Natural and Synthetic Cannabinoids for Agitation and Aggression in Alzheimer’s Disease: A Meta-Analysis

Natural and Synthetic Cannabinoids for Agitation and Aggression in Alzheimer’s Disease:

A Meta-Analysis

Vertical divider


Objective: This meta-analysis investigated the efficacy of cannabinoids on agitation and aggression in patients with Alzheimer’s disease (AD).

Data Sources: Electronic records up to August 2018 were searched from MEDLINE, EMBASE, and PsycINFO. Search terms included Alzheimer’s disease, agitation, aggression, and cannabinoids.

Study Selection: Double-blind, placebo-controlled studies investigating the effect of cannabinoids on agitation in patients with AD were included. Of the 1,336 records returned, 123 were reviewed and 6 (N = 251 participants) were included.

Data Extraction: Data on demographics, study setting, trial length, intervention, outcomes, and dropouts were extracted.

Results: There was no effect of cannabinoids as a group on agitation (standard mean difference: 0.69, P = .10), though there was significant heterogeneity (χ26 = 43.53, P < .00001, I2 = 86%). There was a trend for greater difference in agitation with synthetic cannabinoids over tetrahydrocannabinol (χ21 = 3.05, P = .08). Cannabinoids had a larger effect on agitation with greater cognitive impairment (B = 0.27, t6 = 2.93, P = .03). Cannabinoids did not change overall neuropsychiatric symptoms or body mass index (BMI). However, there was a significant difference in patients with a lower BMI compared to patients with a higher BMI (χ21 = 4.63, P = .03). Sedation was significantly greater with cannabinoids compared to placebo (risk ratio = 1.73, P = .04), but there were no differences in the occurrence of adverse events or dropouts due to an adverse event between treatment groups.

Conclusions: The efficacy of cannabinoids on agitation and aggression in patients with AD remains inconclusive, though there may be a signal for a potential benefit of synthetic cannabinoids. Safety should be closely monitored as cannabinoid treatment was associated with increased sedation.

J Clin Psychiatry 2019;80(2):18r12617

To cite: Ruthirakuhan M, Lanct×´t KL, Vieira D, et al. Natural and synthetic cannabinoids for agitation and aggression in Alzheimer’s disease: a meta-analysis. J Clin Psychiatry. 2019;80(2):18r12617.

To share:

aHurvitz Brain Sciences Program, Sunnybrook Research Institute, Toronto, Ontario, Canada

bDepartment of Pharmacology and Toxicology, University of Toronto, Toronto, Ontario, Canada

cDepartment of Psychiatry, University of Toronto, Toronto, Ontario, Canada

dDepartment of Psychiatry, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada

*Corresponding author: Nathan Herrmann, MD, Department of Psychiatry, Sunnybrook Health Sciences Centre, 2075 Bayview Ave, Ste FG08, Toronto, Ontario, Canada, M4N 3M5 (

Along with the cognitive deficits that are characteristic of Alzheimer’s disease (AD), neuropsychiatric symptoms commonly occur, and they typically increase in prevalence and severity with AD progression.1 An estimated 60%-90% of AD patients will experience at least one neuropsychiatric symptom during the course of their illness.2 Agitation is one of the most common and challenging neuropsychiatric symptoms to treat, occurring in 20%-50% of patients with moderate to severe AD.2-4 In addition to being associated with increased caregiver burden, agitation is also associated with more rapid AD progression, increased risk of falls, weight loss, and mortality.5-7 Therefore, this neuropsychiatric symptom is important to treat.

Current pharmacologic treatments of agitation in AD have been associated with limited to modest efficacy and questionable safety profiles. For example, despite the modest efficacy of atypical antipsychotics, their associated side effects, such as increased risk of mortality, parkinsonism, and cerebrovascular events, warrants caution when prescribing them to those with AD.8 Other psychotropic medications that have been investigated for the treatment of agitation in AD include typical antipsychotics, antidepressants, cholinesterase inhibitors, memantine, and benzodiazepines. However, studies have reported contradictory findings regarding their safety and efficacy for the treatment of agitation.9,10 As such, there is a clinical need to identify a treatment for agitation in AD that is both safe and effective.

The endocannabinoid system has been implicated in modulating AD pathology and symptomatology, positioning itself as a novel pharmacologic target for the treatment of agitation in AD.11-13 The endocannabinoid system pathway begins with the synthesis of endogenous cannabinoids, which are retrograde messengers that regulate neurotransmission.14,15 Following postsynaptic release, endogenous cannabinoids bind to G protein-coupled cannabinoid receptors 1 and 2 (CB1 and CB2).16 CB1 receptors are highly abundant in the central nervous system, particularly in the cerebral cortex and hippocampus, 2 brain structures that are essential in learning and memory function and that are also affected by AD pathology.17-21 CB2 receptors, which are more abundant in the cells and tissues of the immune system, have been associated with decreases in the production of proinflammatory molecules in vitro and with the removal of amyloid-β plaques in the brain.22-24 Animal model studies have demonstrated that deletion of the CB1 gene is associated with cognitive impairment.25,26 However, a number of clinical studies have demonstrated that acute and chronic cannabinoid use are associated with impaired cognitive function27 in healthy individuals and patient populations such as those with multiple sclerosis28 and schizophrenia.29 Cannabinoids such as tetrahydrocannabinol (THC) have psychotropic benefits that may benefit agitation in AD.30 More specifically, the full and partial CB1 agonist THC has been reported to have antianxiety, antidepressant, analgesic, and sedative effects in a variety of clinical populations. With the development of oral THC and synthetic cannabinoids, such as nabilone and dronabinol, the therapeutic potential of cannabinoids is now being evaluated.


Search Strategy

A literature search was conducted using the methodology recommended by the PRISMA guidelines.31 MEDLINE, EMBASE, PsycINFO, and Cochrane databases were searched for articles investigating THC or cannabinoids for AD up to August 2018. A sample search strategy (EMBASE) is presented in Supplementary Table 1.

