Loading ...

MADRID — Recent research on the progression and treatment of psychosis is prompting some experts to ask: Are substance-induced psychosis (SIP) and primary psychosis fundamentally different? And if they are, does it matter?

Treatment of first-episode psychosis (FEP) in patients who also use cannabis is often a challenge because it is frequently unclear to clinicians whether a patient’s cannabis use preceded or followed the psychosis, and treatment guidelines are vastly different for each.

While treatment for SIP involves abstinence and sometimes a short hospitalization in an acute psychiatric ward, the recommended treatment for FEP is often up to 1 year of antipsychotic medication. But studies also show that 1 in 3 with cannabis-induced psychosis (CIP) is later diagnosed with schizophrenia.

Given the diagnostic overlap and evidence from studies showing transition from SIP to schizophrenia, “How should we understand what’s going on in our data?” Eline Borger Rognli, PhD, Department of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway, asked at a recent symposium at the European Psychiatric Association Congress 2025.

“Would these people have developed schizophrenia anyway? Or did they do so because they developed substance-induced psychosis and then schizophrenia?” Rognli said. “Were these, in fact, transitions from one condition to a different condition, or were these cases developing…but masked by cannabis use?”

Other experts at the symposium raised the same questions. Given the growing recognition of the close, bidirectional association of these conditions and emerging evidence that treatment should be the same for both, should the diagnostic characterization of psychosis be more fluid?

Blurring the Diagnostic Lines

In a recent paper, Rognli and other presenters at the meeting questioned the diagnostic term “substance-induced psychosis.”

“Just by the name, it’s assumed that the psychosis is caused by substance use, and that type of explanatory diagnostic terms is something that we’ve broadly left behind,” she said. “With the introduction of the DSM-III [Diagnostic and Statistical Manual of Mental Disorders, Third Edition], diagnosis should be descriptive, acknowledging that we don’t completely understand the reason why disorders develop; we can just describe them.”

Classifying a formal diagnosis as substance-induced vs primary psychosis has important implications, she added.

“In Norway, if you have committed a severe, violent crime while you were psychotic, whether you have a primary psychosis versus a substance-induced psychosis will determine if you will be sentenced to receive treatment at the hospital or if you will be incarcerated,” Rognli said.

In terms of treatment, the distinction may actually be clinically irrelevant on the basis of several studies suggesting the benefits of antipsychotic medications are similar in both SIP and primary psychosis.

New data presented at the meeting from a population-based cohort of patients with CIP showed that subsequent treatment with antipsychotics decreased their risk of future hospitalization for any psychotic relapse, whether the psychosis was substance-induced.

Led by Heidi Taipale, PhD, senior researcher at Niuvanniemi Hospital, Kuopio, Finland, and Karolinska Institutet, Stockholm, Sweden, the analysis included data gathered from the prescribed drug register in Sweden from 2005 to 2023 and included 1772 patients with CIP (mean age, 26.6 years; 84% men). Using a within-individual design, where each individual was compared with themselves, periods of medication were compared with non-use periods over a mean follow-up of 8 years.

Overall, any antipsychotic use was associated with a 25% decreased risk for psychotic relapse (hazard ratio [HR], 0.75), she reported. Long-acting injectables (LAIs) were associated with the lowest risk (LAI formulations of aripiprazole HR, 0.27; olanzapine HR, 0.28; and risperidone HR, 0.52). Among oral medications, clozapine and aripiprazole were associated with the best outcomes (HR, 0.55, and HR, 0.64, respectively).

For risk of hospitalization due to any substance use disorder, results were quite similar, with any antipsychotic use being associated with similarly reduced risk (HR, 0.78), and similar results for clozapine (HR, 0.27) and LAI formulations of olanzapine (HR, 0.39), aripiprazole (HR, 0.42), and paliperidone (HR, 0.46), she said.

Whether cannabis use problems precede or co-occur with psychosis may also be irrelevant on the basis of findings of another recent Swedish cohort study, reported Solja Niemelä, MD, PhD, professor of psychiatry at the University of Turku, Turku, Finland.

In the study, 1820 patients with FEP (84.7% men; mean age, 26.8 years) were followed for a mean of approximately 6 years. Antipsychotic use over this period resulted in a 33% reduction of psychosis relapse (HR, 0.67), a 23% risk reduction of psychiatric hospitalization (HR, 0.76), and a 23% reduction of hospitalization for substance use disorder (HR, 0.76).

