Cannabis is the most widely used controlled substance in the world, with more than 244 million people estimated to use it. As more countries change their cannabis laws, a key question for public health is whether the way people access cannabis changes its effect on their mental health.
A new study from Switzerland, published in the journal Drug and Alcohol Dependency, followed 374 adults who could buy regulated cannabis from pharmacies for a year. It looked at what happened to their symptoms of depression and anxiety, and to how much and how often they used cannabis.
This article summarises the findings and what they do, and do not, tell us. It is about non-medical cannabis accessed through a Swiss pilot scheme, not about cannabis-based medicines prescribed in the UK. It is general information, not medical advice.
Key Takeaways:
- The study followed 374 adults in Basel who already used cannabis regularly and who could buy quality-controlled cannabis from pharmacies over 12 months.
- Average scores for depression, anxiety and problematic cannabis use fell slightly over the year. The changes were statistically significant but small.
- The proportion of people above clinical thresholds for depression, anxiety or possible cannabis use disorder did not change significantly.
- People used cannabis on slightly more days per month by the end of the year, but the amount per day did not change.
- People at high risk of cannabis use disorder at the start showed the largest fall in problematic use. Even so, most were still in the high-risk group after a year.
- The study had no comparison group, so it cannot show that regulated access caused these changes. The authors state the results are not an endorsement of cannabis as a treatment for depression or anxiety.
What do we know about cannabis and mental health?
The relationship between cannabis and mental health is complex. Research has linked cannabis use, especially frequent use of high-strength products, to a higher risk of depression, anxiety, psychosis and cannabis use disorder. Around one in ten people who use cannabis are expected to develop cannabis use disorder. Among weekly users, it is around one in five.
Most of this research comes from countries where cannabis is illegal. In those settings, product strength, contamination and the stress of buying illegally are all mixed in with any effect of cannabis itself. That makes it hard to separate the drug from the way it is used.
What about medical cannabis and mental health?
Cannabis-based medicines are a separate question from recreational use. They are prescribed products with a known cannabinoid content, taken under medical supervision. The evidence for their use in mental health conditions is still limited, and it is important to be clear about what it shows.
Two systematic reviews in The Lancet Psychiatry have pooled the randomised controlled trials in this area. The first, by Black and colleagues in 2019, found scarce evidence that cannabinoids improve depression, anxiety, PTSD or psychosis. It found very low-quality evidence of a small improvement in anxiety symptoms, mostly in people taking THC-containing medicines for other conditions such as chronic pain.
The most recent, by Wilson and colleagues in 2026, reviewed 54 trials and found no significant effects on anxiety, PTSD or psychotic disorders, and no randomised trials at all in depression. It did find some low-quality evidence of benefit in cannabis use disorder, insomnia and tic disorders. In both reviews, cannabinoids were linked to more side effects than placebo, though not to more serious adverse events. The 2026 authors concluded that routine use of cannabinoids for mental health conditions is “currently rarely justified” on the evidence available.
Real-world data from the UK add a different type of evidence. Data from the UK Medical Cannabis Registry, which collects outcomes from patients prescribed cannabis-based medicines, were published in a two-year case series in the Journal of Affective Disorders in 2026. Among 698 patients prescribed these medicines for depression, average scores for depression, anxiety, sleep quality and quality of life improved from baseline at every follow-up point up to 24 months, with the largest changes in the first three months. Around 9% of patients reported at least one side effect, most of which were mild or moderate.
As with the Swiss study, this was an observational study with no comparison group, so it cannot show that the medicines caused the improvement. The authors describe the findings as support for future clinical trials rather than proof of effectiveness.
Taken together, the trial evidence for cannabis-based medicines in mental health is limited.. Observational data suggest that some patients report improvements, but controlled trials are needed to test this properly. Established treatments such as talking therapies and prescribed antidepressants remain first-line.
What is the Swiss Weed Care study?
Switzerland does not have a legal recreational cannabis market. Instead, it is running a series of scientific pilot trials in different cities to test what regulated access does to cannabis use and health. The Weed Care study in Basel-Stadt is one of these pilots.
Under the Basel model:
- Cannabis was sold only through pharmacies by trained pharmacists.
