For specialist psychiatrists, GPs in shared-care arrangements, mental-health pharmacists and clinic governance leads: this article reviews what UK clinicians need to understand about the intersection of cannabis-based products for medicinal use (CBPMs) and mental health. It covers the risk literature, the much smaller therapeutic literature, current UK position statements, screening at initiation, monitoring, and what a defensible prescribing posture looks like under conditions of genuine evidence uncertainty. It is written for clinicians; it does not promote any specific product, and it is not clinical advice.
Start with the position statements, not the marketing
Three UK reference points should anchor any mental-health-adjacent CBPM conversation:
- NICE NG144 (2019, minor update 2021) makes no recommendation in favour of CBPMs for any primary mental-health indication. The guideline scope is intractable nausea and vomiting, chronic pain, spasticity and severe treatment-resistant epilepsy. Mental health is outside its remit, and silence is not endorsement.
- Royal College of Psychiatrists Position Statement PS05/19 (“Cannabis-based medicinal products”) states that CBPMs may be of potential benefit in some psychiatric contexts but that there is a paucity of high-quality evidence supporting their use in psychiatric disorders, and recommends caution. In May 2025, in its response to the London Drugs Commission, the RCPsych restated that even small amounts of daily cannabis use can cause significant harm, and that vulnerable people can be up to five times more likely to develop a psychotic disorder.
- The Medical Cannabis Clinicians Society Good Practice Guide was updated in 2025 and reinforces the expectations around training, peer-review of higher-risk prescribing and structured outcome monitoring. It is not a clinical guideline but it is a useful operational framework.
A clinician initiating a CBPM in a patient with significant mental-health history is doing so in a space where no UK clinical guideline recommends the intervention, and the relevant specialist college continues to counsel caution. That is the starting point.
The psychosis risk literature, plainly stated
The strongest signal in the cannabis and mental health evidence base is not therapeutic. It is the association between heavy, high-potency cannabis use and psychotic illness, and it must be addressed directly.
Key reference points:
- Moore et al., Lancet 2007 — a systematic review and meta-analysis reporting an approximately 40% increased risk of any psychotic outcome in individuals who had ever used cannabis, with a dose-response relationship.
- Di Forti et al., EU-GEI study, Lancet Psychiatry 2019 — daily cannabis use was associated with roughly three-fold increased odds of psychotic disorder versus never-use, and daily use of high-potency cannabis with approximately five-fold increased odds. The study estimated that, in London, if high-potency cannabis were unavailable, around 30% of first-episode psychosis cases could be prevented; in Amsterdam, around 50%.
- EU-GEI follow-up analyses (Bond et al., 2024) on cannabis cessation and psychosis risk suggest that risk declines with sustained abstinence, but elevated risk may persist in former high-potency heavy users for prolonged periods.
The mechanistic interpretation remains debated and the causal architecture is not fully established, but the population-level association is robust across designs, cohorts and countries. A clinician who frames cannabis as therapeutically neutral with respect to psychosis is not representing the evidence.
Two practical implications follow:
- A personal or strong family history of a primary psychotic disorder (schizophrenia, schizoaffective disorder, psychotic depression, drug-induced psychosis) is, in current UK practice, a substantive contraindication to initiating a THC-containing CBPM. Some clinicians will extend this to bipolar disorder with psychotic features. Where a clinician proceeds despite such a history, the rationale, the consent discussion and the alternatives considered should be documented in detail.
- The potency of the prescribed product matters. The MCCS 2025 guidance recommends peer-panel oversight for flower prescriptions exceeding 2g/day or above 25% THC. Mental-health-history patients should not, as a default, be initiated at the high end of either axis.
The therapeutic exploration literature: what is, and is not, there
Set against the risk literature, the published evidence for CBPMs as treatment for mental-health conditions is observational, early-stage and predominantly UK registry data.
PTSD. The UK Medical Cannabis Registry has published two case-series analyses on PTSD outcomes. The 2022 analysis of 162 patients reported significant improvements in PTSD symptoms, sleep and anxiety across follow-up. An updated 2025 analysis (Sapphire / Curaleaf cohort, ~269 patients, published in Expert Review of Neurotherapeutics) extended these findings to 18 months, with adverse events reported in around a quarter of patients, most commonly insomnia and fatigue. The signal is consistent and clinically interesting; the design is observational with no comparator arm and meaningful loss to follow-up. It supports a hypothesis, not a guideline-level recommendation.
Generalised anxiety disorder. A 2021 UK Medical Cannabis Registry analysis (67 GAD patients) reported significant improvements in GAD-7, EQ-5D-5L and sleep scores at 1, 3 and 6 months. A 2024 follow-up extended the cohort. To date, only a small number of randomised controlled trials have examined CBPMs in anxiety disorders, and those have largely examined high-dose CBD isolate in social anxiety — pooled, they do not show a significant effect on anxiety as a primary endpoint.
Depression. A 2025 two-year UK Medical Cannabis Registry case series in depression (Journal of Affective Disorders) reported improvements in patient-reported depression scores over time, again without a comparator arm. The implications for treatment-resistant depression specifically remain speculative.
