A leading psychosis researcher has laid out some of the clearest evidence yet on the high potency cannabis psychosis link. Daily use sharply raises the odds of a psychotic disorder. The pattern holds even once genetic vulnerability, geography and time are all factored in.

Dr Marta Di Forti delivered the findings in a keynote at the 2024 ADAI Symposium on Cannabis, Schizophrenia and Other Psychotic Disorders, held on 19 September that year. Co-moderators Kristen Haley, a public health education liaison at the Washington State Liquor and Cannabis Board, and Beatriz Carlini introduced the session. Di Forti thanked them, mentioned she was still recovering from a bout of laryngitis, and joked that her cat had joined the call because British pet shops now sell cannabis-infused treats for animals too.

Who Dr Di Forti Is

Di Forti is a clinical reader in psychosis research at Kings College London. She also leads the UK’s first Cannabis Clinic for people with psychotic disorders, which opened in 2020. An MRC senior research fellowship funds her work expanding the study of cannabis in psychosis and its underlying biology.

Her research draws on virtual reality, genetic and epigenetic data, and peripheral measures of endocannabinoids from human participants. It also uses animal models of adolescent cannabinoid exposure, all aimed at finding biological markers of vulnerability among heavy cannabis users. She trained in medicine in Palermo, Sicily, before moving to London, where she practises as a psychiatrist and leads this research.

The Scale Of The Risk

Di Forti opened with the existing meta-analyses of cannabis and psychosis research. They converge on a similar figure, which she called unusually consistent for this field: people who have ever used cannabis in their lifetime carry roughly double the risk of later developing a psychotic disorder.

A simple yes-or-no question about lifetime use tells clinicians very little, she argued. It’s the same problem as asking a room whether anyone has ever drunk alcohol. Almost everyone puts a hand up, and that question reveals nothing about individual risk of liver disease. So researchers began measuring the amount and frequency of cannabis use instead, and the picture changed sharply.

A continuous measure of exposure produced an overall estimate of a fourfold increase in psychosis risk, not a doubling. London provided the highest odds ratio of any site in that analysis. Di Forti linked this to the type of cannabis available in her own catchment area.

What Cannabis-Associated Psychosis Looks Like

Di Forti described the experience in some detail, drawing on a cartoon her own patients helped create. The onset is usually a subtle, pervasive shift in how someone understands the world around them. Streets begin to feel threatening. People believe others are watching them with hostile intent. Some patients describe the television reporting on their own lives; others come to believe neighbours have planted microchips under the floorboards. Auditory hallucinations are common too: many patients hear voices nobody else can hear, alongside a sense that the brain itself is “on fire”.

These experiences are intrusive and disabling. They cut people off from relationships and the outside world. Di Forti stressed that psychosis is a multifactorial condition, and cannabis is never the only risk factor for any individual.

But when researchers compared symptom profiles, one group stood out. People who had used daily high potency cannabis before their first episode were far more likely to show positive symptoms, including persecutory delusions and hallucinations. This high potency cannabis psychosis pattern held consistently across the clinical picture Di Forti described.

Why Frequency And Potency Matter

A large European collaboration spanning London, Amsterdam and Paris recruited people presenting for the first time with a psychotic disorder between 2011 and 2013. Researchers also recruited a general population comparison group and took blood samples for later genetic analysis. Across all three cities, one pattern stood out: daily use of cannabis with a THC concentration of ten per cent or higher carried the greatest risk. London’s odds ratio for this group was 5.16. Amsterdam’s was 9.43. Paris came in at 4.29, each measured against people who had never used cannabis.

This finding underpins the high THC cannabis risk that Di Forti wants clinicians to explain more plainly to young people. Even people using cannabis more than once a week at lower potency showed a real rise in risk. She put this down to cumulative THC exposure building up over time, whether from frequency, strength, or both together.

A Fourfold Gap Between Cities

The same study found at least a fourfold variation in the incidence of psychotic disorder between sites, once researchers accounted for age, gender and minority status. Di Forti noted, with some humour, that her own home city of Palermo had one of the lowest incidence rates in Europe. She joked that moving to London hadn’t been the smartest career choice, given her own risk profile as a woman entering the age range for a second peak in psychosis onset. Researchers plotted the proportion of the general population using high potency cannabis against the incidence rate of psychotic disorder over time, and the two lines rose and fell together.

