhood conduct disorder symptoms, and use of other substances. Sedatives and inhalants/solvents have effects on high school noncompletion that cannot be explained by polysubstance use or familial factors.