A large international meta-analysis has found that people who use cannabis develop psychosis an average of two and a half years earlier than people who do not use the drug. The difference becomes substantially larger among those whose illness emerges later in life: in studies involving participants who developed psychosis at age 30 or older, cannabis users experienced their first episode more than six years earlier than non-users. The findings, led by researchers at UNSW Sydney and published in Biological Psychiatry, add to growing evidence that cannabis exposure may influence not only whether psychosis develops, but also when it begins.
Psychosis is a clinical syndrome in which a person loses contact with shared reality. It can include delusions, hallucinations, disorganised thought, unusual behaviour and severe disturbances in perception or reasoning. Psychosis occurs in schizophrenia and several other psychiatric disorders, and usually first appears during late adolescence or early adulthood, although its onset can occur across a much wider age range. The researchers analysed data from 149 studies involving more than 71,000 people, most of whom had schizophrenia or related psychotic disorders. The studies were conducted primarily in Europe, North America and Australia and were published from 2011 onward.
The analysis identified a clear age-related pattern. In studies where psychosis began at an average age of 20 to 24, cannabis users experienced their first psychotic episode approximately 1.6 years earlier than non-users. Among studies in which the average age of onset was between 25 and 29, the gap widened to 4.43 years. In studies involving people whose psychosis began at age 30 or older, cannabis users developed symptoms an average of 6.3 years earlier. Across the full dataset, the average difference was approximately two and a half years, suggesting that the relationship between cannabis exposure and psychosis onset cannot be understood solely as a risk confined to adolescence.
Matthew Large, a psychiatrist and Conjoint Professor with the UNSW School of Clinical Medicine, said the results challenge the idea that cannabis-related psychiatric risks disappear once a person reaches their twenties. The biological explanation remains uncertain, but one possibility is that repeated exposure to delta-9-tetrahydrocannabinol, or THC, gradually alters brain systems involved in motivation, perception, learning and the regulation of salience—the process by which the brain decides which internal or external events deserve attention. THC acts primarily through cannabinoid receptors, particularly CB1 receptors, which are distributed widely throughout the brain and interact with dopamine, glutamate and other neurotransmitter systems implicated in psychosis.
The researchers emphasise that their findings do not prove that cannabis causes psychosis or directly advances its onset. A randomised experiment assigning people to prolonged cannabis use would be unethical, leaving scientists dependent on observational evidence. Such studies can be affected by confounding factors, including genetic vulnerability, childhood trauma, socioeconomic disadvantage, use of other substances, sleep disruption and differences in access to mental-health care. People who are already developing subtle psychiatric symptoms may also use cannabis in an attempt to relieve distress, raise questions about causality and create the possibility of reverse causation. Even so, the consistency of the association across many studies, together with the stronger effect observed among older participants, supports the hypothesis that cumulative cannabis exposure may contribute to the timing of psychosis in at least some vulnerable individuals.
The absence of a statistically significant difference among teenagers should not be interpreted as evidence that cannabis is harmless during adolescence. Psychosis is relatively uncommon before the mid-teens, which creates a limited opportunity for cannabis use to shift the age of onset further downward. The brain is also undergoing major structural and functional changes during this period, and powerful genetic or developmental risk factors may dominate the timing of illness. In statistical terms, a biological lower limit may make it difficult to detect an additional age shift in the youngest groups. The data therefore suggest a possible floor effect rather than a lack of risk.
The new study updates earlier work by the same research group. A 2011 meta-analysis found that psychosis began about 2.7 years earlier in cannabis users, while a 2015 analysis estimated that psychosis typically occurred 6.3 years after cannabis use began. The larger dataset now allows researchers to compare different stages of life and examine whether the association changes with age. Carly Stevens, the study’s lead author from the UNSW School of Biomedical Sciences, said the findings challenge the assumption that postponing cannabis use until the mid-twenties eliminates the danger. The analysis indicates that warnings focused exclusively on adolescent use may fail to communicate the potential risks faced by adults.
The practical consequences of an earlier psychotic episode can be profound even when the difference is measured in months rather than years. The onset of psychosis may interrupt education, employment, travel, relationships and the development of social support. Treatment commonly involves antipsychotic medication, which can reduce hallucinations and delusions but may also produce adverse effects such as weight gain, sedation, movement disorders and metabolic changes. Earlier illness can also expose a person to a longer period of functional impairment and increase the risk of repeated episodes, especially when treatment is delayed or cannabis use continues. Conversely, reducing or stopping cannabis use may remove one potentially modifiable factor and create more time for people to reach educational, occupational and social milestones before serious illness takes hold.
The researchers say public-health communication must also account for changes in the cannabis market. Medical cannabis has expanded rapidly, and its therapeutic framing may lead some users to underestimate the psychiatric risks associated with THC. High-potency products and edibles can deliver large doses, sometimes with delayed effects that encourage people to consume more before the initial dose has taken effect. The concentration and composition of cannabis products vary widely, making exposure difficult to compare across studies. The researchers are calling for standardised measures that record frequency of use, duration, age at first use, THC concentration, cannabidiol content and the amount consumed. Such information would help determine whether risk rises in proportion to cumulative THC exposure or depends more strongly on particular patterns of use.
Legalisation would not remove the potential for harm, but the authors argue that regulation could provide tools for reducing it. Governments could impose potency limits, require accurate product labelling, restrict marketing, fund education and use taxation to support treatment and prevention. Clear warnings about psychosis would be particularly important for people with a personal or family history of psychotic disorders, mood disorders or other indicators of vulnerability. The researchers also want more studies of older cannabis users, a population that may face increasing exposure as medical cannabis and concentrated products become more common. Their central message is not that every cannabis user will develop psychosis, but that the association is strong enough to warrant serious attention: for some people, repeated cannabis exposure may bring a life-altering illness forward by years.
Subject of Research: People
Article Title: Cannabis Use and Age-Related Acceleration of Psychosis Onset: A Meta-Analysis
News Publication Date: 18-Aug-2026
Web References: UNSW Sydney, https://www.unsw.edu.au/newsroom/ ; Psychosis Australia, https://psychosisaustralia.com.au/
References: Biological Psychiatry, DOI: 10.1016/j.biopsych.2026.06.025
Keywords: Cannabis, psychosis, schizophrenia, THC, mental health, age of onset, psychiatric disorders, meta-analysis, public health, substance use
Tags: age of psychosis onset and substance usecannabis and schizophrenia risk factorscannabis consumption and mental health outcomescannabis use and early psychosis onsetcannabis-related psychosis onset delayearly intervention in psychosis among cannabis usersepidemiology of psychosis and substance useimpact of cannabis on mental healthinfluence of cannabis on age of psychosis diagnosisinternational meta-analysis of psychosis developmentlong-term effects of cannabis on psychiatric disordersneuropsychiatric effects of cannabis