The honest summary: the observational signal is positive across PTSD, anxiety and depression cohorts, the controlled evidence is thin, and the publication picture is dominated by a small number of registry groups. None of this meets the threshold for guideline recommendation, and prescribing in this space remains private, off-licence and observational.
Who should not be initiated, or should be initiated only with extreme care
A practical contraindication and caution list for clinicians evaluating mental-health-history patients:
- Absolute caution: personal history of schizophrenia, schizoaffective disorder, psychotic depression or drug-induced psychosis; active first-episode psychosis or recent recovery; strong first-degree family history of primary psychotic illness.
- Substantial caution: bipolar disorder (particularly bipolar I with psychotic features); current active substance-use disorder, including cannabis use disorder; severe personality disorder with affective instability; pregnancy or breastfeeding.
- Care required: adolescents and young adults (<25), given the developmental psychosis-risk profile; patients on multiple psychotropics with metabolic interactions; patients with cardiovascular disease (THC has cardiovascular effects independent of mental-health considerations).
None of these categories is an automatic bar in every case — the GMC expects individualised decisions within a clinician’s competence — but each warrants a documented rationale and, in most clinic governance frameworks, peer or MDT review before initiation.
Screening at initiation
Before any THC-containing CBPM is initiated in a patient with current or historical mental-health involvement, a defensible baseline assessment includes:
- A structured psychiatric history covering personal and family history of psychosis, bipolar disorder, severe depression and substance-use disorder
- Current medications, including any psychotropics and CYP-relevant agents
- Baseline self-report measures appropriate to the indication: PHQ-9 for depressive symptoms, GAD-7 for anxiety, PCL-5 for PTSD symptoms, HADS where a generic affective screen is useful
- Explicit screening questions for current or recent psychotic experiences (perceptual abnormalities, ideas of reference, paranoid ideation) and a low threshold for psychiatric review where any are positive
- An AUDIT or equivalent alcohol screen and a substance-use history including prior cannabis exposure, frequency, age of first use and any prior adverse psychiatric reaction
- A documented consent discussion that includes psychosis risk, dependence potential, the unlicensed status of the product, the absence of guideline recommendation for mental-health indications, and the alternatives considered
This is consistent with the structured-assessment expectations set out in the UK specialist prescribing pathway and aligns with the MCCS 2025 good-practice framework. Patients moving between EU jurisdictions — relevant where care is coordinated across the European medical cannabis landscape — should have the same screening applied at re-initiation.
Shared care with primary mental-health teams
Most CBPM prescribing in the UK is private and specialist-initiated. Where a patient is concurrently under NHS primary or secondary mental-health care, shared-care considerations are non-trivial:
- The patient’s NHS GP and any community mental-health team should be informed in writing, with the rationale, the product, the dose and the monitoring plan. Most indemnifiers expect this; many GPs will decline to participate further, and that is a legitimate position.
- Where a patient is on existing psychotropics, interactions and the possibility that subjective improvement on a CBPM may prompt unilateral psychotropic discontinuation should be discussed and recorded.
- Escalation pathways for emergent psychotic symptoms — whom the patient contacts, who reviews, on what timescale — should be specified at initiation, not improvised after an incident.
Monitoring after initiation
A reasonable monitoring framework in a mental-health-history patient on a CBPM includes:
- Repeat PHQ-9 / GAD-7 / PCL-5 at 1 month, 3 months, 6 months and at least 6-monthly thereafter
- Direct enquiry at each review about perceptual changes, paranoid ideation, sleep disruption, suicidality and substance-use escalation
- A low threshold to pause or taper if emergent psychotic symptoms, manic switch, marked anxiety worsening or dependence pattern develops
- Documented dose, formulation, batch, and the patient’s adherence and use pattern (frequency, timing, route)
- Periodic review of whether the prescription continues to meet its initial clinical objective, and whether equivalent licensed-medicine options have become accessible
If a patient deteriorates psychiatrically while on a CBPM, the clinical default in current UK practice is to taper or stop the CBPM and to escalate the mental-health management through the appropriate NHS or specialist pathway. Continuing a CBPM through emergent psychosis is not a defensible position.
The defensible posture, summarised
For UK clinicians working at the intersection of CBPMs and mental health, a defensible posture has five components:
- Be honest about the evidence asymmetry. The risk literature is substantial. The therapeutic literature is observational and early-stage. Patients deserve both pictures.
- Do not present CBPMs as a recommended treatment for any primary mental-health condition. No UK guideline currently makes that recommendation. Honest practice acknowledges that.
- Screen properly at initiation. Psychiatric history, family history, substance-use history, validated baseline measures, explicit psychosis screening, documented consent.
- Monitor continuously, with a low threshold to stop. Mental-health prescribing in this space is not a one-time decision; it is a continuing one.
- Communicate with the patient’s wider care team. Private specialist prescribing does not exempt a clinician from the standards of communication and shared-care discipline applied elsewhere in psychiatry.
Done within those constraints, observational CBPM prescribing in carefully selected mental-health-history patients can be a defensible part of UK specialist practice. Done outside them, it is not.
About MUZO Health
MUZO Health was created to raise the standard of cannabis-based medicines in the UK, with a focus on quality, consistency and clinical integrity. Our mission is to support clinicians, empower patients and help move cannabis medicine forward responsibly, transparently and without compromise. Learn more about MUZO Health.