Selection Criteria

Studies were included in the quantitative analysis if they enrolled patients who met diagnostic criteria for possible or probable AD, if agitation and aggression were assessed using a standardized neuropsychological tool before and after drug treatment, and if the study was a placebo-controlled trial. Two independent reviewers examined each article for eligibility, and disagreements regarding inclusion were settled through consensus.

clinical points

  • Medications used to treat agitation in patients with Alzheimer’s disease (AD) have limited to modest efficacy and are associated with harmful side effects.
  • Cannabinoids may offer benefits for agitation in AD due to their unique pharmacologic mechanism.
  • Evidence for cannabinoid treatment of agitation in AD remains inconclusive, though there may be a signal for a potential benefit of synthetic cannabinoids.

Data Extraction

Two independent reviewers examined each retrieved article for eligibility and extracted data using a data extraction form. We obtained missing data from the study authors when possible. One author of this report entered data into Review Manager (RevMan; Copenhagen, Denmark; The Nordic Cochrane Center, The Cochrane Collaboration; 2015), and the coauthors checked for accuracy. Discrepancies were resolved by consensus.

Evaluation of Quality Reporting

The quality of evidence was evaluated by 2 independent raters using the Newcastle Ottawa Scale and the Cochrane Collaboration’s risk of bias tool for assessing quality and risk of bias.32 Items were modified to better reflect the study population covered in this review. If the majority of reporting and control quality items were satisfied (> 50%), the study was rated as having high quality of evidence.

Statistical Analysis

Data from measures of agitation (Cohen Mansfield Agitation Inventory and Neuropsychiatric Inventory [NPI] agitation subscale), overall neuropsychiatric symptoms (NPI total) and cognition (Mini-Mental Status Examination [MMSE]) were extracted from the included studies. Data on changes in weight, body mass index (BMI), occurrence of sedation and somnolence, and adverse events were also extracted.

For continuous data, standardized mean differences (SMDs) or mean differences and 95% confidence intervals (CIs) were calculated using a random-effects model. Random-effects models were used for all analyses, as this method accounted for variable underlying effects in estimates of uncertainty, including differences within and between studies. Additionally, this method is preferable when significant heterogeneity is expected across studies.

Heterogeneity among results was evaluated in a χ2 analysis. The impact of heterogeneity was quantified using an I2 index as a measure of inconsistency.33 When substantial heterogeneity (I2 > 40%) was encountered, subgroup analyses were completed to examine the effects of study design, cannabinoid type, or patient characteristics. When subgroup analyses did not reveal sources of potential heterogeneity, meta-regression analyses were completed to assess the association between treatment outcomes and demographic characteristics such as baseline MMSE scores, age, and sex. As there were fewer than 10 studies that met inclusion criteria for this meta-analysis, publication bias was not assessed. Imprecision of results was evaluated through an assessment of 95% CIs. When the 95% CIs were wide and associated with a null result, the finding was associated with imprecision.

Effect sizes and heterogeneity were calculated using RevMan (2015). Meta-regressions were calculated using Stata (Release 10.1, StataCorp, College Station, Texas).


Literature Search Results

An initial database search identified 123 primary publications from a total of 1,336 records returned. Of these, 7 articles and 1 abstract investigated synthetic cannabinoids or THC for the treatment of agitation and/or aggression in patients with AD. Two articles34,35 did not report on placebo-controlled trials and were excluded from review. Of the 6 studies included (see Table 1), 2 investigated THC, 3 investigated dronabinol, and 1 investigated nabilone. Missing data were requested from the author(s) of the original studies when needed.36-38 There were no clinical trials reporting on the efficacy of cannabidiol for agitation in patients with AD (Supplementary Figure 1).

Table 1

Click figure to enlarge

Assessment of Risk of Bias and Quality of Evidence

Risk of bias from the Newcastle Ottawa Scale and the Cochrane Collaboration’s risk of bias tool are shown in Supplementary Table 2. Five of 7 studies satisfied the majority of quality reporting items. All patients in the included studies met diagnostic criteria for AD. All studies but 139 reported on demographic characteristics in patients treated with drug and placebo. All neuropsychological and cognitive tests were administered by a trained study staff member, and outcomes were assessed objectively. Five studies reported double-blinding of personnel, while 1 study39,40 compared data from an open-label trial with dronabinol against placebo.

Patient and Study Characteristics

The patient and study characteristics of included studies are summarized in Table 1. A total of 143 AD participants from 6 studies were included. As data from 3 of the 4 crossover trials37,38,41,42 did not include crossover or treatment-order effects, data from both phases of each participant were included. Additionally, 1 study36 contributed to 2 effect sizes, resulting in an overall sample size of 251 participants. The participants were elderly and predominantly male and had moderate to severe cognitive impairment (Table 1).

The mean study duration was 4.7 weeks (range, 2-6 weeks). The mean dose for trials with THC was 1.75 mg daily (range, 1.5-4.5 mg, daily), the dose used by all trials with dronabinol was 2.5 mg daily, and the mean therapeutic dose in the nabilone trial was 1.6 mg daily.

Four of the 6 studies were crossover randomized controlled trials (RCTs).37,38,41,42 As Volicer et al41 reported significant crossover effects, data from the first phase of the study, and not the second phase, were extracted for analysis. Van den Elsen et al37 contributed to 2 effect sizes as patients were randomized to 0.75 mg THC (or placebo) twice daily, followed by a 4-day washout period, and then randomized to 1.5 mg of THC (or placebo) twice daily; no significant crossover effects were reported. Lanct×´t et al42 reported no crossover or treatment-order effects in their trial with nabilone.42 Walther et al38 did not investigate crossover or treatment-order effects in their crossover trial of 2 patients.

