rethinkTHC Search
Menu
Study breakdown

Cannabis Use Creates Opposite Brain Inflammation Patterns in People With and Without Psychosis

Cross SectionalModerate evidence
The takeaway

Cannabis use was linked to lower brain inflammation markers in healthy controls but higher inflammation in people with early psychosis — and may blunt the brain effects of antipsychotic medication.

Psychosis clinicians, early intervention service providers, neuroimaging researchers

What the researchers found

Past cannabis use was associated with lower extracellular free water (FW, an inflammation marker) in controls but elevated FW in people with recent-onset psychosis, particularly in temporal and parietal cortex. Within psychosis patients, antipsychotic exposure reduced FW only in non-cannabis users — cannabis users showed no antipsychotic benefit on FW.

Why it matters

If cannabis blocks antipsychotics from reducing brain inflammation in psychosis, it could explain why cannabis-using psychosis patients have poorer treatment outcomes — and suggests cannabis cessation may be particularly important for medication effectiveness.

The numbers in context

62 ROP and 38 controls. Group × cannabis interaction significant in average GM (p=0.049) and temporal-parietal cortex (TFCE p-FWE<0.05). Antipsychotic × cannabis interaction in WM (p=0.005) and GM (p=0.073). Antipsychotics reduced FW only in non-cannabis-using patients.

How the study worked

Cross-sectional diffusion MRI measuring extracellular free water in 62 individuals with recent-onset psychosis and 38 controls, stratified by cannabis use history. Group × cannabis interactions analyzed with TFCE correction.

What this study cannot tell us

Cross-sectional design limits causal interpretation. Moderate sample size. Cannabis use was historical, not current. Free water is a proxy for inflammation, not a direct measure. Multiple comparisons may inflate some findings.

How to read the evidence

Innovative diffusion MRI approach with appropriate corrections, limited by cross-sectional design and moderate sample requiring replication.

When this study was published

Published 2026, using extracellular free water as a neuroinflammation biomarker.

The bigger picture

This study adds a neurobiological mechanism to the well-known clinical observation that cannabis worsens psychosis outcomes. The finding that cannabis may block antipsychotic brain effects provides a concrete target for improving treatment through cannabis cessation support.

Questions still open

  • Does cannabis cessation restore antipsychotic effectiveness on brain inflammation? Is the cannabis-psychosis FW interaction mediated by the endocannabinoid system? Could adjunctive anti-inflammatory treatments help cannabis-using psychosis patients?

Common questions

Does cannabis cause brain inflammation in people with psychosis?
This study found cannabis use was associated with higher brain inflammation markers in people with early psychosis but lower markers in healthy people — suggesting cannabis interacts with the psychotic brain in a fundamentally different way.
Can cannabis affect how well antipsychotic medications work?
The data suggest yes — antipsychotic medication reduced brain inflammation markers only in psychosis patients who didn't use cannabis. In cannabis users, antipsychotics showed no effect on these markers, potentially explaining poorer treatment outcomes.

Read the original research

Differential effect of cannabis use and antipsychotic medication on extracellular free-water in the brain of individuals with early psychosis and controls.

Molecular psychiatry, 31(1), 362-373

Citation

Martínez-Sadurní, Laura; Barrera-Conde, Marta; Robledo, Patricia; Veza-Estevez, Emma; Garcia-Quintana, Jordi; Mané, Anna; Toll, Alba; Trabsa, Amira; Lesh, Tyler A; Carter, Cameron S; Bergé, Daniel. (2026). Differential effect of cannabis use and antipsychotic medication on extracellular free-water in the brain of individuals with early psychosis and controls.. Molecular psychiatry, 31(1), 362-373. https://doi.org/10.1038/s41380-025-03287-4

Explore the wider topic