rethinkTHC Search
Menu
Study breakdown

Finding the Right Antipsychotic Dose After Cannabis-Induced Psychosis

CohortModerate evidence
The takeaway

A nationwide Swedish study found that moderate antipsychotic doses were most effective at preventing psychotic relapse after cannabis-induced psychosis — higher doses didn't add benefit and likely increased side effects.

Psychiatrists treating cannabis-induced psychosis, patients prescribed antipsychotics after a psychotic episode, pharmacologists studying dose-response optimization.

What the researchers found

Using linked Swedish health registers, researchers identified all individuals with a first diagnosis of cannabis-induced psychosis and conducted a dose-response analysis of oral antipsychotic medications.

The analysis modeled antipsychotic exposure as time-dependent across three dose categories (low: <0.6 DDD, moderate: 0.6–<1.4 DDD, high: ≥1.4 DDD) using within-individual comparisons — meaning each person served as their own control across different exposure periods.

The primary outcome was hospitalization for any psychotic episode (schizophrenia-spectrum disorder or substance-induced psychosis). The results showed a dose-response curve that plateaued at moderate doses: low doses provided some protection, moderate doses were most effective, and high doses did not add further benefit.

The within-individual design is particularly strong because it eliminates confounding from differences between patients (genetics, severity, comorbidities) — it only compares how the same person fares at different dose levels.

Why it matters

Cannabis-induced psychosis carries a high conversion rate to chronic psychotic illness (RTHC-00251 found ~34% convert to schizophrenia). Antipsychotics can prevent relapse, but they carry dose-related side effects including weight gain, metabolic syndrome, and movement disorders. This study provides the first real-world dose optimization data, suggesting clinicians can achieve maximum benefit at moderate doses without the added burden of high-dose treatment.

The numbers in context

Three dose categories: <0.6 DDD (low), 0.6–<1.4 DDD (moderate), ≥1.4 DDD (high). Medications analyzed: aripiprazole, clozapine, risperidone, olanzapine, quetiapine, polytherapy, and others. Outcome: hospitalization for schizophrenia-spectrum or substance-induced psychosis. Moderate doses showed optimal protection; high doses showed no additional benefit.

How the study worked

Nationwide cohort study using linked Swedish administrative and healthcare registers. All first diagnoses of cannabis-induced psychosis (ICD-10 F12.5) identified. Antipsychotic exposure modeled as time-dependent using validated PRE2DUP method. Dose-response analyzed across three DDD categories. Within-individual stratified Cox regression. Primary outcome: hospitalization for psychotic episode.

Who was studied

N=1,772 individuals aged 16-64 years with first-time cannabis-induced psychosis in Sweden from 2006 to 2021

What this study cannot tell us

Register-based study — medication exposure is estimated from dispensing records, not confirmed adherence. Cannabis use during follow-up was not measured and may confound results. The Swedish healthcare system may not generalize to other settings. Within-individual design requires that patients contribute periods at different dose levels, which may introduce selection effects.

How to read the evidence

Nationwide cohort study with within-individual design — one of the strongest observational designs available, leveraging Sweden's comprehensive health registers.

When this study was published

Published in 2026 using Swedish register data, providing real-world evidence on antipsychotic dosing that complements clinical trial data.

The bigger picture

This connects directly to the psychosis conversion meta-analysis (RTHC-00251) by addressing what to do about that ~34% conversion rate. If moderate-dose antipsychotics can reduce relapse without the full side-effect burden of high doses, that changes the risk-benefit calculation for treating cannabis-induced psychosis — potentially making clinicians and patients more willing to use preventive medication during the critical post-episode period.

Replication

Not stated in abstract.

Funding

Not reported in abstract.

Conflicts of interest

Not reported in abstract.

Questions still open

  • Does the optimal dose differ by specific antipsychotic (e.g., aripiprazole vs. olanzapine)? How long should antipsychotic treatment continue after cannabis-induced psychosis? Would combining moderate-dose antipsychotics with cannabis cessation support further reduce relapse rates?

Read the original research

Optimizing antipsychotic dosing for relapse prevention in cannabis-induced psychosis: A nationwide cohort study.

Psychiatry research, 358, 116966

Psychiatry Research is a reputable journal focusing on mental health and psychiatric studies.

Citation

Mustonen, Antti; Niemelä, Solja; Denissoff, Alexander; Forti, Marta Di; Tanskanen, Antti; Mittendorfer-Rutz, Ellenor; Tiihonen, Jari; Taipale, Heidi. (2026). Optimizing antipsychotic dosing for relapse prevention in cannabis-induced psychosis: A nationwide cohort study.. Psychiatry research, 358, 116966. https://doi.org/10.1016/j.psychres.2026.116966

Explore the wider topic