Evidence From Portugal And Denmark

Di Forti also pointed to two country-level natural experiments. Portugal decriminalised cannabis possession in 2001 and began redirecting people caught with cannabis towards psychological support rather than arrest. Hospital admissions for psychotic disorder associated with cannabis use tell a striking story: they rose from 20 cases in 2000 to 588 cases in 2015. Di Forti admitted she was initially sceptical of that scale of increase, until other countries echoed it.

Denmark’s population registry links birth, school, health and general practice records over a person’s lifetime, and it showed a similar pattern. Schizophrenia incidence rose sharply through the 1990s and again in the early 2000s. Cannabis use rose over the same period, alongside rising availability of high potency strains. Researchers calculated a population attributable risk fraction, adjusting for other contributing factors, and traced a growing share of new schizophrenia cases over that period back to cannabis use.

The High Potency Cannabis Psychosis Link And Genetics

Critics have raised a persistent challenge to this research. Perhaps people with a higher genetic predisposition to schizophrenia simply choose high potency cannabis more often. If so, cannabis would mark risk rather than cause it.

Di Forti’s data doesn’t support that explanation. Her team calculated a polygenic risk score for schizophrenia from participants’ DNA and found no evidence that people with a higher score were more likely to use high potency cannabis in the first place. The analysis then adjusted for that genetic score, plus age, gender, employment, education and other drug use. Daily users of high potency cannabis still showed an almost fivefold increase in risk.

Genetic load still matters, and risk climbs for everyone as polygenic risk increases. But daily high potency cannabis psychosis risk remained the highest curve on the graph, even among people with comparatively low genetic liability. Di Forti said this points to heavy cannabis exposure pushing some people towards psychosis on its own; a strong genetic starting point isn’t required.

Weighing The Genetic Overlap

Di Forti added nuance rather than claiming the genetics question is fully settled. Large genetic studies show that liability to cannabis use disorder and liability to schizophrenia overlap, and researchers still debate what that overlap means. Some Mendelian randomisation studies probe causal direction using genetic data; some point one way, some the other, and some suggest the relationship runs both ways at once.

Di Forti referenced a commentary by researchers Gillespie and Kendler on this debate. They set out three possibilities: cannabis use could be entirely causal, partially causal and partially explained by shared genetic or family factors, or entirely non-causal. Di Forti’s own reading, echoed by the Portugal and Denmark data, favours the middle option. Genetics dominates for some people, environment does for others, and many fall somewhere in between.

The Baking Analogy

To make this accessible to a lay audience, Di Forti’s team produced a short explainer video using a baking analogy. Genetics play the role of the cake mixture, shaping the raw ingredients a person starts with. Social background acts as the cake tin, shaping those ingredients further. Baking temperature and time stand in for the context in which someone uses cannabis, so even identical genetics and identical ingredients can produce different outcomes under different conditions. Cannabis itself is the icing. Icing comes in many forms, and cannabis varies enormously in strength; frequent use of potent strains is what tips the balance towards harm.

Cannabis Use After Diagnosis

Di Forti was clear that continuing cannabis use after a first psychotic episode carries serious costs. She said this part of the evidence base attracts far less controversy than the causation debate. Patients who continue to use cannabis relapse faster and more often, including people who others had told that cannabis might relax them or ease their anxiety.

The pattern held true whether or not patients took their prescribed antipsychotic medication as directed. That ruled out poor medication adherence as the explanation: cannabis users who took their medication religiously still relapsed more, and faster, than non-users who didn’t.

She shared figures from her own trust in south London, where cannabis use and psychotic disorder frequently coexist. Among all psychotic patients under the trust’s care, 60 per cent also had a history of cannabis use. That figure rose to 80 per cent among patients under intensive psychiatric care. Among patients also known to forensic mental health services, it reached 96 per cent.

Di Forti set this against a wider backdrop of shifting cannabis policy in her own city. She referenced, with a wry aside, London’s mayor piloting a cannabis legalisation scheme in her own catchment area after a visit to California. She also made a passing joke about Elon Musk reportedly considering growing cannabis in space because Earth is running out of room for plantations.

The Cannabis Clinic And Its Results

In response, Di Forti and colleagues set up a dedicated service for young adults with psychosis to help patients change their cannabis use. Charity money funded the service initially, rather than NHS resources, since new services typically need proof of impact before health system funding follows. It brings together clinicians experienced in both addiction and psychosis, rather than inventing a new model from scratch.

Patients receive at least 20 individual sessions over several weeks. Alongside this, a weekly peer support group runs online every Tuesday. Young people in the community, or on inpatient wards, hear from international experts and from each other there, including fellow keynote speaker Professor Cerato Deusa, whom Di Forti credited as a regular contributor to the group.

