Daily use of high-potency cannabis (10%+ THC) was linked to nearly 5 times higher odds of first-episode psychosis, and the effect varied dramatically by city.
Anyone wondering about the connection between strong weed and psychosis, particularly in the context of legalization and potency regulation debates.
~5xHigher odds of first-episode psychosis with daily high-potency (10%+ THC) cannabis use
The Backstory
In London, you can buy cannabis that regularly exceeds 20% THC from a dealer down the street. In Bologna, the hash available is closer to 5% THC. Both cities have cannabis users. But London has substantially more people developing psychosis for the first time.
Marta Di Forti wanted to know if that difference was a coincidence.
Her team recruited patients experiencing first-episode psychosis across eleven cities in six European countries and Brazil, alongside population controls from the same cities. What emerged was the largest study ever to connect cannabis potency at the market level with psychosis incidence at the population level. The numbers were stark enough to shift policy conversations across Europe.
The EU-GEI Study
The European Network of National Schizophrenia Networks Studying Gene-Environment Interactions (EU-GEI) was designed to answer a question that single-city studies couldn't: does the type of cannabis available in a city affect how many people in that city develop psychosis?
Process
Study Design: From Cities to Causation
Eleven sites across Europe and Brazil
London, Amsterdam, Paris, Barcelona, Madrid, Bologna, Palermo, Gouda, Ribeirao Preto (Brazil), Puy-de-Dome (France), and GAP (South London). Each site had different cannabis markets with different dominant products.
901 first-episode psychosis cases
Patients aged 18-64 presenting to psychiatric services with their first psychotic episode. Assessed for cannabis use history, frequency, and type of product used.
1,237 population controls
Representative of the local populations at each site. Same cannabis use assessments as cases.
Potency classification from market data
Using national and European data on expected THC concentrations, cannabis types were classified as low potency (<10% THC) or high potency (10%+ THC). This was ecological — estimated from what was available, not measured in what participants actually smoked.
Di Forti et al. (2019), Lancet Psychiatry 6(5):427-436
The Numbers
Psychosis Risk by Cannabis Use Pattern
Odds of First-Episode Psychosis
5x
Daily high-potency use
vs never-users (OR 4.8, 95% CI 2.5-6.3)
3x
Daily any-potency use
vs never-users (OR 3.2, 95% CI 2.2-4.1)
~1
Less-than-weekly use
no significantly increased risk
0
Never users
reference group
Di Forti et al. (2019), Lancet Psychiatry
Daily use of high-potency cannabis (10%+ THC) was associated with nearly five times the odds of experiencing a first psychotic episode compared to people who had never used cannabis. Daily use of any potency carried about three times the odds. Less-than-weekly use carried no significantly increased risk.
But the finding that distinguished this study from everything before it was the city-level analysis.
The Geographic Pattern
Amsterdam
50.3%If high-potency cannabis were unavailable, an estimated half of new psychosis cases could theoretically be prevented
London
30.3%Nearly a third of new cases attributable to high-potency daily use
Paris
~19%Lower-potency hash market, lower attributable fraction
All 11 sites
12.2%Average across Europe — roughly 1 in 8 cases
Bologna/Palermo
<5%Lower-potency markets, lowest attributable fractions
Di Forti et al. (2019), Lancet Psychiatry
Amsterdam — where high-potency cannabis has dominated the coffeeshop market for decades — had the highest estimated attributable fraction: 50.3%. If high-potency cannabis were no longer available, the researchers estimated that half of new psychosis cases in the city could be prevented. London followed at 30.3%. Cities with lower-potency markets had lower attributable fractions.
The correlation between prevalence of high-potency cannabis use among controls and the incidence rate of psychosis across the eleven sites was striking (r = 0.7, p = 0.03). Cities where more people used potent cannabis had more psychosis.
r = 0.7
correlation between prevalence of daily high-potency cannabis use in a city's population and that city's incidence rate of psychotic disorder — across eleven sites in Europe and Brazil.
This ecological correlation doesn't prove causation, but it shows something individual-level studies cannot: the cannabis market itself — specifically what potency is available — appears to shape population-level psychosis rates. This is a public health finding, not just a clinical one.
Di Forti et al. (2019), Lancet Psychiatry
The Causal Debate
The study calculated population attributable fractions (PAFs) — a measure that assumes a causal relationship between exposure and outcome. This assumption drew immediate criticism.