- Products had a maximum THC content of 20% and a monthly limit of 10 g of THC per person.
- Prices were set to match the illegal market, so cost did not drive behaviour.
- At every purchase, pharmacists gave counselling on lower-risk use and could refer people to specialist services.
- Taking part was not exclusive. Participants could still buy cannabis illegally. Over the year, about half (51%) of the cannabis they used came from pharmacies and half (49%) from the illegal market.
There was no commercial marketing, and no psychological or medical treatment was given as part of the scheme.
Who took part?
Participants were adults living in Basel-Stadt who had used cannabis at least monthly for the previous six months and tested positive for metabolites of tetrahydrocannabinol (THC). People with acute suicidal thoughts or psychosis, a current psychiatric admission, severe cognitive impairment, or who were pregnant or breastfeeding were excluded.
The 374 participants had an average age of 35.8 years, and 81% were men. At the start, a third (33.7%) scored in the high-risk range for cannabis use disorder on a validated screening questionnaire (the CUDIT-R). Around 9.5% dropped out before the one-year follow-up.
How was the study carried out?
The Weed Care study was originally designed as a randomised trial. Half of the participants received pharmacy access straight away, and the other half after six months. This analysis covers the full year, when everyone had access. It is therefore best described as a before-and-after study within the same group of people. The original allocation was adjusted for in the statistics. That design matters when reading the results, as explained below.
Participants completed online questionnaires at the start, at six months and at 12 months. The questionnaires measured:
- Depressive symptoms using the PHQ-9
- Anxiety symptoms using the GAD-7
- Psychotic symptoms using an adapted early-recognition checklist
- Problematic cannabis use using the CUDIT-R
- How often they used cannabis (days in the past month) and how much (grams per day)
On the PHQ-9 and GAD-7, a score of 10 or more is the usual threshold for moderate-to-severe symptoms. On the CUDIT-R, a score of 13 or more suggests possible cannabis use disorder.
What were the results?
Depression and anxiety scores fell slightly
Across the whole group, the average PHQ-9 depression score fell from 5.37 to 4.53. The average GAD-7 anxiety score fell from 3.81 to 3.20. Both changes were statistically significant, but they were small. The starting scores were already in the mild range.
The share of people above the clinical threshold did not change significantly. Moderate-to-severe depressive symptoms fell from 12.7% to 8.9% and moderate-to-severe anxiety from 5.3% to 4.4%, but neither change reached statistical significance. Psychotic symptoms did not change.
Problematic cannabis use fell, mainly in the high-risk group
The average CUDIT-R score fell from 11.02 to 10.25. The proportion of people scoring in the possible cannabis use disorder range fell from 32.7% to 28.6%, which was not statistically significant.
The pattern differed by starting risk. People in the high-risk group at the start saw their average CUDIT-R score fall from 16.1 to 13.9. That is larger than the two-point change usually considered meaningful. People in the low-risk group did not change. Even so, 87% of those who started in the high-risk group were still in it after a year. Their average score stayed above the threshold for possible cannabis use disorder.
Use became slightly more frequent, but not heavier
People used cannabis on more days per month by the end of the year: 18.7 rising to 20.1. This was a small but statistically significant increase. The amount used per day (around 1.2 g) did not change.
Time explained very little of the difference between people
One technical finding is worth highlighting. Time accounted for less than 1% of the variation in symptom scores. Most of the variation (between 57% and 70%) came from stable differences between individuals. In plain terms, the year of regulated access made only a small average difference, and people started and stayed at very different levels.
Adjusting for any antidepressants, sedatives, antipsychotics, stimulants or cannabis counselling that participants were also receiving did not change these results.
What do the findings mean?
The authors describe the central finding as a divergence. People used cannabis slightly more often, yet their scores for problematic use, depression and anxiety went down rather than up. This runs against the common assumption that more frequent use always means more harm.
Their suggested explanation is that the regulated model itself may have offset some of the usual risks of frequent use. Known THC content, consistent quality, less contact with the illegal market and regular conversations with a pharmacist could all have played a part. They conclude that how access is designed matters more than legalisation itself.
