Cannabinoids for the treatment of dementia


Dementia is the name for a range of conditions associated with an ongoing loss of thinking ability, memory, and other mental abilities. Medical treatments for dementia are limited and confer modest benefits. Many medications and non-pharmacological interventions are used for behavioural and psychological symptoms of dementia. However,  there are continuing problems with their lack of efficacy, safety, and feasibility. Accordingly, new, safe, and more effective treatments are needed for dementia and its associated symptoms. 

Our review question

The cannabinoids are one potential agent under investigation for the treatment of dementia. The purpose of this systematic review was to investigate whether cannabinoids could help people with dementia, and whether they have any potential harmful effects.

What we did

We searched databases of scientific studies to find studies that had to randomly decide whether people would be treated with cannabinoids or a comparator. We combined the results of included studies to estimate the effects of cannabinoids. We also assessed how well these studies were conducted and how credible their results were.

What we found

We searched for relevant studies that had been published up to June 2020. We found four trials that met the inclusion criteria for this review. A total of 126 people were included in those four trials. Most participants had Alzheimer's disease, and a few had vascular dementia or mixed dementia.

One trial was undertaken in the USA, one in Canada, and two in The Netherlands. Trials used different types of cannabinoids. Studies reported that cannabinoids have little or no effect on memory and thinking. They reported different results regarding overall behavioural and psychological symptoms of dementia, based on the types of individuals who were reporting data. Family caregivers did not report a beneficial effect of cannabinoids on behaviour and psychological symptoms, but nursing staff reported improvement of symptoms among participants receiving cannabinoids. Harms were reported in all studies, among both participants taking cannabinoids and those taking placebo. However, we could not combine the total number of harmful events due to problems with data reporting. There was no significant difference between participants taking cannabinoids and those given placebo in the total numbers of harmful events belonging to nervous system disorders, psychiatric disorders, and gastrointestinal disorders. Sedation (including lethargy) was more frequent in participants taking cannabinoids, but these results were uncertain. We have very low confidence in these results because studies included a small number of participants, there were differences between the studies, and their results were uncertain.

Our conclusions

Based on data from four small trials of short duration, it is uncertain whether cannabinoids have any beneficial or harmful effects on dementia, compared to placebo. Even if the benefit reported in these studies is real, the effect was modest and may not be important to people living with dementia. Furthermore, available studies were very short, with efficacy examined over 3 to 14 weeks, and one study did not report its methods and results completely. A large, well‐conducted study is needed to understand better if cannabinoids are a useful treatment for people living with dementia.

Authors' conclusions: 

Based on data from four small, short, and heterogeneous placebo-controlled trials, we cannot be certain whether cannabinoids have any beneficial or harmful effects on dementia. If there are benefits of cannabinoids for people with dementia, the effects may be too small to be clinically meaningful. Adequately powered, methodologically robust trials with longer follow-up are needed to properly assess the effects of cannabinoids in dementia.

Read the full abstract...

Dementia is a common chronic condition, mainly affecting older adults, characterised by a progressive decline in cognitive and functional abilities. Medical treatments for dementia are limited. Cannabinoids are being investigated for the treatment of dementia.


To determine the efficacy and safety of cannabinoids for the treatment of dementia.

Search strategy: 

We searched ALOIS - the Cochrane Dementia and Cognitive Improvement Group’s Specialised Register - on 8 July 2021, using the terms cannabis or cannabinoid or endocannabinoid or cannabidiol or THC or CBD or dronabinol or delta-9-tetrahydrocannabinol or marijuana or marihuana or hashish. The register contains records from all major healthcare databases (the Cochrane Library, MEDLINE, Embase, PsycINFO, CINAHL, LILACS), as well as from many clinical trials registries and grey literature sources.

Selection criteria: 

We included all randomised controlled trials (RCTs) of cannabinoids for the treatment of dementia. We included participants of any age and of either sex with diagnosed dementia of any subtype, or with unspecified dementia of any severity, from any setting. We considered studies of cannabinoids administered by any route, at any dose, for any duration, compared with placebo, no treatment, or any active control intervention.

Data collection and analysis: 

Two review authors independently screened and selected studies for inclusion, extracted data, and assessed the risk of bias in included studies. When necessary, other review authors were involved in reaching consensus decisions. We conducted meta‐analyses using a generic inverse variance fixed‐effect model to derive estimates of effect size. We used GRADE methods to assess our confidence in the effect estimates.

Main results: 

We included four studies (126 participants) in this review. Most participants had Alzheimer's disease; a few had vascular dementia or mixed dementia. Three studies had low risk of bias across all domains; one study had unclear risk of bias for the majority of domains. 

The included studies tested natural delta-9-tetrahydrocannabinol (THC) (Namisol) and two types of synthetic THC analogue (dronabinol and nabilone). Three trials had a cross-over design. Interventions were applied over 3 to 14 weeks; one study reported adverse events over 70 weeks of follow-up. One trial was undertaken in the USA, one in Canada, and two in The Netherlands. Two studies reported non-commercial funding, and two studies were conducted with the support of both commercial and non-commercial funding.

Primary outcomes in this review were changes in global and specific cognitive function, overall behavioural and psychological symptoms of dementia (BPSD), and adverse events.

We found very low-certainty evidence suggesting there may be little or no clinically important effect of a synthetic THC analogue on cognition assessed with the standardised Mini-Mental State Examination (sMMSE) (mean difference (MD) 1.1 points, 95% confidence interval (CI) 0.1 to 2.1; 1 cross-over trial, 28 participants). 

We found low-certainty evidence suggesting there may be little or no clinically important effect of cannabinoids on overall behavioural and psychological symptoms of dementia assessed with the Neuropsychiatric Inventory (or its modified nursing home version) (MD -1.97, 95% CI -3.87 to -0.07; 1 parallel group and 2 cross-over studies, 110 participants).

All included studies reported data on adverse events. However, the total number of adverse events, the total numbers of mild and moderate adverse events, and the total number of serious adverse events (SAEs) were not reported in a way that permitted meta-analysis.  There were no clear differences between groups in numbers of adverse events, with the exception of sedation (including lethargy), which was more frequent among participants taking nabilone (N = 17) than placebo (N = 6) (odds ratio (OR) 2.83, 95% CI 1.07 to 7.48; 1 cross-over study, 38 participants). We judged the certainty of evidence for adverse event outcomes to be low or very low due to serious concerns regarding imprecision and indirectness.