Mahlberg and Walther39 conducted a pooled-analysis of 2 studies40,43 to compare the efficacy of dronabinol or melatonin versus placebo for agitation in patients with AD. For the purposes of this meta-analysis, dronabinol data from this trial were extracted from Walther et al40 and placebo data were extracted from Mahlberg and Walther.39

Effect of Drug on Agitation

Six studies were included in this analysis, with 1 study37 providing 2 effect sizes. There was no significant difference between cannabinoids and placebo (SMD = −0.69; 95% CI, −1.50 to 0.13; P = .10). However, as significant heterogeneity was present (χ26 = 43.53, P < .00001, I2 = 86%), we completed subgroup analyses with drug type, trial duration, inclusion of an RCT design, and AD severity.

THC versus synthetic cannabinoids. Two studies,36,37 1 of which37 contributed to 2 effect sizes, investigated the efficacy of THC for agitation, whereas 4 studies38,40-42 investigated the efficacy of synthetic cannabinoids for agitation. There was no significant benefit of drug over placebo within each subgroup, but there was a signal for an improvement in agitation in those randomized to synthetic cannabinoid compared to placebo (SMD: 1.67; 95% CI, 3.65 to 0.30; P = .10). In addition, there was a trending difference between subgroups that favored trials with synthetic cannabinoids over those with THC (χ21 = 3.05, P = .08) (Figure 1). However, the synthetic cannabinoids subgroup had significant heterogeneity (χ23 = 29.16, P < .00001, I2 = 90%) and imprecision due to a wide 95% CI.

Figure 1

Click figure to enlarge

Trial duration. Three studies36,38,40 had a trial duration of less than 6 weeks, whereas 3 studies, 37,41,42 1 of which37 contributed to 2 effect sizes, had a trial duration of 6 weeks. There was a signal for a potential benefit on agitation only in studies that had a trial duration of 6 weeks (SMD = 1.10; 95% CI, 2.35 to 0.16; 181 patients, P = .09). However, this finding was associated with significant heterogeneity (χ23 = 40.58, P < .00001, I2 = 93%) and imprecision due to a wide 95% CI. Furthermore, there were no significant differences between subgroups.

Randomized versus nonrandomized controlled trials. Four studies36,37,41,42 had an RCT trial design, 1 of which37 contributed to 2 effect sizes, and 2 studies did not have an RCT design.38,40 There was no significant effect of drug on agitation in either subgroup, and there were no significant differences between subgroups. Within the RCT subgroup, there was significant heterogeneity (χ24 = 43.25, P < .00001, I2 = 91%).

AD severity. Four studies36-38,40 included patients with moderate AD (mean ± SD baseline MMSE score = 11.5 ± 6.3), 1 of which37 contributed to 2 effect sizes. Two studies41,42 included patients in the more severe stages of AD (mean ± SD baseline MMSE score = 5.9 ± 6.9). Within the moderate (SMD: 0.06; 95% CI, 0.27 to 0.38; P = .73; I2 = 0%) and severe AD (SMD = 0.69; 95% CI, 1.50 to 0.13; P = .20; I2 = 96%) subgroups, there were no significant differences in change in agitation between treatment groups. Due to the lack of additional studies within this subgroup, a meta-regression was completed with mean baseline MMSE scores. This analysis indicated that lower mean MMSE score at baseline was significantly associated with greater improvement in agitation and aggression in drug compared to placebo (B = 0.27, t6 = 2.93, P = .03). A meta-regression with mean age (B = 0.15, t6 = 0.83, P = .44) at baseline and proportion of males (B = 1.43, t6 = 0.31, P = .77) identified no significant associations.

Effect of Drug on Overall Neuropsychiatric Symptoms

Five studies36-38,40,42 were included in this analysis, 1 of which37 contributed to 2 effect sizes. There were no significant differences in change in neuropsychiatric symptoms between treatment groups (mean difference = 0.67; 95% CI, −2.82 to 4.16; 229 patients; P = .71; I2 = 0%). Although there were no concerns with heterogeneity, due to a wide 95% CI there were concerns with imprecision of results.

Effect of Drug on BMI

For change in BMI over trial duration, 1 study41 reported data and 2 other studies36,42 provided data on request. There were no significant differences in the change in BMI between treatment groups (mean difference = 0.05; 95% CI, −0.15 to 0.25; n = 144 patients; P = .64). Due to heterogeneity (χ22 = 5.23, P = .07, I2 = 62%), a subgroup analysis with baseline BMI (high versus low) was completed. Although patients with a higher BMI36,42 did not report differences in change in BMI between treatment groups, in patients with a lower BMI,41 there was a significant increase in BMI in those who were randomized to drug, compared to placebo (mean difference = 2.00; 95% CI, 0.21 to 3.79; n = 30 patients; P = .03). There were also significant subgroup differences, (χ21 = 4.63, P = .03, I2 = 78%).

Effect of Drug on Occurrence of Sedation/Somnolence

Four studies,36,37,41,42 1 of which37 contributed to 2 effect sizes, reported on the number of patients who experienced sedation or somnolence during study participation. There were significant differences in the number of patients who experienced somnolence and/or sedation, favoring placebo (risk ratio [RR] = 1.73; 95% CI, 1.02 to 2.93;, n = 244 patients; P = .04). Although there were no serious concerns with heterogeneity, there were concerns with the imprecision due to a wide 95% CI.

Effect of Drug on Occurrence of Adverse Events and Dropouts Due to Adverse Events

Five studies,36-38,40 1 of which37 contributed to 2 effect sizes, reported on the number of patients who experienced 1 or more adverse events during study participation. However, as both studies by Walther et al38,40 reported no adverse events in either treatment group, they did not contribute to an effect size in this analysis. There were no significant differences in the number of patients who experienced an adverse event between treatment groups (RR = 1.26; 95% CI, 0.93 to 1.71; 210 patients, P = .68).

There were also no significant differences in the number of patients who discontinued the study early due to an adverse event between treatment groups (RR = 1.31; 95% CI, 0.34 to 5.13; 152 patients; P = .70). In both analyses, there were no serious concerns with heterogeneity or imprecision of results.