Evidence for a Causal Role
moderate-to-strong- Dose-response: daily use riskier than weekly, high-potency riskier than low — consistent with a causal agent
- Biological plausibility: THC activates CB1 receptors in dopaminergic circuits linked to psychosis
- Temporal sequence: cannabis use precedes psychosis onset in most cases
- City-level ecological correlation: more potent markets = more psychosis incidence
- Consistent with earlier Di Forti (2012) study showing AKT1 gene interaction with daily cannabis use
- Mendelian randomization studies show some support for cannabis use disorder increasing schizophrenia risk
Challenges to Causation
moderate- Case-control design cannot establish temporal causation — people developing psychosis may increase cannabis use in the prodromal phase (reverse causation)
- Shared genetic liability: genes predisposing to schizophrenia also predispose to cannabis use (Pasman GWAS, 2018)
- Potency was estimated from market data, not measured in what participants actually consumed
- Population attributable fractions assume causality — the paper's critics argue this is circular
- Most daily high-potency users never develop psychosis — individual genetic vulnerability matters more than cannabis itself
- Mendelian randomization results are mixed — some analyses find bidirectional effects, others find no clear causation
Di Forti et al. (2019); Gage & Hickman (2019), Lancet Psychiatry correspondence; Pasman et al. (2018)
Suzi Gage and Matthew Hickman published a letter in The Lancet Psychiatry titled "High-potency cannabis and incident psychosis: correcting the causal assumption." They argued that calculating how many psychosis cases could be "prevented" by removing high-potency cannabis presupposes the very causal link the study was trying to establish.
“The authors present population attributable fractions, but these assume a causal effect that has not been established. The fundamental question — whether cannabis causes psychosis or shared genetic and environmental factors explain the association — remains open.”
— Suzi Gage & Matthew Hickman
University of Liverpool / University of Bristol
Letter to The Lancet Psychiatry criticizing the causal framing of the Di Forti study
Di Forti's team replied thoughtfully: they acknowledged that psychosis has complex, multifactorial origins and that they do not believe cannabis is the sole cause. But they argued that PAFs are standard epidemiological tools used for other risk factors (smoking and lung cancer, for example), and that waiting for perfect causal proof before acting on a clear dose-response signal would be irresponsible public health.
The Genetic Vulnerability Layer
This study's findings make more sense when combined with Di Forti's earlier genetic work. In 2012, her team showed that individuals carrying specific variants of the AKT1 gene — involved in dopamine signaling in the striatum — had up to 7-fold increased risk of psychosis when using cannabis daily.
Gene-Environment Interaction
Why Most Cannabis Users Never Develop Psychosis
Base rate
Roughly 3% of the population will experience a psychotic episode in their lifetime. Cannabis doesn't change this dramatically for most users.
AKT1 vulnerability
Carriers of the AKT1 rs2494732 C/C genotype showed up to 7x increased psychosis risk with daily cannabis use. This genotype is carried by roughly 25% of the population.
The interaction
It's not cannabis alone or genes alone — it's the combination. An AKT1 risk carrier who uses cannabis daily has a dramatically different risk profile than a non-carrier using the same amount.
Clinical reality
AKT1 testing isn't clinically available. You can't know your genotype. The practical advice remains: daily high-potency use carries real risk, and people with family history of psychosis should be especially cautious.
Di Forti et al. (2012), Biol Psychiatry; Di Forti et al. (2019), Lancet Psychiatry
This gene-environment interaction reframes the question from "does cannabis cause psychosis?" to "does cannabis cause psychosis in genetically susceptible people who use high-potency products daily?" The answer to that second, more precise question is almost certainly yes — though proving it definitively remains difficult.
What This Means for Policy
The study landed in the middle of European cannabis legalization debates and provided ammunition for a specific policy approach: potency regulation.
Myth vs. Reality
Cannabis causes schizophrenia.
Cannabis — specifically daily use of high-potency products — is a risk factor for psychosis in vulnerable individuals, not a sufficient cause of schizophrenia. Most people who use cannabis daily, even high-potency products, never develop psychosis. But for the subset with genetic vulnerability (including AKT1 risk variants and family history), the risk multiplication is real and clinically significant. The molecule matters: THC increases psychosis risk while CBD may be protective.
The Evidence
Di Forti EU-GEI study: 5x odds with daily high-potency use. But absolute risk remains low — the vast majority of daily users never develop psychosis. AKT1 gene-environment interaction shows up to 7x risk in carriers. CBD has shown antipsychotic properties in clinical trials (Leweke 2012, McGuire 2018).
Di Forti et al. (2019); Di Forti et al. (2012); Leweke et al. (2012); McGuire et al. (2018)
If the relationship is causal — and the weight of evidence increasingly supports at least a partial causal role — then THC potency caps in legal markets could meaningfully reduce psychosis incidence. Some European countries have implemented or discussed exactly this. The study also strengthens the argument for requiring CBD content alongside THC in commercial products, since CBD appears to have antipsychotic properties that may counterbalance THC's psychotogenic effects.