In terms of relapse reduction, LAI olanzapine “wasn’t that effective” (HR, 0.68), she said. But clozapine was (HR, 0.43), as were other long-acting second-generation injectables, other than olanzapine (LAI aripiprazole and paliperidone, HRs, 0.45 and 0.43, respectively). This breakdown was similar for decreased risk of hospitalization.

The Cannabis Link

In the current classification of CIP, “too much emphasis is placed on the cannabis use,” said Rognli. “And this categorization of substance use psychosis as a type of substance use disorder, not together with other psychotic disorders, should be questioned.”

“The concept of substance-induced psychosis not being central in the psychosis continuum may lead to underuse of treatments and poor prognosis,” added Taipale.

Lena Palaniyappan, MD, professor of psychiatry at McGill University and researcher at the Douglas Research Centre in Montreal, Quebec, Canada, agreed that using the word “induced” places disproportionate emphasis on cannabis.

“The word evoke has less causal connotation to it, but it has a temporal meaning, in that cannabis was given, or taken, and it resulted in an episode of psychosis," he told Medscape Medical News. “So, we still keep cannabis in our consciousness w hen we talk about this; we still know that this person needs an intervention for cannabis, but we’re not blaming all of their psychosis on cannabis.”

Palaniyappan recently published research in JAMA Psychiatry that shows the clearest connection to date between cannabis use and psychosis.

The longitudinal observational cohort study included 61 young patients (mean age, 22 years; 81% men), all of whom were taking antipsychotic medication as part of an early intervention psychosis program. Within the group, 25 had a diagnosis of cannabis use disorder (CUD), and 36 had no CUD.

Using MRI to assess neuromelanin, a marker for dopamine levels in the brain, the study showed that CUD was independently associated with higher neuromelanin levels.

“It occurred in the same area of the midbrain where previous studies have shown any increase in dopamine is linked with psychotic symptom severity,” Palaniyappan explained. “So, we were able to draw a line between cannabis, dopamine, and psychosis, which wasn’t possible in the past. At the end of the day, the final common pathway for these two is the same.”

Traveling on the Same Road

Finding the common pathway softens the diagnostic distinctions that have been criticized by Rognli and her colleagues.

“What it says about cannabis-related psychosis is clearly that the mechanisms of what we think of in schizophrenia and cannabis-related psychosis mechanisms travel on the same road. They belong to the same category, and they’re not two distinct phenomena,” said Palaniyappan.

But he stops short of suggesting that cannabis should be completely exonerated.

“What we can say now to young people and their families is we can help reduce the dopamine levels with antipsychotic medications, but the evidence is pretty clear: If people continue to smoke weed in a large amount, it increases their risk of relapse,” Palaniyappan said. “The outcomes are really bad in people who continue to smoke the same amount of weed over a longer period of time after developing an episode of psychosis, whether you call that psychosis schizophrenia or cannabis-induced or cannabis-evoked.”

If a patient experiencing an episode of psychosis needs antipsychotics or early clinical intervention, the treatment decision is the same regardless of whether the psychosis was cannabis induced, he said.

“But the reason for the causal tagging that we do is because there is also another very important intervention to do in these people, which is asking them to be abstinent. If you don’t, if you’re shy of calling an episode cannabis-related, then you’re losing the opportunity,” Palaniyappan said.

“If you’re not differentiating schizophrenia from cannabis-related problems, then you actually treat everything as schizophrenia, which should include some advice on drug addiction and detoxification, but not as much emphasis as with a condition like cannabis-related psychosis.” 

Rognli and Niemelä reported no conflicts of interest. Taipale reported lecture fees from Gedeon Richter, Janssen, Lundbeck, and Otsuka, and a research collaboration with Janssen. Palaniyappan reported receiving personal fees from Janssen Canada, Otsuka Canada, SPMM Course Limited, and the Canadian Psychiatric Association; grants from Sunovion, Janssen Canada, and Otsuka Canada; and book royalties from Oxford University Press.

Kate Johnson is a Montreal-based journalist with more than 30 years of experience writing about all areas of medicine.