Substance Use and Mental Health in Homeschooled Adolescents in the United States

52,089 adolescents (NSDUH 2015-2018)

Nationally representative US adolescents aged 12-17, classified as home-schooled; public/private school at low dropout risk; public/private school at risk for dropout; not in school

The design point is the comparison group. Previous work treated non-home-schooled adolescents as homogeneous; this study splits them by dropout risk, which is the single change that reverses the headline.

Substance use rates were LOWEST in schooled adolescents at low dropout risk, not in home-schoolers: that group had significantly lower past-year prescription opioid misuse, tobacco use, non-marijuana illicit drug use and nicotine dependence than home-schooled adolescents. Adolescents at risk for dropout had the highest rates. Psychopathology treatment prevalence was lowest among home-schoolers — which is at least as consistent with less access to and less detection by school-based referral as with better mental health.

High for the levels; low for the methodological point, which is about comparison groups rather than about causation.

Consistent with Green-Hennessy 2014 and Hodge 2016 on within-group heterogeneity, and directly reframes Vaughn 2015 using the same survey series.