For individuals, the practical implications are clear: daily use of high-potency cannabis carries a meaningful risk of psychosis, especially for those with family history of psychotic disorders. Less-than-weekly use carries no significantly increased risk. Potency and frequency are the variables that matter most.
Related Research
Cannabis, Psychosis, and the Brain
The Di Forti study sits within a broader literature connecting cannabis to psychotic disorders — a relationship that involves genetics, dose, potency, and timing.
THC and CBD have opposite effects on the brain
Bhattacharyya et al. (2010)
fMRI evidence that CBD counteracts THC's effects on psychosis-relevant brain regions — the pharmacological basis for why THC:CBD ratio matters
CBD as an antipsychotic
Leweke et al. (2012)
CBD matched a standard antipsychotic in reducing psychotic symptoms — the therapeutic flip side of the THC-psychosis story
The Dunedin IQ Study
Meier et al. (2012)
Another landmark study showing adolescent-specific vulnerability to cannabis — cognitive effects parallel the psychosis age-dependence
THC potency has been rising for decades
ElSohly et al. (2016)
Documents the potency escalation that makes the Di Forti findings increasingly relevant — today's products are even stronger than what was studied
Cannabis GWAS: genes predict use
Pasman et al. (2018)
Shared genetic architecture between cannabis use and schizophrenia — the confounding that makes causation so hard to establish
Does cannabis cause schizophrenia?
Not in a simple sense. Cannabis — specifically daily use of high-potency THC products — is a risk factor that increases the odds of psychosis, especially in genetically vulnerable individuals. Most daily users never develop psychosis. But for those with risk factors (AKT1 gene variants, family history of psychotic disorders), the risk multiplication is significant. The relationship is almost certainly partly causal and partly driven by shared genetic vulnerability. The precise contribution of each remains debated.
What does 5x risk actually mean for me?
A five-fold increase in odds sounds dramatic, but context matters. If the baseline risk of developing psychosis is roughly 3% over a lifetime, a 5x increase raises that to roughly 15% — significant but still meaning 85% of daily high-potency users would never develop psychosis. The risk is concentrated in those with genetic vulnerability. If you have no family history of psychosis and no personal prodromal symptoms, your absolute risk remains relatively low even with daily use. If you do have family history, the risk calculation changes substantially.
Would lower-potency cannabis be safer?
The data strongly suggest yes. The study found that daily use of any-potency cannabis carried about 3x odds, while daily use of high-potency cannabis carried about 5x odds. Less-than-weekly use of any potency carried no significantly increased risk. The dose-response pattern suggests that both frequency and potency matter, and that reducing either would reduce risk. This is the scientific basis for THC potency cap policies being discussed in several countries.
Does CBD protect against cannabis-induced psychosis?
Probably, at least partially. Clinical trials have shown CBD has antipsychotic properties (Leweke 2012, McGuire 2018), and fMRI studies show CBD counteracts THC's effects on psychosis-relevant brain regions (Bhattacharyya 2010). The dominance of high-THC, zero-CBD products in modern markets removes this natural pharmacological buffer. Whole-plant cannabis containing both THC and CBD may carry less psychosis risk than THC-dominant products — though this hasn't been directly tested in a psychosis-prevention trial.
What should I do if I have a family history of psychosis?
The evidence strongly suggests caution. Family history of psychotic disorders is one of the strongest known risk factors for cannabis-associated psychosis. If you have a first-degree relative (parent, sibling) with schizophrenia or another psychotic disorder, daily cannabis use — especially high-potency products — carries a meaningfully elevated risk. This doesn't mean any cannabis use will trigger psychosis, but the risk-benefit calculation is very different for you than for someone without family history. Talk to a healthcare provider, and if you do use cannabis, low-potency, CBD-containing products used infrequently represent the lowest-risk approach.
What the researchers found
This multicentre study across 11 sites in Europe and Brazil compared 901 people experiencing first-episode psychosis with 1,237 population controls. The pattern was consistent: daily cannabis use was associated with 3.2 times higher odds of psychotic disorder compared to never-users. For daily users of high-potency cannabis (THC above 10%), the odds jumped to nearly 5 times higher.
But the most striking finding was geographic. In cities where high-potency cannabis dominated the market — London and Amsterdam — the population attributable fractions were highest. The researchers estimated that if high-potency cannabis were no longer available, 30% of new psychosis cases in London and 50% in Amsterdam could theoretically be prevented. In cities where lower-potency products were more common, the estimated impact was much smaller.
This didn't prove causation, but it showed something beyond individual risk: cannabis potency at the market level appeared to shape psychosis incidence at the population level.
Why it matters
Previous studies linked cannabis to psychosis risk at the individual level. This study scaled that question up to entire cities and found that the local cannabis market — specifically how potent the available product was — appeared to influence how many people developed psychotic disorders. That's a different kind of finding. It suggests that potency regulation at the policy level could be a public health lever.
