Psychosis Symptoms Increased Before Teens Started Using Cannabis, Not Just After | Research

The takeaway

In a large ABCD Study analysis, psychosis spectrum symptoms were rising in the period leading up to cannabis initiation, supporting self-medication and shared vulnerability models rather than cannabis as a simple causal trigger.

Psychiatrists, cannabis-psychosis researchers, adolescent mental health clinicians, drug policy analysts.

Psychosis symptoms increased BEFORE teens started cannabis, not just after

What the researchers found

Adolescents who used cannabis at any point had more psychosis symptoms (B=0.86) and distress from symptoms (B=1.17) than never-users, consistent with shared vulnerability. Symptoms increased in the time leading up to cannabis initiation (B=0.16 for symptoms, B=0.23 for distress), consistent with self-medication. Evidence for post-initiation increases (contributing risk) was mixed.

Why it matters

This is one of the most rigorous tests of the cannabis-psychosis causal question. By modeling symptom trajectories before and after cannabis initiation, it provides evidence that the relationship is more complex than "cannabis causes psychosis," with shared vulnerability and self-medication playing important roles.

The numbers in context

n=11,858; mean age 9.5 at wave 1; 52% male; 4-year follow-up; shared vulnerability: B=0.86 (symptoms), B=1.17 (distress); pre-initiation increase: B=0.16 (symptoms), B=0.23 (distress); mixed evidence for post-initiation contributing risk.

How the study worked

Cohort study using 5 waves across 4 years from the ABCD Study (n=11,858 adolescents aged 9-10 at baseline). Discontinuous growth curve modeling assessed psychosis symptom trajectories before and after cannabis initiation, adjusting for age, sex, other substance use, SES, and parental mental health.

What this study cannot tell us

ABCD Study follow-up is still relatively short (4 years, ages ~10-15). Cannabis use in this young sample may not represent adult patterns. Self-reported psychosis symptoms are not the same as clinical psychotic disorders. Cannot fully rule out unmeasured confounders.

How to read the evidence

Strong: Large longitudinal ABCD Study cohort published in JAMA Psychiatry with sophisticated growth curve modeling and comprehensive adjustment for confounders.

When this study was published

Published in 2025 using ABCD Study data with 4-year follow-up.

The bigger picture

This JAMA Psychiatry study shifts the narrative from a simple causal model (cannabis causes psychosis) to a more nuanced understanding where genetic vulnerability, pre-existing symptoms, and cannabis use may all interact. This has implications for both clinical screening and policy.

Questions still open

Will longer follow-up reveal clearer post-initiation effects? Should psychosis screening precede cannabis prevention efforts? How should clinical guidelines incorporate the self-medication finding?

Common questions

Does this mean cannabis does not cause psychosis?

Not exactly. The study found mixed evidence for post-initiation symptom increases. What it clearly shows is that the relationship is more complex than simple causation: teens who later use cannabis already have more psychosis symptoms and these symptoms are rising before they start using, suggesting shared vulnerability and self-medication are also at play.

What is the self-medication hypothesis?

The self-medication hypothesis suggests that some people use cannabis to cope with existing psychological distress. This study found psychosis symptoms and associated distress were increasing in the time leading up to cannabis initiation, consistent with teens turning to cannabis to manage emerging symptoms.

Read the original research

Psychosis Spectrum Symptoms Before and After Adolescent Cannabis Use Initiation.

JAMA psychiatry, 82(2), 181-190

View on PubMed ↗Publisher’s paper ↗

Citation

Osborne, K Juston; Barch, Deanna M; Jackson, Joshua J; Karcher, Nicole R. (2025). Psychosis Spectrum Symptoms Before and After Adolescent Cannabis Use Initiation.. JAMA psychiatry, 82(2), 181-190. https://doi.org/10.1001/jamapsychiatry.2024.3525