This meta-analysis quantitatively investigated the efficacy and safety of cannabinoids for the treatment of agitation in patients with AD. Mechanistically, the endocannabinoid system has been implicated as a novel therapeutic target that may have benefits for agitation and aggression. CB1 receptors, located in brain areas associated with the limbic system such as the hypothalamus, have been linked to mood-related disorders and AD-related neuropsychiatric symptoms such as aggression.44 CB1 hyperactivity and the hypothalamus have also been linked to monoamine neurotransmitter systems such as the serotonergic and dopaminergic systems, both of which have also been linked to aggression and have been targeted for pharmacologic intervention with antidepressants and antipsychotics.45-47 Despite mechanistic support for an association between the endocannabinoid system and agitation, we did not identify a significant benefit of cannabinoids over placebo for the treatment of agitation.

Post hoc analyses suggested several potential reasons for this finding. In a comparison of THC to synthetic cannabinoids, the efficacy of THC on agitation has been consistently negative, whereas there was a signal for a potential benefit of synthetic cannabinoids on agitation. Pharmacologic differences between THC and synthetic cannabinoids may contribute to the differences observed between the 2 groups. Specifically, compared to synthetic cannabinoids, the oral THC used by van den Elsen et al36,37 has a shorter half-life and a shorter time to peak effect.48 Although these pharmacologic properties may be preferred for the acute symptomatic management of agitation, synthetic cannabinoids may be beneficial for persistent and recurrent agitation that is clinically significant. It is also difficult to compare doses used, as there are no THC equivalents. However, THC may have been relatively underdosed. Compared to the synthetic cannabinoids, THC has a shorter time to maximum concentration, a faster absorption period, and a 2-5 times smaller maximal concentration.49 Therefore, a higher dose may be required to observe a clinically significant improvement on agitation. The dronabinol trials included in this meta-analysis may have also been underdosed. However, as the retrospective chart review by Passmore35 noticed a benefit on agitation with a mean dose of 7.03 mg/d, future studies should consider dosing higher than 2.5 mg daily.

Post hoc analyses also suggested that patients with greater AD severity demonstrated greater improvements in agitation when treated with cannabinoids. Consistent with the finding that agitation increases in frequency and severity with AD progression,50 postmortem studies have reported that anandamide, an endogenous cannabinoid that acts on CB1 and CB2, is reduced in the midfrontal and temporal cortices of patients with AD compared to healthy individuals and is correlated with increased amyloid-β pathophysiology and cognitive decline.51 In addition, fatty acid amide hydrolase, an endocannabinoid metabolizing enzyme, has been shown to be up-regulated in neuritic plaque-associated glia52 and in peripheral blood mononuclear cells53 of AD patients, further contributing to the decline in circulating endogenous cannabinoids. As such, exogenously activating the endocannabinoid system through pharmacologic agents may offer greater benefits to patients with greater AD severity. Conversely, postmortem studies have demonstrated that CB1 density decreases with Braak staging in patients with AD.54 Although this finding may suggest that patients in the more severe stages of AD may not be able to benefit as well from treatment with an exogenous ligand, the studies included in this meta-analysis collected antemortem data, and subjects were not followed up postmortem. As such, we are unable to comment on whether CB1 density was associated with our findings. Previous trials with psychotropics have had similar findings with regard to efficacy for agitation and AD severity. For example, a meta-analysis investigating the efficacy of atypical antipsychotics for agitation in patients with AD8 reported greater improvements in agitation in patients with severe AD compared to those with mild-to-moderate AD. The lack of efficacy in mild-to-moderate AD was very likely due to a floor effect introduced by an overall lower severity of agitation. Additional studies with larger sample sizes are required to provide confirmatory evidence regarding the relationship between efficacy of cannabinoids and severity of AD.

In this meta-analysis, cannabinoid treatment did not demonstrate a significant effect on total neuropsychiatric symptoms compared to placebo. The study by Volicer et al41 was the only one not included in this analysis as they did not report on change in neuropsychiatric symptoms over treatment duration. However, as patients included in the study by Volicer et al41 had the lowest mean MMSE scores at baseline, those patients may have also had more frequent and severe neuropsychiatric symptoms as AD severity has been positively correlated with the severity of neuropsychiatric symptoms.1,55 Therefore, it is possible that cannabinoids may improve neuropsychiatric symptoms in patients with severe AD with more severe neuropsychiatric symptoms. Furthermore, as the endocannabinoid system has been implicated in the regulation of neuropsychiatric symptoms in patients with schizophrenia, depression, bipolar disorder, and posttraumatic stress disorder, additional studies in patients with AD and agitation and aggression are warranted.56

We found no significant effect of cannabinoids on BMI in 336,41,42 of the 6 studies that reported on the change in BMI over treatment duration. However, Volicer et al41 reported that dronabinol significantly increased BMI over 6 weeks compared to placebo, a finding associated with a large effect size (2.00). Of relevance, the mean ± SD BMI of patients included in the study by Volicer et al41 (22.6 ± 2.5) was much lower compared to that of patients included in the studies by van den Elsen et al36 (25.0 ± 3.5) and Lanct×´t et al42 (24.5 ± 3.9), suggesting that cannabinoids may offer benefits for weight gain in patients who have a low BMI. CB1 agonism has been associated with increased feeding behavior, food craving, enjoyment during feeding, and energy deposition of fat into adipose tissues.57 Therefore, endocannabinoid system modulation may offer benefits to patients with AD, since weight loss and anorexia commonly occur in the elderly and have been associated with mortality, increased neuropsychiatric symptoms, and reduced quality of life.58

