Autism and Problematic Substance Use in Adolescence: Evidence From the Millennium Cohort Study

J Adolesc Health. 2026 Jul 10:S1054-139X(26)00187-4. doi: 10.1016/j.jadohealth.2026.05.003. Online ahead of print.

ABSTRACT

PURPOSE: Prior research on the relationship between autism and substance use in adolescence is limited, and the findings are inconsistent. This study examined the association between autism and substance use in adolescence using a large representative sample.

METHODS: Data came from the UK Millennium Cohort Study, a nationally representative sample of children born in the UK in 2000-2002. Problematic use of alcohol, tobacco smoking, cannabis, and other substances was measured at ages of 14 and 17 years (n = 12,243). The sample was 3.75% autistic (n = 459). Binary and proportional odds of logistic regression models assessed the association between autism and problematic substance use. Consistency of problematic use between ages of 14 and 17 years was also tested.

RESULTS: At age of 14 years, any problematic substance use was reported by 18% of autistic adolescents versus 22% of nonautistic peers; at age of 17 years, rates were 38% versus 56%. Autism significantly predicted lower odds of overall problematic use at both ages, independent of intelligence and other confounders. Autism also predicted lower odds of problematic alcohol use at both ages of 14 and 17 years. No differences by autism status were found for tobacco smoking, cannabis, or other substances. Consistency of problematic use between ages of 14 and 17 years was similar for autistic and nonautistic adolescents.

DISCUSSION: Autistic adolescents showed lower odds of problematic substance use, independently of intelligence. The contrast between this finding and those from studies using adult samples indicates that the relationship may reverse during adulthood.

PMID:42429702 | DOI:10.1016/j.jadohealth.2026.05.003