Study finds psychosis onset gap widens with age
A global analysis of 149 studies involving more than 71,000 people has found that cannabis use is linked to earlier-onset psychosis, but the effect is not uniform across age groups. The research, led by University of New South Wales (UNSW) scientists, challenges the long-held assumption that the developing brains of people under 25 are most at risk.
The meta-analysis, published in the journal Biological Psychiatry on 18 August 2026, found that among people who developed psychosis in their 20s, cannabis users did so nearly 4.5 years earlier on average than non-users. For those who developed psychosis after age 30, the gap widened to around 6.5 years.
However, the study found no evidence of earlier onset in cannabis users under 20. Overall, cannabis users tended to develop psychosis about 2.5 years earlier than non-users, a finding consistent with previous research.
Building on prior evidence
The work builds on an analysis published in JAMA 15 years ago, incorporating five times more data than was available at that time. Despite the much larger dataset, the researchers found a strikingly similar average. "Despite this massive increase in sample size, we found a strikingly similar result: just like in 2011, we found that on average, cannabis users tended to develop psychosis earlier by about two and a half years," said lead author Carly Stevens, a PhD candidate at the UNSW School of Biomedical Sciences and trainee doctor.
Stevens noted that the consistency of the finding strengthens confidence in the association. "At this point, we can pretty confidently say this is a robust association," she said, adding that the age-related variation was the more revealing aspect. "Because of the increase in data, we were able to see if this association varied across different life stages, and we found that it did, but not in the way that we thought it would."
She explained that the initial assumption was that effects would be concentrated in young people, given the vulnerability of the adolescent brain and experimental evidence from animals. "We found no evidence that cannabis users under 20 develop psychosis earlier," Stevens said.
Why younger users may not show a gap
Researchers suggest a "floor effect" may explain the lack of association in under-20s, meaning psychosis is already likely to be early-onset in that age group. However, they stressed that this does not indicate cannabis is safe for teenagers.
Co-author Matthew Large, a UNSW Conjoint Professor and medical superintendent of Mental Health in the Eastern Suburbs Mental Health Service, called cannabis a "long-term cumulative neurotoxin". He argued that the findings make a case against the medicalisation of cannabis, which he described as "one of the biggest public health bungles in Australian history".
Large, who was also a co-author of the 2011 paper, noted that psychotic disorders account for about 3% of total healthcare expenditure in developed nations.
Medical cannabis concerns
The researchers also pointed to the widespread prescribing of medical cannabis. About 30% of medical cannabis prescriptions are for anxiety, they said, despite little evidence that it works. In Australia, cannabis is regulated and must be prescribed by a registered healthcare professional, with guidelines that it cannot be used as a first-line treatment. Most medicinal cannabis products are classed as unapproved therapeutic goods under the Therapeutic Goods Administration, meaning they have not been assessed for safety, quality or efficacy.
In Australia, an estimated 2.5 million people use cannabis each year. The study's findings are "concerning given the ease of access to increasingly potent cannabis products in Australia," said Jack Wilson, a research fellow at The University of Sydney's Matilda Centre, who was not involved in the research.
Limitations
One limitation of the study, the authors acknowledged, was that it did not examine the type of cannabis products used or their potency.
This article was based on reporting from Medical Republic.