With regard to side effects, the quantitative findings of our meta-analysis suggest cannabinoids are associated with a greater risk of sedation and somnolence compared to placebo and thus should be carefully monitored in future trials. Given the advanced age of this patient population and high risk of sedation associated with cannabinoid use,59 future trials should consider flexible dosing based on tolerability. The study by Lanct×´t et al42 was the only one included in this meta-analysis that employed a flexible dosing regimen. Although mild sedation may be a therapeutic benefit in the treatment of a patient with severe agitation or aggression, persistent and excessive sedation is a major concern in patients with advanced AD, as it has been associated with cognitive worsening, impaired gait and balance, increased risk of falls, and reduced food intake.60

Two additional reports34,35 investigated the efficacy of synthetic cannabinoids for agitation or aggression in patients with AD. However, since 1 report was a retrospective chart review34 and 1 was a single case study,35 neither was included in the quantitative analysis of our meta-analysis. Woodward et al34 reported that dronabinol (mean dose = 7.03 mg daily) significantly reduced agitated and aggressive behaviors in patients with AD. They also reported an overall increase in meal completion with dronabinol, though there was no significant increase in weight. The investigators attributed this finding to a short observation period (mean = 16.88 days; range, 4-50 days), and missing data. In a single case study with nabilone (dose = 0.5 mg twice daily), Passmore35 reported improved agitation and aggression over the 6 weeks of treatment, which persisted for over 3 months. There are currently 3 ongoing placebo-controlled RCTs investigating the use of nabilone (NCT02351882), dronabinol (NCT0272257), and Avidekel oil (20:1 cannabidiol:THC) (NCT03328676) for the treatment of agitation/aggression. Cannabidiol may also have psychotropic benefits for AD patients with agitation, as it has demonstrated anxiolytic benefits in patients with schizophrenia and social anxiety disorder.61 However, to date cannabidiol has not been investigated in patients with AD. The results of those studies will provide efficacy and safety data on the use of cannabinoids for agitation/aggression and will also provide feasibility data for future trials investigating the therapeutic use of cannabinoids in this patient population.


This meta-analysis quantified the existing evidence on the efficacy and safety of cannabinoid use for agitation and aggression in patients with AD. The limitations in the context of this review should be considered when interpreting the evidence. One major limitation is the small number of studies included in this meta-analysis. However, the results of this meta-analysis add to previous reports based on limited evidence30 and are timely, as the use of cannabinoids for the treatment of agitation in AD has been gaining interest over the past decade due to the lack of efficacy associated with current pharmacotherapies for agitation and recent changes in the legislation of THC in North America. Furthermore, as cannabinoids are available in capsule form, their therapeutic potential can be studied in a controlled manner.

This meta-analysis included only 251 participants, and the small number of studies contributed to limitations in our subgroup analyses. For example, in the subgroup analysis evaluating efficacy of cannabinoids for agitation in patients with moderate AD and severe AD, only 2 studies41,42 recruited patients with severe AD. The included studies were also limited by the relatively small sample sizes with large variability, which likely resulted in wide 95% CIs. Additional studies with larger sample sizes that include patients with severe AD are needed to confirm the findings of this meta-analysis.

The short trial duration of the included studies may also be considered a limitation of this meta-analysis. Only 3 studies37,41,42 had a study duration of 6 weeks, whereas the remaining studies had durations of 2,40 3,36 and 538 weeks. As such, the studies included in this meta-analysis may have not had a sufficient trial duration to notice an observable benefit on agitation or aggression, neuropsychiatric symptoms, and weight.

One of the studies included in this meta-analysis was a pooled-analysis of an open-label trial with dronabinol40 and a trial with a placebo control.39 As a result, the quality of these data were lowered since the dronabinol trial was not randomized and was not necessarily placebo-controlled. However, as this meta-analysis was limited by the number of studies included, we thought the strengths associated with including this study outweighed its associated limitations.

Three of the studies37,41,42 included in this meta-analysis were crossover RCTs. An RCT with a crossover design offers advantages in a study population that is difficult to recruit, as each patient acts as his or her own control, meaning required sample size can be halved. However, since Volicer et al41 reported significant crossover effects, data from the first, and not second, phase of that study were included in our quantitative analysis. Carryover effects from Volicer et al41 were most likely due to the lack of a washout period between treatment phases. Because cannabinoids are lipophilic compounds, a sufficient washout period is required to reduce carryover effects in a crossover RCT.

None of the included studies investigated or reported on pain. Pain occurs in approximately 50% of patients with AD62 and in 60%-80% of institutionalized patients.63 In addition, the majority of these patients experience chronic pain for a minimum of 3-6 months.64 Since untreated pain is associated with a reduced quality of life, functional status, sleep, and appetite disturbances and increased severity of neuropsychiatric symptoms, the treatment of pain in patients with AD is an important clinical need.63,65,66 Preclinical studies and clinical studies in non-AD patient populations have also implicated the endocannabinoid system in having benefits for pain, a symptom that is difficult to both treat and identify in patients with AD. THC, a cannabinoid that is a partial and full agonist at CB1 on glutamatergic and γ-aminobutyric neurons, respectively, and a partial agonist at CB2, has demonstrated anti-inflammatory properties that may have analgesic effects.67,68 Furthermore, THC has been shown to have an impact on opioid, serotonergic, and glutamatergic receptors, all of which have been shown to have a role in the development and maintenance of neuropathic pain.67 As current therapeutic recommendations for pain management in AD have modest efficacy and are associated with side effects, and as pain has been linked to the exacerbation of agitation, future trials investigating cannabinoids in AD should consider including pain as an outcome measure.69

There have been no studies that have investigated the efficacy of cannabinoids on cognitive function as a primary outcome measure in those with AD. Since cannabinoids have been associated with cognitive worsening in other patient populations,27 future trials should also assess cognition over time as a measurement of safety. Findings from those studies would also inform clinicians and researchers on whether patients who have mild, moderate, or severe AD with agitation should be targeted for treatment with a cannabinoid.


