Cannabis Use in Bipolar Disorder Was Linked to Earlier Onset and More Psychosis, But Not More Relapses | Research

The takeaway

In a 3-year study of 279 bipolar patients, cannabis use was associated with earlier illness onset, more psychotic symptoms, and higher treatment discontinuation, but not with more hospital readmissions.

Psychiatrists treating bipolar disorder, addiction psychiatrists, researchers studying substance-mood disorder comorbidity.

Cannabis predicted worse psychotic symptoms in bipolar disorder, but tobacco predicted more relapses

What the researchers found

Cannabis was associated with earlier onset of bipolar disorder, current manic polarity, presence of psychotic symptoms, and higher likelihood of discontinuing treatment. However, cannabis was NOT associated with higher rates of ER visits or hospital readmissions. Interestingly, tobacco use showed significantly higher readmission rates. Alcohol, cocaine, and stimulants showed no association with studied variables.

Why it matters

The distinction between cannabis affecting illness severity (earlier onset, more psychosis) but not relapse rate (readmissions) is clinically important. It suggests cannabis may influence the type and quality of bipolar episodes more than their frequency.

The numbers in context

n=279 patients; 3-year follow-up; cannabis associated with earlier onset, manic polarity, psychotic symptoms, and treatment discontinuation; tobacco associated with higher ER and readmission rates; no significant associations for alcohol, cocaine, or stimulants.

How the study worked

Prospective cohort study of 279 patients admitted to Hospital Clinic of Barcelona for manic or mixed episodes (2015-2019), with 3-year follow-up examining hospital readmissions, ER visits, and clinical characteristics.

What this study cannot tell us

Single-center study in Barcelona. Lacks detailed information on cannabis use patterns, history, and consumption amounts. Loss to follow-up from patients leaving the region or switching to private care. Cannot establish causation.

How to read the evidence

Moderate: Prospective cohort with 3-year follow-up from a major academic center, though single-site design and incomplete consumption data are limitations.

When this study was published

Published in 2025 with data from 2015-2019.

The bigger picture

This study adds nuance to the substance use and bipolar disorder literature by examining multiple substances and multiple outcomes simultaneously. The finding that tobacco, not cannabis, predicted relapse rate challenges assumptions about which substances are most harmful for bipolar prognosis.

Questions still open

Why did tobacco predict readmissions while cannabis did not? Does cannabis affect the quality of bipolar episodes (more psychotic features) without affecting frequency? Would cannabis cessation reduce psychotic symptoms in bipolar patients?

Common questions

Does cannabis make bipolar disorder worse?

In this study, cannabis was associated with earlier onset and more psychotic symptoms during bipolar episodes, but not with more frequent hospitalizations. This suggests cannabis may affect the severity of episodes rather than how often they occur.

Why was tobacco a stronger predictor of relapse than cannabis?

The study found tobacco use, not cannabis, was significantly associated with more ER visits and hospital readmissions. The reasons are unclear but may relate to differences in dependence patterns, health effects, or the role of nicotine in bipolar neurobiology.

Read the original research

Cannabis and tobacco use in bipolar disorder: Associations with early onset, psychotic symptoms, and relapse risk (2015-2019).

Journal of affective disorders, 382, 30-38

View on PubMed ↗Publisher’s paper ↗

Citation

Olivier, Luis; Andreu, Helena; de Juan, Oscar; Ochandiano, Iñaki; Salmerón, Sergi; Fernández-Plaza, Tábatha; Colomer, Lluc; Vieta, Eduard; Giménez-Palomo, Anna; Pacchiarotti, Isabella. (2025). Cannabis and tobacco use in bipolar disorder: Associations with early onset, psychotic symptoms, and relapse risk (2015-2019).. Journal of affective disorders, 382, 30-38. https://doi.org/10.1016/j.jad.2025.04.026