rethinkTHC Search
Menu
Study breakdown

ABCD study tracked substance use in 9-13 year olds, finding increasing but still low rates

Longitudinal CohortStrong evidence
The takeaway

In the ABCD study of nearly 12,000 children followed from ages 9-13, substance use rates increased over time with 1.1% reporting cannabis use by age 12-13, while externalizing behaviors and parental drug problems predicted earlier initiation.

Parents, prevention researchers, school counselors, and policymakers concerned about youth substance use.

1.1% cannabis use by ages 12-13 in a nationally representative sample

What the researchers found

By age 12-13, 39.7% of children had experimented with any substance (mostly sipping alcohol), while 7.4% reported "full use" in their lifetime including 1.1% cannabis use. Yearly any-substance-use rates increased from 13.9% to 18.4%. Greater externalizing symptoms and parental substance problems predicted use initiation.

Why it matters

The ABCD study is the largest longitudinal study of brain development and substance use in American youth. Tracking when and how early substance experimentation begins provides essential data for prevention efforts.

The numbers in context

11,876 enrolled at baseline. By Y3 (ages 12-13): 39.7% experimented with any substance. Full use rates: 1.9% alcohol, 2.1% nicotine, 1.1% cannabis, 1.2% other. Externalizing symptoms and parental drug problems were significant longitudinal predictors.

How the study worked

Longitudinal analysis of the ABCD Study (11,876 participants at baseline, ages 9-10, with 6,251 at the third follow-up, ages 12-13). Descriptive statistics for all psychoactive substances over time. GEE models assessed predictors of substance use initiation.

What this study cannot tell us

Data only through age 12-13, when substance use rates are still low. Self-reported use may underestimate actual rates. Attrition could bias results if high-risk youth are more likely to drop out.

How to read the evidence

Strong: very large nationally representative longitudinal cohort with standardized assessments and multiple follow-ups.

When this study was published

Published in 2022.

The bigger picture

These data provide a baseline for understanding how adolescent substance use unfolds, confirming that experimentation begins early but full use remains uncommon through age 13. The identified risk factors can guide targeted prevention.

Questions still open

  • How will these rates change as the cohort enters high school? Do the identified risk factors become more or less predictive with age? Will early experimenters be more likely to develop problems later?

Common questions

How common was cannabis use among 12-13 year olds?
Only 1.1% reported cannabis use in their lifetime by ages 12-13. Alcohol sipping and nicotine experimentation were more common at this age.
What predicted earlier substance use?
Greater externalizing symptoms (behavioral problems like aggression and rule-breaking), parental substance use problems, and certain sociodemographic factors were associated with higher odds of initiating substance use.
Is experimentation the same as regular use?
No. The study distinguishes between "experimentation" (e.g., sipping alcohol, which 39.7% had done) and "full use" (a complete drink, nicotine product use, or cannabis use, which 7.4% reported). Most early exposure was experimentation.

Read the original research

Substance use patterns in 9 to 13-year-olds: Longitudinal findings from the Adolescent Brain Cognitive Development (ABCD) study.

Drug and alcohol dependence reports, 5

Citation

Sullivan, Ryan M; Wade, Natasha E; Wallace, Alexander L; Tapert, Susan F; Pelham, William E; Brown, Sandra A; Cloak, Christine C; Feldstein Ewing, Sarah W; Madden, Pamela A F; Martz, Meghan E; Ross, J Megan; Kaiver, Christine M; Wirtz, Hailey G; Heitzeg, Mary M; Lisdahl, Krista M. (2022). Substance use patterns in 9 to 13-year-olds: Longitudinal findings from the Adolescent Brain Cognitive Development (ABCD) study.. Drug and alcohol dependence reports, 5. https://doi.org/10.1016/j.dadr.2022.100120

Explore the wider topic