The results of this meta-analysis provide little evidence of efficacy for the effect of cannabinoids on agitation in AD. While our findings suggest that THC for the treatment of agitation has been consistently negative, results with synthetic cannabinoids are inconclusive due to substantial heterogeneity. In spite of potential mechanistic linkages between the endocannabinoid system and agitation, other neuropsychiatric symptoms, weight, and pain, there are to date only a small number of studies with small numbers of patients and weak methodologies. Whether this small number is a result of a negative bias toward cannabinoids in mainstream medicine or a healthy skepticism to a class of compounds that might have important adverse effects in frail elderly patients with neurodegenerative disorders is unclear. What is clear, however, is that treatment of agitation and aggression in AD is still an important public health priority, and this potentially useful class of compounds urgently deserves more attention with rigorous studies.

Submitted: October 17, 2018; accepted December 12, 2018.

Published online: January 29, 2019.

Potential conflicts of interests: Dr Herrmann has received research grants from the Alzheimer Drug Discovery Fund, the Alzheimer Society of Canada, the National Institutes of Health Research, Canadian Institute of Health Research (CIHR), Brain Canada, Alzheimer’s Association, Lundbeck, Axovant Sciences Ltd, and Roche and consultation fees from Lilly, Merck, and Astellas. Ms Ruthirakuhan is funded by a CIHR Doctoral Research Award (the Frederick Banting and Charles Best Canada Graduate Scholarships Doctoral Award). Dr Lanct×´t has received research grants from the National Institute of Aging, Alzheimer’s Drug Discovery Fund, the Alzheimer Society of Canada, Alzheimer’s Association, CIHR, Brain Canada, AbbVie, Lundbeck, and Axovant Sciences Ltd and honoraria from AbbVie, Otsuka, and Lundbeck. Ms Vieira has no known conflicts of interest.

Funding/support: Alzheimer’s Drug Discovery Foundation, Alzheimer Society of Canada, and Canadian Consortium on Neurodegeneration in Aging provided support for the nabilone study (NCT02351882).

Role of the sponsor: The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of this article.

Supplementary material: Available at PSYCHIATRIST.COM.


1. Lyketsos CG, Carrillo MC, Ryan JM, et al. Neuropsychiatric symptoms in Alzheimer’s disease. Alzheimers Dement. 2011;7(5):532-539. PubMed CrossRef

2. Lyketsos CG, Lopez O, Jones B, et al. Prevalence of neuropsychiatric symptoms in dementia and mild cognitive impairment: results from the cardiovascular health study. JAMA. 2002;288(12):1475-1483. PubMed CrossRef

3. McKeith I, Cummings J. Behavioural changes and psychological symptoms in dementia disorders. Lancet Neurol. 2005;4(11):735-742. PubMed CrossRef

4. Pitkala KH, Laurila JV, Strandberg TE, et al. Behavioral symptoms and the administration of psychotropic drugs to aged patients with dementia in nursing homes and in acute geriatric wards. Int Psychogeriatr. 2004;16(1):61-74. PubMed CrossRef

5. Shin IS, Carter M, Masterman D, et al. Neuropsychiatric symptoms and quality of life in Alzheimer disease. Am J Geriatr Psychiatry. 2005;13(6):469-474. PubMed CrossRef

6. Miller EA, Schneider LS, Rosenheck RA. Predictors of nursing home admission among Alzheimer’s disease patients with psychosis and/or agitation. Int Psychogeriatr. 2011;23(1):44-53. PubMed CrossRef

7. Mohamed S, Rosenheck R, Lyketsos CG, et al. Effect of second-generation antipsychotics on caregiver burden in Alzheimer’s disease. J Clin Psychiatry. 2012;73(1):121-128. PubMed CrossRef

8. Schneider LS, Dagerman K, Insel PS. Efficacy and adverse effects of atypical antipsychotics for dementia: meta-analysis of randomized, placebo-controlled trials. Am J Geriatr Psychiatry. 2006;14(3):191-210. PubMed CrossRef

9. Herrmann N, Gauthier S. Diagnosis and treatment of dementia, 6: management of severe Alzheimer disease. CMAJ. 2008;179(12):1279-1287. PubMed CrossRef

10. Hogan DB, Bailey P, Black S, et al. Diagnosis and treatment of dementia, 5: nonpharmacologic and pharmacologic therapy for mild to moderate dementia. CMAJ. 2008;179(10):1019-1026. PubMed CrossRef

11. Sherman C, Ruthirakuhan M, Vieira D, et al. Cannabinoids for the treatment of neuropsychiatric symptoms, pain and weight loss in dementia. Curr Opin Psychiatry. 2018;31(2):140-146. PubMed CrossRef

12. Liu CS, Ruthirakuhan M, Chau SA, et al. Pharmacological management of agitation and aggression in Alzheimer’s disease: a review of current and novel treatments. Curr Alzheimer Res. 2016;13(10):1134-1144. PubMed CrossRef

13. Liu CS, Chau SA, Ruthirakuhan M, et al. Cannabinoids for the treatment of agitation and aggression in Alzheimer’s disease. CNS Drugs. 2015;29(8):615-623. PubMed CrossRef

14. Di Marzo V, Fontana A, Cadas H, et al. Formation and inactivation of endogenous cannabinoid anandamide in central neurons. Nature. 1994;372(6507):686-691. PubMed CrossRef

15. Stella N, Schweitzer P, Piomelli D. A second endogenous cannabinoid that modulates long-term potentiation. Nature. 1997;388(6644):773-778. PubMed CrossRef

16. Lauckner JE, Jensen JB, Chen HY, et al. GPR55 is a cannabinoid receptor that increases intracellular calcium and inhibits M current. Proc Natl Acad Sci U S A. 2008;105(7):2699-2704. PubMed CrossRef

17. Glass M, Dragunow M, Faull RL. Cannabinoid receptors in the human brain: a detailed anatomical and quantitative autoradiographic study in the fetal, neonatal and adult human brain. Neuroscience. 1997;77(2):299-318. PubMed CrossRef

