A major new review in The Lancet Psychiatry has delivered one of the clearest verdicts yet on cannabis and psychosis, concluding that regular use is a contributory cause of the illness rather than simply something that happens to occur alongside it.

The review, led by Professor Wayne Hall and colleagues from the University of Queensland, the University of Bath and several other institutions, pulled together evidence from epidemiological, genetic, experimental and animal studies published up to April 2026. Rather than relying on any single type of study, the team looked for what they called coherence: whether different research methods, each vulnerable to different sources of bias, pointed towards the same conclusion.

On cannabis and psychosis, they did.

A threefold rise in risk

A 2024 meta analysis cited in the review found that any measure of recreational cannabis use, from having tried it once to using it daily, was associated with a threefold increase in the incidence of psychosis. The researchers then tested the two most obvious alternative explanations: that people with early psychosis turn to cannabis to ease their own symptoms, and that some hidden factor explains both the drug use and the illness.

Neither explanation held up well. In the largest study to track individuals against themselves over time, cannabis use was a far stronger predictor of later psychotic symptoms, with an odds ratio of 7.03, than psychotic symptoms were of later cannabis use, which came in at just 0.59. That points one way: the drug tends to come first. And when the authors calculated how strong a hidden confounding factor would need to be to fully explain away the link between marijuana and psychosis, they found it would need to raise the risk of both cannabis use and psychosis by almost six times on its own, something they describe as implausible.

There is a dose relationship too. The earlier cannabis use begins, the more often it is used, the higher its THC content and the longer someone uses daily, the greater the risk of psychosis appears to be. In Denmark, a nationwide psychiatric registry found that the proportion of schizophrenia cases attributable to cannabis use disorder has climbed since 2010, which the authors link to rising rates of cannabis use itself.

Young people carry the heaviest load

Cannabis use disorder, the review notes, is far more likely to take hold when use begins in the teenage years and continues into young adulthood, precisely the period when anxiety, depression, psychosis and bipolar disorder typically first appear. Among young people who have ever tried cannabis, 27% go on to develop a cannabis use disorder; among those using it at least weekly, that figure climbs to 33%.

Beyond psychosis: a murkier picture

Away from psychosis, the evidence connecting cannabis and mental health more broadly is far less settled. Cannabis use is linked to poorer outcomes in people who already have bipolar disorder, including a younger age of onset, more frequent mood episodes and less improvement with treatment, though few genetically informed studies exist to rule out shared underlying risk.

For depression, one meta analysis found that cannabis use in adolescence predicted later depression more strongly, with an odds ratio of 1.33, than depression predicted later cannabis use, at just 1.03, a pattern the authors say points towards cannabis contributing to the illness rather than simply following from it. Anxiety findings are mixed: one meta analysis of longitudinal studies found no significant increase in risk, while another restricted to adolescents found more than double the odds of a later anxiety disorder.

The most sobering figures concern suicide. A meta analysis of longitudinal studies found cannabis use in adolescence was associated with a 50% rise in the odds of suicidal thoughts and more than a threefold increase in the odds of a suicide attempt. People diagnosed with a cannabis use disorder had a suicide mortality rate more than three times the expected level. The authors are careful to note these estimates rest on only a handful of studies, and that depression may explain some, though not all, of the relationship between cannabis and suicidal behaviour.

“No universally safe level”

The review points to the Lower Risk Cannabis Use Guidelines as the clearest public health advice currently available, and its opening line is blunt: people who use cannabis need to know that there is no universally safe level of use, and that abstinence is the only way to avoid all risk of harm. The guidelines go on to recommend delaying the start of use until at least late adolescence, avoiding high potency products, and steering clear of cannabis altogether during pregnancy or breastfeeding.

What comes next

The authors are candid about the gaps that remain. Research into cannabis and mental health still lacks large, long running studies capable of tracking rare outcomes such as suicide, and few clinical trials have tested treatments for people facing both a cannabis use disorder and a concurrent mental illness. Early results from integrated treatment programmes, which bring cannabis focused and mental health focused care together within a single team, are promising, the review notes, but remain too small and too few in number to be considered settled practice.

What the review does settle, on the strength of six converging lines of evidence, is the question its authors set out to answer. The link between cannabis and psychosis is not a coincidence of timing, nor simply a case of people using the drug to cope with symptoms that were already there. It is, the evidence increasingly suggests, part of the cause.

Source: dbrecoveryresources