The team measures cannabis use disorder severity, mood, anxiety and functioning before and after the intervention. Among the first 46 people to complete it, 73 per cent had stopped cannabis entirely by the end, and most of the rest had cut down to once a week or less. Cannabis use disorder scores fell significantly, as did measures of persecutory delusions, paranoia, anxiety and depression.

Di Forti called the anxiety improvement notable in itself, since it contradicts the common assumption that cannabis calms people with psychosis. Functioning improved sharply too: most patients moved from doing very little day to day at the start of treatment to returning to work or study by the end. Patients also save a considerable amount of money by no longer buying cannabis, Di Forti added, though she framed this as a secondary benefit rather than the point of the service.

In A Patient’s Own Words

Di Forti closed her results section with quotes from a patient who had recently completed the programme. Asked what helped him reduce his use, he credited cutting down with building his confidence. Getting emotional support when he was struggling had mattered most, particularly while he was still working. He said the team never pressured him with a fixed timeline, which gave him space to work out his own pace of change. He also valued not feeling judged, or told that what he was doing was good or bad; making that decision himself mattered to him.

Di Forti then shared a short recorded exchange with another patient, referred to as Marcus, who had also worked with the clinic. Asked what had been most beneficial, Marcus said that simply being listened to mattered most. He’d always felt able to talk about his drug use with the team without feeling judged, and sharing his mind, his feelings and his thoughts freely felt central to his progress.

What Patients Say Cannabis Gives Them

During the question and answer session, chaired by Kristen Haley, someone asked Di Forti what people with psychosis get from continuing to use cannabis, and what might replace it. She described two components. The high itself can dull the distress of hallucinations and delusions during intoxication, though symptoms often flare up more intensely once it wears off, pulling people into a cycle of escalating use. The second component is social: cannabis can remain a link to friends, and a shared context a young person is reluctant to give up, particularly where a partner or peer group is still using.

On substitutes, Di Forti was cautious about one-size-fits-all advice. She described patients who gradually replaced cannabis with exercise or with music. Others took up structured activities, such as a monthly breakfast club her service runs, which she framed as small proof points that a morning without cannabis is achievable. On medication, she noted there’s no consensus treatment for cannabis craving. Heavy dopamine-blocking antipsychotics can sometimes worsen craving and drive further use, while partial dopamine agonists may support reduction better by allowing some normal dopamine function to continue.

Someone also asked about policy and prevention. Di Forti noted the UK currently has no public education campaign on cannabis harms, unlike the messaging that already exists for alcohol and tobacco. She suggested non-judgemental, youth-facing content would likely work better than clinicians lecturing young people directly. Content featuring musicians or public figures who’ve spoken about their own experiences could open a conversation into which the evidence gets introduced, much as public health messaging has worked for alcohol, tobacco and healthy eating over time.

Frequently Asked Questions On Cannabis And Psychosis

Do genetic studies disprove the cannabis psychosis link? No. Vulnerability to psychosis and vulnerability to problematic cannabis use appear to partly overlap genetically. That overlap doesn’t erase the observed association between cannabis exposure, particularly daily high THC exposure, and psychosis.

Does genetic risk explain the association away? Not in the data Di Forti presented. The association between daily high potency cannabis use and psychotic disorder remained after researchers adjusted for polygenic liability to schizophrenia and other relevant factors.

If genetics and cannabis use overlap, is cannabis harmless for people with no family history of psychosis? No. Genetic liability is only one part of the picture. High frequency, high potency cannabis use can raise risk even among people with lower measured genetic liability.

Can genes tell us exactly who will develop psychosis from cannabis? No. Current polygenic scores aren’t deterministic tests. They capture only part of genetic liability and can’t reliably predict an individual outcome on their own. Age of first use, trauma, social conditions, other substance use, and cannabis dose and potency all matter too.

Is the relationship causal? The overall evidence supports a contributory causal role for cannabis in at least some psychosis cases, especially with intensive high THC use. The relationship isn’t identical for everyone, and genetics, reverse causation and shared vulnerabilities remain active research questions.

The Bottom Line

Di Forti’s overall message is that psychosis risk works as an interaction between vulnerability and exposure. Daily high potency cannabis psychosis risk stands out as one particularly significant and preventable exposure. Stopping, even after years of heavy use, offers a genuine route to reducing risk and supporting recovery, and the evidence suggests it works best the earlier it happens.

Kristen Haley closed the session by thanking Di Forti. She noted that the full slides and recording would go online within about a week, and invited the audience’s applause for the presentation.

Source: youtube