more_vert
The number of Australians using medicinal cannabis has surged over the past five years. Around 700,000 Australians have used cannabis for their health in the past year.
And since 2022, medicinal cannabis sales have increased four-fold. But the majority of products prescribed in the Australian market are not registered with the Therapeutic Goods Administration. This means they have not been rigorously tested.
So, is medicinal cannabis safe? And is it actually effective?
Our new research, published today in Lancet Psychiatry, is the largest ever review to look at the safety and effectiveness of medicinal cannabis for mental health and substance use disorders.
These make up six of the top ten reasons cannabis is prescribed, specifically: anxiety, sleep disorders, post-traumatic stress disorder (PTSD), insomnia, depression, and attention deficit and hyperactivity disorder (ADHD).
But we found little evidence medicinal cannabis effectively treats these conditions. And while most side effects were mild to moderate, some serious questions about safety remain.
What evidence was available?
Between 1980 and 2025, we found 54 randomised controlled trials that looked at whether medicinal cannabis reduced or treated mental health disorders (including psychotic disorders, anxiety, insomnia, anorexia and PTSD) or substance use disorders (including cannabis, cocaine and opioids). This kind of trial is the gold standard for understanding the direct effects of a medicine.
The most common cannabinoid being evaluated was cannabidiol (CBD), followed by tetrahydrocannabinol (THC) and a combination of THC and CBD.
CBD is non-intoxicating and typically safe, whereas THC is psychoactive, and linked to harms in the short term, such as paranoia, and longer term, such as the development of a cannabis use disorder.
Cannabis may help people quit cannabis
We found cannabis medicines were no more effective than a placebo at treating symptoms of psychotic disorders (such as schizophrenia), anxiety, PTSD, anorexia or opioid use disorder.
However, there were promising findings that medicinal cannabis may be effective in reducing cannabis use among those with a cannabis use disorder.
While this may sound strange, the medicines largely consisted of an oil-based combination of CBD and THC that was taken orally. As these medicines reduce craving, patients may use less of their usual cannabis. So for people who regularly smoke high-THC cannabis, using medicinal cannabis instead may reduce their risk of related health problems, such as lung conditions.
But there are limitations
We must be careful when interpreting the positive findings.
For example, some evidence suggested medicinal cannabis could help treat symptoms associated with tic or Tourette’s syndrome, insomnia, and autism spectrum disorder. But only a small number of studies focused on these conditions and many were low quality.
In randomised controlled trials, we don’t want participants to know whether they are consuming the medicine or placebo. But as cannabis is often intoxicating, participants may be aware of what they have been given, and this can introduce bias.
Some of these studies also reported conflicts of interest, which may have influenced their results. So it may be too soon to tell whether medicinal cannabis is effective in treating these conditions.
How about safety?
The combined data showed cannabis medicines were linked to mild side effects such as nausea, dry mouth and fatigue.
But the risk of serious side effects, such as a psychotic episode, was no greater among those taking cannabis medicines or a placebo.
The data alone seems to suggest cannabis medicines are relatively safe. But this may not be reflected in real-world use.
The average length of treatment in these studies was only five weeks – and we know regular cannabis use can cause long-term harms.
One 2024 review found around one-quarter of those using medicinal cannabis will develop a cannabis use disorder. This is similar to the rate among those using cannabis for non-medical use.
The cannabis medicines used in these studies were also low in THC. But data from the TGA shows Australians have access to a wide range of cannabis medicines that are often high in THC content. Chronic consumption of high THC cannabis has been linked to a greater risk of worsening mental health symptoms, particularly among young people.
So what does this mean?
Similar reviews have been conducted in the past. But many focus on a smaller number of health conditions and have not combined data to reveal a single estimate.
Reviews have also commonly drawn conclusions on cannabis as a treatment for mental health when it was being used to treat other conditions, such as chronic pain.
Nevertheless, our findings are largely consistent with previous reviews: there is little evidence medicinal cannabis is an effective treatment for mental health and substance use disorders.
Currently, there is a mismatch between the research evidence on medicinal cannabis – mainly short-term trials and CBD formulations – and real-world use, which is longer term and often using high-THC products.
We need more research on cannabis medicines, particularly for conditions with limited alternative treatments, and monitoring over longer periods.
As the TGA conducts a review of medical cannabis prescribing in Australia, these findings should inform future regulation. The long-term use of these medicines could result in harm and delay the use of more effective treatments.
The takeaway
For those who believe their medicinal cannabis is beneficial for these conditions, our review does not mean to contradict your experience.
However, we encourage you to regularly discuss your circumstances with a doctor, and if possible, consider alternative evidence-based treatments.
Jack Wilson is a Postdoctoral Research Fellow at the Matilda Centre for Research in Mental Health and Substance Use, University of Sydney.
Emily Stockings is an Associate Professor of Medicine, University of Sydney.
This article is republished from The Conversation under a Creative Commons license. Read the original article.
The statements or opinions expressed in this article reflect the views of the authors and do not necessarily represent the official policy of the AMA, the MJA or InSight+ unless so stated.
Subscribe to the free InSight+ weekly newsletter here. It is available to all readers, not just registered medical practitioners.
If you would like to submit an article for consideration, send a Word version to mjainsight-editor@ampco.com.au.
