Home > Cannabis: the evidence for medical use.

Barnes, Michael P and Barnes, Jennifer C (2016) Cannabis: the evidence for medical use. London: All-Party Parliamentary Group for Drug Policy Reform.

PDF (Cannabis: the medical evidence)

Cannabis has been used as a medical product for many centuries. In recent decades it has been discovered that the human brain and other organs contain naturally occurring cannabinoid receptors as well chemicals that bind to those receptors. This is called the endocannabinoid system. It is known that the endocannabinoid system has a range of important natural functions, including modulation of pain, control of movement, protection of nerve cells and a role in natural brain adaptability (plasticity), as well as a role in various metabolic, immune and inflammatory processes and a possible role in the control of tumour growth.


Plant cannabis probably works in man by “mimicking” the effects of the human endocannabinoid system. The main plant cannabinoids (phytocannabinoids) studied, and thought to be the most important in terms of efficacy, are tetrahydrocannabinol (THC) and cannabidiol (CBD), although many others exist and a role for them may become clearer in due course. In this paper we have analysed and graded the evidence for efficacy of cannabis and various licenced cannabis products for a number of different indications.


We have found good evidence for one or more of the cannabis products or “natural” cannabis in; the management of chronic pain, including neuropathic pain; spasticity; nausea and vomiting, particularly in the context of chemotherapy; and in the management of anxiety. We have found moderate evidence in; sleep disorders; appetite stimulation in the context of chemotherapy; fibromyalgia; post-traumatic stress disorder; and for some symptoms of Parkinson’s disease. We have found some limited evidence of efficacy, but further studies are required, in; the management of agitation in dementia; epilepsy, particularly drug resistant childhood epilepsies; bladder dysfunction; glaucoma; and in Tourette’s syndrome. We have found that there is a theoretical basis, but so far no convincing evidence of efficacy; for the management of dystonia; Huntington’s disease; headache; brain protection in the context of traumatic brain injury; depression; obsessive compulsive disorder; gastrointestinal disorders; anti-psychotic agent (CBD); and a role in cancer/tumour control.


We have summarised the short term effects of cannabis, which are generally mild and well tolerated. We have looked at the evidence for a causal link between cannabis use and schizophrenia and find that there is probably a link in those who start using cannabis at an early age and also if the individual has a genetic predisposition to psychosis. Thus we recommend caution with regard to prescription of cannabis for such individuals. We found there is a small dependency rate with cannabis at around 9%, which needs to be taken seriously but compares to a rate of around 32% for dependency in tobacco use and 15% dependency with regard to alcohol. There may be a, as yet unproven, risk of respiratory cancer for smoked cannabis but nevertheless this route of administration is not recommended. The evidence for cognitive impairment in long term users is not clear but it is wise to be cautious in prescribing cannabis to younger people, given the possible susceptibility of the developing brain.


Overall, we conclude there is considerable literature demonstrating the efficacy of cannabis and/or available cannabis products in a number of important indications. Clearly there needs to be much further work with regard to the formulation of cannabis and the best THC:CBD ratio for different conditions and better and further studies are needed on both short and, more particularly, longer term effects. We consider that these studies will be facilitated by legalisation of cannabis for medical indications in strictly controlled circumstances with a quality-controlled product and a secure supply chain.

Repository Staff Only: item control page