It is important to be clear about what the study does not show. In the authors’ own words, the findings “should not be read as an endorsement of cannabis use as a treatment for depressive or anxiety symptoms”. The improvements were modest, their cause is uncertain, and participants remained exposed to the known harms of cannabis, including dependence and effects on memory and thinking. The authors state that established treatments for depression and anxiety, meaning psychological therapy and prescribed medicines, remain the appropriate first-line approach.
What are the limitations of this study?
The authors set out several limitations:
- No control group. Everyone eventually had pharmacy access, so the study compares people with themselves over time. Improvements could reflect the natural ups and downs of symptoms. They could also reflect “regression to the mean”, where people who score high at the start tend to score lower next time, or simply the effect of taking part in a study.
- Low starting scores. Average depression and anxiety scores were mild to begin with, which limits how much they could fall and makes small changes harder to interpret.
- Half of the cannabis came from the illegal market. Exposure to the regulated model was partial, so the results cannot be attributed to the pharmacy scheme alone.
- Self-reported data. Cannabis use and symptoms were reported by participants, which can be affected by memory and by wanting to give a favourable answer.
- Vulnerable groups were excluded. People with severe mental illness were not eligible, and the average age of 35.8 years is past the period of highest risk for cannabis-related psychosis. The study cannot say much about psychosis risk in younger or more vulnerable people.
- Only one year of follow-up. Rare or slow-developing harms would not be captured.
- Psychological therapy was not recorded, so it remains a possible alternative explanation for the improvements.
Why does this matter in the UK?
Recreational cannabis remains illegal in the UK. This study describes a tightly controlled public-health pilot in one Swiss city. Its findings cannot simply be transferred to other countries or to other ways of regulating cannabis.
The study is nonetheless relevant to a wider debate. Its findings suggest that the conditions under which people use cannabis may influence the risks they face. Product quality, potency limits and contact with a healthcare professional all appear to matter. That is relevant wherever cannabis is used, whether legally or not. Longer and controlled studies are needed before firmer conclusions can be drawn.
Frequently asked questions
Does regulated cannabis access improve mental health?
This study found small average improvements in depression and anxiety scores over a year of pharmacy-based access in Switzerland. The proportion of people with clinically significant symptoms did not change. There was no comparison group, so the study cannot show that regulated access caused the improvement.
Did people use more cannabis once it was available from pharmacies?
Participants used cannabis on slightly more days per month (18.7 rising to 20.1), but the amount they used per day did not change. About half of their cannabis still came from the illegal market.
Is recreational cannabis legal in Switzerland?
No. Switzerland has not legalised recreational cannabis. Regulated sales are only permitted within approved scientific pilot trials, such as the Weed Care study in Basel. These trials are designed to test the effects of different access models.
Is recreational cannabis legal in the UK?
No. In the UK, possession, cultivation and supply of cannabis for non-medical use are criminal offences. Cannabis-based medicines are a separate category and can only be prescribed by a specialist doctor for eligible patients.
Is this study about medical cannabis?
No. The participants were adults who already used cannabis recreationally and were not being treated for a medical condition. The study did not test cannabis as a treatment for depression, anxiety or any other condition. Its authors say the results should not be read that way. The evidence for cannabis-based medicines in mental health is still developing and is a separate question. You can read about how our specialists approach anxiety, depression and post-traumatic stress disorder on our condition pages.
Where was the research published?
The study was published in Drug and Alcohol Dependence in 2026 by Pichler and colleagues from Psychiatric Services Aargau and the University Psychiatric Clinics Basel.
Looking after your mental health
If you are struggling with low mood, anxiety or your cannabis use, speak to your GP or a qualified healthcare professional. Talking therapies and prescribed medicines are the established first-line treatments for depression and anxiety. Free, confidential support for drug use is available through the NHS and local drug and alcohol services. In an emergency, or if you feel unable to keep yourself safe, contact your local emergency services or a crisis helpline.
Curaleaf Clinic is a specialist clinic. To find out whether a consultation with one of our specialist doctors may be appropriate for your circumstances, you can complete our online eligibility check. Any treatment decisions are made by a specialist on an individual basis after a full assessment.