As Professor Mather has already pointed out, the medicinal use of cannabis is an unusually difficult intervention to evaluate. There are several reasons for this. Firstly, the very large number of psychoactive ingredients makes evaluation very complicated and protracted. Secondly, the lack of standardisation of what patients receives complicates the aggregation of research experience. Thirdly, research into medicinal cannabis was delayed for many decades because of the prohibition of recreational cannabis which continues in many countries. Fourth, conventional research designs often appear to underestimate the benefits which many patients receiving medicinal cannabis report. The low (but not zero) risk of medicinal cannabis should encourage researchers and funders of research to work through these problems. In all likelihood, medicinal cannabis supplied through regulated sources will be safer than when supplied by unregulated sources.
the absence of proof is not the proof of absence.
It’s such an old fashioned dogma – designed to intentionally mislead the unwary.
A more balanced view of the topic would be more helpful.
Medicinal Cannabis undoubtedly has helped many genuine patients to date.
On the other hand , unregulated doses and black market recreational cannabis use / abuse of unquantified doses and strains has been linked to mental health disorders like schizophrenia and more – hence the resistance by the psychiatrists interests groups. That is not contested. However, Cannabis has a legitimate place in Australia for the genuine treatment of various ailments, ..especially where other pharmacologic agents have FAILED ! . It is true that more research will only benefit the sector; but it is an over-reach, soaked in stigma and fear, to simply slam the door shut and say no …. when there is much to be learned and to be gained to improve healthcare. I am personally interested in the research & treatment of frozen shoulder – simply because currently we do not have a valid solution that can be agreed upon. How absurd is that ? Don’t point the finger, accuse & / or be blinded by the lack of RCT & published evidence. Start building the evidence, create the evidence , and solve the problem – it is science.
I understand from the literature that cannabis is not effective at lessening pain intensity but I wonder if the “suffering” components have been sufficiently looked at. Designing trials that prospectively target and measure those dimensions would be a logical next step, and current evidence leaves that as an open and important research gap rather than a settled negative.
Thank you Professor Mather for your valuable comments and I concur with the salient points you raise.
The homogenization that goes into an all-encompassing meta-analysis will likely prevent the fidelity needed to identify the benefits that might accrue to individual patients with individual doses of individual compounds for individual problems in the biopsychosocial milieu of that individual patient.
The military in World War 1 learned the devastating and lethal effect of homogenization of tactics on the battlefield.
Economics learned that lesson in the mid 20th century.
The reproducibility crisis has shown that even the most basis systems are hard to predict.
And various medical specialties have been abandoning the large homogenized clinical trial as lacking the fidelity to address meaningful intervention in a complex clinical situation (for example the NIH in their sepsis research strategy).
But the most important issue is the devastating disenfranchisement of pulling a patient’s lifeline because of overly simplistic (and perhaps even authoritarian) pronouncements of “there’s no evidence……”.
Mass homogenization, demanding statistical surety (at 95%), and failure to concede the severe consequences of false negative trial outcomes, will likely lead to the disempowerment, disenfranchisement and psychological torture of many patients who already struggle with life. Not everyone was born lucky enough to be neurotypical and lead a “normal” life.
The work that went into this project evinced in the supplementary data provided to the original Lancet Psychiatry paper can only be described “Herculean”.
However, the fundamental issues not recognized in this, and the now-frequent similar papers based on meta analyses of “medicinal cannabis” pharmacotherapy, are these.
“Cannabis” is not a drug but a chemical “fruit salad”. While the authors recognize the well-known ingredients THC and CBD, these are but two of the hundreds of components of cannabis that regulate the plant’s growth and sustenance, that differ according to a multitude of variables including strain and growth conditions, in post harvesting storage and processing, and in the pharmaceutical preparation, even if the two most-recognized principal substances THC and CBD are regulated. Moreover, the delivery of a variety of pharmacologically active principals to the cannabinoid receptor system within the recipient depends on the route and timing of administration, and is subject to individual characteristics of drug uptake, distribution and metabolism. And that is apart from considerations of individual dose-response, or more precisely blood concentration-response, relationships (Mather LE, Rauwendaal E, Moxham-Hall V, Wodak A. The issue of medicinal cannabis in contemporary Australia. Griffith J Law 3: 286-313, 2015).
The authors in their “takeaway” concede “For those who believe their medicinal cannabis is beneficial for these conditions, our review does not mean to contradict your experience.” While that is comforting, it is not surprising that when studies and patients are combined in meta analyses, that “no evidence” is found for “this benefit” or “that effect”. Meta-analysis has been the “gold standard” now for some time. However, by combining and amalgamating data from patients with differing personal conditions and experiences, they abrogate their ability to cater for individual patients. In comparison, the “N-of-1 approach” for a particular patient with a particular dosage form” is surely the most reliable manner of assessing whether patients, with their particular state and condition, benefit from the medication, and the dosage form used, the dosing regimen, or whether a different prescription might be better. (Lillie EO, Patay B, Diamant J, Issell B, Topol EJ, Schork NJ. The n-of-1 clinical trial: the ultimate strategy for individualizing medicine? Per Med 8: 161-173, 2011).
On the other hand, the continual presentations of formulaic meta analyses of medicinal cannabis that regularly conclude that “no evidence for this or that effect can be found”, to me, can only be described as “Sisyphean”.