The largest meta-analysis to date highlights inconsistent benefits, frequent adverse effects and widespread bias across psychiatric and substance use trials.
There is little high-quality evidence supporting cannabinoids for most mental disorders and substance use conditions despite widespread prescribing, new research suggests.
Cannabinoid medicines are approved for the treatment of mental disorders and substance use disorders (SUDs) in Australia, the US and Canada, yet data on their efficacy and safety in treating these conditions remains limited.
Over a quarter of the population aged 16–65 years in the US and Canada reported ever using cannabis for medical purposes, half of which were for mental health management.
In Australia, where these medications have been legalised more recently, over one million applications have been approved for their use, and mental disorders account for six of the top ten most common indications for their prescription.
A team of researchers from Australia and the UK have conducted the largest and most comprehensive systematic review and meta-analysis of randomised controlled trials (RCTs) of cannabinoids for the treatment for these disorders.
They identified large gaps in evidence for their use, as well as major issues of potential bias.
Overall, 54 studies were included: 12 for cannabis use disorder, eight for psychotic disorders, six for anxiety disorders, five for tic or Tourette’s syndrome, four for opioid use disorder, four for insomnia, three for cocaine use disorder, three for post-traumatic stress disorder (PTSD), two for anorexia nervosa, two for autism spectrum disorder and two for obsessive compulsive disorder (OCD). There was one study each for attention-deficit/hyperactivity disorder (ADHD), bipolar disorder and tobacco use disorder.
Despite depression being one of the leading indications for cannabinoid use, no trials which examined its efficacy were identified.
For cannabis use disorder, there was a moderate reduction in withdrawal symptoms associated with cannabinoids (standardised mean difference [SMD] –0.70), although this was no longer statistically significant after excluding high-risk-of-bias studies. When stratified by formulation, only combined CBD plus THC (nabiximols) demonstrated a small benefit (SMD –0.29), while CBD or THC alone showed no effect. Cannabinoids also reduced overall cannabis use but had no significant impact on craving, cannabis-related problems, or abstinence. There was no difference in adverse effects compared with placebo.
For autism spectrum disorder, there was a small improvement in autistic traits overall (SMD –0.36), although this was not consistent when different cannabinoid formulations were analysed separately. Adverse effects were similar to placebo.
For cocaine use disorder, cannabinoids were associated with a worsening of craving (SMD 0.69) and an increased risk of adverse events (OR 3.76) compared with placebo.
Cannabinoids did not significantly improve overall insomnia symptoms. Although there was some evidence of increased sleep duration, findings were inconsistent, and adverse events were significantly more common (OR 11.04).
Cannabinoids significantly reduced tic severity in Tourette’s syndrome and other tic disorders (SMD –0.62), but this effect was limited to combined CBD plus THC formulations. However, adverse events were significantly more frequent (OR 4.93), while withdrawal rates were similar between groups.
Cannabinoids showed no meaningful effect on symptoms of schizophrenia and other psychotic disorders, anxiety disorders, opioid use disorder, OCD, ADHD, bipolar disorder, tobacco use disorder, PTSD or anorexia nervosa, including no significant effects on weight or physical activity.
Related
Most studies were parallel group design (n=32), with an average treatment duration of around five weeks, a median sample size of 31 and a median age of 33 years. Participants across the trials were mostly male (70.7%).
Cannabidiol was the most common medication (n=24), followed by THC (n=18) and a combination of THC and cannabidiol (n=12).
Overall, 44% of the studies were identified as at high risk of bias, 37% raised concerns and 18% were considered low risk, based on stated conflicts of interest, author industry involvement and financial relationships, and unclear roles of sponsors in study design and reporting.
“There was some evidence that cannabinoids can reduce symptoms of cannabis use disorder, insomnia, tic or Tourette’s syndrome, and autism spectrum disorder, but the quality of this evidence was generally low,” authors wrote.
“Given the scarcity of evidence, the routine use of cannabinoids for the treatment of mental disorders and SUDs is currently rarely justified.”
