People who use cannabis are developing psychosis years earlier than non-users, new research shows, prompting calls for greater health warnings about one of the nation’s most widely used substances.
Cannabis users develop psychosis an average of two-and-a-half years earlier than non-users, a new meta-analysis from UNSW Sydney researchers shows.
This earlier onset of psychosis becomes more pronounced with age, stretching to more than six years difference between cannabis users and non-users who develop psychosis in their 30s, the global analysis published in Biological Psychiatry found.
Psychosis is a symptom of schizophrenia and other mental illnesses, and can involve delusions, hallucinations, disordered thinking and behaviour. It typically emerges in late adolescence or early adulthood but can occur across a wide age span. Up to 3% of adults experience a diagnosed disorder involving psychosis in their lifetimes, according to Psychosis Australia.
Researchers from UNSW Medicine examined 149 studies involving more than 71,000 people, largely conducted in Europe, North America and Australia since 2011. The majority focused on people with schizophrenia.
In the studies where psychosis developed at an average age of 20 to 24 years, researchers found that cannabis users experienced their first episode an average of 1.6 years earlier than non-users.
The difference increased to 4.43 years in studies where psychosis developed in the mid-to-late 20s (25-29), and to 6.3 years among those aged 30 and over.
Chief investigator and psychiatrist Matthew Large, a Conjoint Professor with UNSW’s School of Clinical Medicine, said the findings showed cannabis use was linked to the earlier onset of psychosis, and its effect appeared to build over time.
“We’ve tended to focus on the dangers of using cannabis in adolescence, when the brain is still developing, but the risk doesn’t disappear once you’re in your 20s,” Prof. Large said. “What we’re seeing suggests a cumulative effect, where harm builds up over time, a bit like alcohol or nicotine.”
The researchers found no significant difference between the age of psychosis onset among teenagers who used cannabis and non-users but noted its rarity before the mid-teens left a limited window for earlier onset.
“There may be a biological floor, in that there is only so much earlier that onset can be shifted. In younger groups, strong development, genetic or other psychosis risk factors may dominate, making any cannabis-related age shift harder to detect,” Prof. Large said.
He noted the study did not prove that cannabis caused psychosis or brought it forward, as to do so would require experimental research that would be unethical.
“But the consistency of the evidence, particularly the strong effect in older adults who likely have been using cannabis for longer, does support the theory that cannabis plays a causal role in the timing of psychosis onset, and possibly the onset of some cases among people who would not have otherwise developed psychosis.”
A closer look at the research
The study follows a smaller 2011 meta-analysis led by Prof. Large which found psychosis began 2.7 years earlier for cannabis users, and a 2015 analysis which found psychosis typically occurred 6.3 years after cannabis use began. There was insufficient data available at the time for age group comparison.
UNSW lead author Carly Stevens, from the School of Biomedical Sciences, said being able to compare life stages revealed a clearer picture of the link between cannabis and psychosis.
“What we’ve found challenges the assumption that you’ll be fine if you wait to smoke weed until your mid 20s, that’s just not the case,” Ms Stevens said.
“There’s been so much emphasis on adolescents not smoking cannabis, which is important, but we need to broaden the warnings for everyone.”
A couple of years can change a life
Delaying psychosis by a couple of years may not sound like much, but the shift can make a lifetime of difference, Ms Stevens said.
“People may miss out on study, jobs opportunities, travel and relationships in those years, and treatment can come with physical side effects that can make life harder as well,” she said.
Prof. Large said reducing or stopping cannabis use had clear benefits.
“It’s a major illness, which robs you of your shared reality with family, friends and all those around you,” he said. “If we can delay psychosis by even a few years people will be more likely to reach crucial milestones in education and training, and develop strong support networks, which can protect against psychosocial decline.”
Further, it would reduce the burden on the healthcare system, given up to 3% of health-related expenditure in developed nations is attributed to psychotic disorders, he said.
Warnings need to expand
While the research does not prove causation, the researchers said the evidence supported the need for stronger public health warning on the potential psychiatric consequences of cannabis use.
“We can’t run direct experiments, so we have to go with the best evidence available,” Prof. Large said. “In my view, the threshold for health warnings about cannabis and psychosis has definitely been met.”
Complicating matters further is the rapid rise of medical cannabis, which Prof. Large said led many people to underestimate the risks.
He noted a growing number of people he treated for psychosis had been prescribed cannabis for conditions where evidence of benefit was limited.
“It’s got this therapeutic halo around it, which blinds people to the risk.”
Also of concern is the rise of cannabis edibles, which can have higher concentration of Tetrahydrocannabinol (THC), the primary psychoactive compound in cannabis.
While the legalisation of cannabis would make access easier, it could help reduce harm by enabling governments to control potency, educate consumers and require informed purchasing.
“You could tax it, regulate it, and require people to show they understand the risks before buying,” Prof. Large said.
More research needed
The researchers say we need to further investigate the potential biological mechanisms underlying cumulative cannabis exposure and psychosis onset.
They also call for more work examining older cannabis users, who may face increasing risks as medical cannabis and high‑potency products become more widely used.
Future studies also need clearer, standardised metrics for measuring cannabis exposure.
DOI link to research: https://doi.org/10.1016/j.biopsych.2026.06.025
Contact details:
Lachlan Gilbert
UNSW News & Content
t: +61 2 9065 5241
e: [email protected]