18. Mackie K. Distribution of cannabinoid receptors in the central and peripheral nervous system. Handb Exp Pharmacol. 2005;168(168):299-325. PubMed CrossRef

19. Kathmann M, Bauer U, Schlicker E. CB1 receptor density and CB1 receptor-mediated functional effects in rat hippocampus are decreased by an intracerebroventricularly administered antisense oligodeoxynucleotide. Naunyn Schmiedebergs Arch Pharmacol. 1999;360(4):421-427. PubMed CrossRef

20. Kפrkkפine E, Tanila H, Laitinen JT. Functional autoradiography shows unaltered cannabinoid CB1 receptor signalling in hippocampus and cortex of APP/PS1 transgenic mice. CNS Neurol Disord Drug Targets. 2012;11(8):1038-1044. PubMed CrossRef

21. Rapp PR, Heindel WC. Memory systems in normal and pathological aging. Curr Opin Neurol. 1994;7(4):294-298. PubMed CrossRef

22. Klegeris A, Bissonnette CJ, McGeer PL. Reduction of human monocytic cell neurotoxicity and cytokine secretion by ligands of the cannabinoid-type CB2 receptor. Br J Pharmacol. 2003;139(4):775-786. PubMed CrossRef

23. Sheng WS, Hu S, Min X, et al. Synthetic cannabinoid WIN55,212-2 inhibits generation of inflammatory mediators by IL-1β-stimulated human astrocytes. Glia. 2005;49(2):211-219. PubMed CrossRef

24. Tolón RM, Nú×±ez E, Pazos MR, et al. The activation of cannabinoid CB2 receptors stimulates in situ and in vitro beta-amyloid removal by human macrophages. Brain Res. 2009;1283:148-154. PubMed CrossRef

25. Saravia R, Flores ׁ, Plaza-Zabala A, et al. CB1 cannabinoid receptors mediate cognitive deficits and structural plasticity changes during nicotine withdrawal. Biol Psychiatry. 2017;81(7):625-634. PubMed CrossRef

26. Degroot A, Nomikos GG. Genetic deletion and pharmacological blockade of CB1 receptors modulates anxiety in the shock-probe burying test. Eur J Neurosci. 2004;20(4):1059-1064. PubMed CrossRef

27. Shrivastava A, Johnston M, Tsuang M. Cannabis use and cognitive dysfunction. Indian J Psychiatry. 2011;53(3):187-191. PubMed CrossRef

28. Honarmand K, Tierney MC, O’ Connor P, et al. Effects of cannabis on cognitive function in patients with multiple sclerosis. Neurology. 2011;76(13):1153-1160. PubMed CrossRef

29. Manseau MW, Goff DC. Cannabinoids and schizophrenia: risks and therapeutic potential. Neurotherapeutics. 2015;12(4):816-824. PubMed CrossRef

30. Krishnan S, Cairns R, Howard R. Cannabinoids for the treatment of dementia. Cochrane Database Syst Rev. 2009;(2):CD007204. PubMed

31. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1-e34. PubMed CrossRef

32. Higgins JP, Altman DG, G׸tzsche PC, et al; Cochrane Bias Methods Group; Cochrane Statistical Methods Group. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. PubMed CrossRef

33. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557-560. PubMed CrossRef

34. Woodward MR, Harper DG, Stolyar A, et al. Dronabinol for the treatment of agitation and aggressive behavior in acutely hospitalized severely demented patients with noncognitive behavioral symptoms. Am J Geriatr Psychiatry. 2014;22(4):415-419. PubMed CrossRef

35. Passmore MJ. The cannabinoid receptor agonist nabilone for the treatment of dementia-related agitation. Int J Geriatr Psychiatry. 2008;23(1):116-117. PubMed CrossRef

36. van den Elsen GA, Ahmed AI, Verkes RJ, et al. Tetrahydrocannabinol for neuropsychiatric symptoms in dementia: a randomized controlled trial. Neurology. 2015;84(23):2338-2346. PubMed CrossRef

37. van den Elsen GAH, Ahmed AIA, Verkes RJ, et al. Tetrahydrocannabinol in behavioral disturbances in dementia: a crossover randomized controlled trial. Am J Geriatr Psychiatry. 2015;23(12):1214-1224. PubMed CrossRef

38. Walther S, Schüpbach B, Seifritz E, et al. Randomized, controlled crossover trial of dronabinol, 2.5 mg, for agitation in 2 patients with dementia. J Clin Psychopharmacol. 2011;31(2):256-258. PubMed CrossRef

39. Mahlberg R, Walther S. Actigraphy in agitated patients with dementia: monitoring treatment outcomes. Z Gerontol Geriatr. 2007;40(3):178-184. PubMed CrossRef

40. Walther S, Mahlberg R, Eichmann U, et al. Delta-9-tetrahydrocannabinol for nighttime agitation in severe dementia. Psychopharmacology (Berl). 2006;185(4):524-528. PubMed CrossRef

41. Volicer L, Stelly M, Morris J, et al. Effects of dronabinol on anorexia and disturbed behavior in patients with Alzheimer’s disease. Int J Geriatr Psychiatry. 1997;12(9):913-919. PubMed CrossRef

42. Lanct×´t KL, Ruthirakuhan M, Gallagher D, et al. Nabilone significantly improves agitation/aggression in patients with moderate-to-severe Alzheimer’s disease: preliminary results of a placebo-controlled, double-blind, cross-over trial. Paper presented at the 2018 Alzheimer’s Association International Conference. July 22-26, 2018; Chicago, Illinois.