The London and Amsterdam numbers were particularly striking because those cities had the highest-potency cannabis markets AND the highest psychosis incidence in the study. The correlation was ecological, not proof, but the consistency across sites made it harder to dismiss.
The numbers in context
• Daily cannabis use: 3.2x higher odds of psychotic disorder vs never-use
• Daily high-potency use (10%+ THC): ~5x higher odds
• London: 30% of new psychosis cases attributable to high-potency cannabis (estimated)
• Amsterdam: 50% attributable (estimated)
• 901 first-episode psychosis cases, 1,237 controls across 11 sites
How the study worked
Multicentre case-control study (EU-GEI project) across 11 sites in 6 European countries and Brazil. Recruited patients aged 18-64 presenting with first-episode psychosis and population-representative controls. Used adjusted logistic regression models. Cannabis potency was classified using national and Europe-wide data on expected THC concentrations: low potency (<10% THC) vs high potency (10%+ THC). Calculated population attributable fractions assuming causality.
Who was studied
N=901 patients aged 18-64 with first-episode psychosis and 1237 controls from 11 sites across Europe and Brazil.
What this study cannot tell us
Case-control design cannot prove causation. People predisposed to psychosis may be drawn to cannabis use (reverse causation). Self-reported cannabis use is unreliable, and potency was estimated from market data rather than measured in what participants actually consumed. Population attributable fractions assume a causal relationship that hasn't been proven. Cultural and diagnostic differences across 11 sites may affect comparability.
How to read the evidence
Large multicentre case-control study published in The Lancet Psychiatry. Strong methodology but cannot prove causation due to study design.
When this study was published
Published in 2019. Remains one of the most influential studies in the cannabis-psychosis literature.
The bigger picture
Published in The Lancet Psychiatry, this became one of the most cited and debated papers in the cannabis-psychosis literature. It landed in the middle of European legalization discussions and strengthened the argument for THC potency caps in regulated markets. The finding that population-level psychosis rates tracked with market-level potency — not just individual use patterns — shifted the conversation from personal risk to public health policy.
Replication
Not stated in abstract.
Funding
Medical Research Council, the European Community's Seventh Framework Program grant, São Paulo Research Foundation, National Institute for Health Research (NIHR) Biomedical Research Centre (BRC) at South London and Maudsley NHS Foundation Trust and King's College London and the NIHR BRC at University College London, Wellcome Trust.
Conflicts of interest
Not reported in abstract.
Questions still open
- Would THC potency caps in legal markets actually reduce psychosis incidence?
- Is there a THC concentration threshold below which psychosis risk doesn't increase?
- Do people who develop psychosis have a pre-existing vulnerability that cannabis triggers, or can cannabis cause psychosis in anyone at high enough doses?
Common questions
Does strong weed cause psychosis?
Does the type of weed matter for psychosis risk?
Read the original research
The contribution of cannabis use to variation in the incidence of psychotic disorder across Europe (EU-GEI): a multicentre case-control study
The Lancet Psychiatry, 6(5), 427-436
The Lancet Psychiatry is a well-respected journal focusing on psychiatric research and clinical practice.
Citation
Di Forti, Marta; Quattrone, Diego; Freeman, Tom P.; Tripoli, Giada; Gayer-Anderson, Charlotte; Quigley, Harriet; et al.. (2019). The contribution of cannabis use to variation in the incidence of psychotic disorder across Europe (EU-GEI): a multicentre case-control study. The Lancet Psychiatry, 6(5), 427-436.
Explore the wider topic
- How THC Affects Your Amygdala: The Brain's Threat Detector and Cannabis
- The Anandamide Connection: Your Body's Natural Bliss Molecule
- How Long for Cannabinoid Receptors to Return to Normal
- Cannabis and the Developing Brain: What Every Teenager (and Parent) Should Know
- Why Can't I Enjoy Anything Without Weed? The Science Behind It
- Dopamine Recovery After Quitting Weed: What the Science Says
- The Endocannabinoid System Explained Simply: What It Does and Why It Matters
- Your Endocannabinoid System Explained: Why Withdrawal Happens
- Your Nervous System After Quitting Weed: Fight or Flight
- Using Weed Under 18: What It Does to Your Developing Brain
- What THC Does to Your Brain: Why Withdrawal Happens
- THC and Your Prefrontal Cortex: What Cannabis Does to Your Decision-Making Brain
- Weed, Cortisol, and Stress: What Cannabis Does to Your Stress Hormones
- Weed and Memory: What the Science Says About THC and Your Hippocampus
- Weed and Motivation: Is Amotivational Syndrome Real?
- Weed and Your Nervous System: What THC Actually Does to Your Brain and Body
- How Weed Rewires Your Reward System (And How to Reset It)