43. Mahlberg R, Kunz D, Sutej I, et al. Melatonin treatment of day-night rhythm disturbances and sundowning in Alzheimer disease: an open-label pilot study using actigraphy. J Clin Psychopharmacol. 2004;24(4):456-459. PubMed CrossRef

44. Rodriguez-Arias M, Navarrete F, Daza-Losada M, et al. CB1 cannabinoid receptor-mediated aggressive behavior. Neuropharmacology. 2013;75:172-180. PubMed CrossRef

45. Haj-Dahmane S, Shen RY. Chronic stress impairs α1-adrenoceptor-induced endocannabinoid-dependent synaptic plasticity in the dorsal raphe nucleus. J Neurosci. 2014;34(44):14560-14570. PubMed CrossRef

46. Lanct×´t KL, Herrmann N, Mazzotta P. Role of serotonin in the behavioral and psychological symptoms of dementia. J Neuropsychiatry Clin Neurosci. 2001;13(1):5-21. PubMed CrossRef

47. Oleson EB, Cheer JF. A brain on cannabinoids: the role of dopamine release in reward seeking. Cold Spring Harb Perspect Med. 2012;2(8):a012229. PubMed CrossRef

48. Klumpers LE, Beumer TL, van Hasselt JG, et al. Novel Δ(9) -tetrahydrocannabinol formulation Namisol has beneficial pharmacokinetics and promising pharmacodynamic effects. Br J Clin Pharmacol. 2012;74(1):42-53. PubMed CrossRef

49. Tai S, Fantegrossi WE. Synthetic cannabinoids: pharmacology, behavioral effects, and abuse potential. Curr Addict Rep. 2014;1(2):129-136. PubMed CrossRef

50. Zahodne LB, Ornstein K, Cosentino S, et al. Longitudinal relationships between Alzheimer disease progression and psychosis, depressed mood, and agitation/aggression. Am J Geriatr Psychiatry. 2015;23(2):130-140. PubMed CrossRef

51. Jung KM, Astarita G, Yasar S, et al. An amyloid β42-dependent deficit in anandamide mobilization is associated with cognitive dysfunction in Alzheimer’s disease. Neurobiol Aging. 2012;33(8):1522-1532. PubMed CrossRef

52. Benito C, Nú×±ez E, Tolón RM, et al. Cannabinoid CB2 receptors and fatty acid amide hydrolase are selectively overexpressed in neuritic plaque-associated glia in Alzheimer’s disease brains. J Neurosci. 2003;23(35):11136-11141. PubMed CrossRef

53. D’ Addario C, Di Francesco A, Arosio B, et al. Epigenetic regulation of fatty acid amide hydrolase in Alzheimer disease. PLoS One. 2012;7(6):e39186. PubMed CrossRef

54. Manuel I, González de San Román E, Giralt MT, et al. Type-1 cannabinoid receptor activity during Alzheimer’s disease progression. J Alzheimers Dis. 2014;42(3):761-766. PubMed CrossRef

55. Hashimoto M, Yatabe Y, Ishikawa T, et al. Relationship between dementia severity and behavioral and psychological symptoms of dementia in dementia with Lewy Bodies and Alzheimer’s disease patients. Dement Geriatr Cogn Dis Extra. 2015;5(2):244-252. PubMed CrossRef

56. Marco EM, Garc×­a-Gutiérrez MS, Bermúdez-Silva FJ, et al. Endocannabinoid system and psychiatry: in search of a neurobiological basis for detrimental and potential therapeutic effects. Front Behav Neurosci. 2011;5:63. PubMed CrossRef

57. Sansone RA, Sansone LA. Marijuana and body weight. Innov Clin Neurosci. 2014;11(7-8):50-54. PubMed

58. Gillette-Guyonnet S, Nourhashemi F, Andrieu S, et al. Weight loss in Alzheimer disease. Am J Clin Nutr. 2000;71(2):637S-642S. PubMed CrossRef

59. Sachs J, McGlade E, Yurgelun-Todd D. Safety and toxicology of cannabinoids. Neurotherapeutics. 2015;12(4):735-746. PubMed CrossRef

60. Vitiello MV, Borson S. Sleep disturbances in patients with Alzheimer’s disease: epidemiology, pathophysiology and treatment. CNS Drugs. 2001;15(10):777-796. PubMed CrossRef

61. Khoury JM, Neves MCLD, Roque MAV, et al. Is there a role for cannabidiol in psychiatry [published online ahead of print February 20, 2017]? World J Biol Psychiatry. PubMed CrossRef

62. Achterberg WP, Pieper MJ, van Dalen-Kok AH, et al. Pain management in patients with dementia. Clin Interv Aging. 2013;8:1471-1482. PubMed CrossRef

63. Corbett A, Husebo B, Malcangio M, et al. Assessment and treatment of pain in people with dementia. Nat Rev Neurol. 2012;8(5):264-274. PubMed CrossRef

64. Miró J, Paredes S, Rull M, et al. Pain in older adults: a prevalence study in the Mediterranean region of Catalonia. Eur J Pain. 2007;11(1):83-92. PubMed CrossRef

65. Foley D, Ancoli-Israel S, Britz P, et al. Sleep disturbances and chronic disease in older adults: results of the 2003 National Sleep Foundation Sleep in America Survey. J Psychosom Res. 2004;56(5):497-502. PubMed CrossRef

66. Bosley BN, Weiner DK, Rudy TE, et al. Is chronic nonmalignant pain associated with decreased appetite in older adults? preliminary evidence. J Am Geriatr Soc. 2004;52(2):247-251. PubMed CrossRef

67. Russo EB. Cannabinoids in the management of difficult to treat pain. Ther Clin Risk Manag. 2008;4(1):245-259. PubMed CrossRef

68. Laaris N, Good CH, Lupica CR. Delta9-tetrahydrocannabinol is a full agonist at CB1 receptors on GABA neuron axon terminals in the hippocampus. Neuropharmacology. 2010;59(1-2):121-127. PubMed CrossRef

69. Husebo BS, Achterberg W, Flo E. Identifying and managing pain in people with Alzheimer’s disease and other types of dementia: a systematic review. CNS Drugs. 2016;30(6):481-497. PubMed CrossRef

Related Articles

Volume: 80

Quick Links: Impulse-Control Disorders , Violence and Aggression


Buy this Article as a PDF