Get Involved
About Us
Our Members

Research Review: Systematic Review of Human Trials Regarding Dosage, Efficacy, and Safety of CBD in Adults

Reviewed: Larsen C, Shahinas J. Dosage, efficacy and safety of cannabidiol administration in adults: A systematic review of human trials. J Clin Med Res. March 2020;12(3):129-141. doi: 10.14740/jocmr4090.

By Mariann Garner-Wizard

Study Details: At a Glance

Study Design

Systematic review

Included Studies

25 studies (N = 927), including 22 controlled clinical trials and three observational studies


Various CBD formulations


Placebo or active control (amisulpride in one study)


The authors declared no conflicts of interest.


Cannabidiol (CBD), a psychoactive but non-intoxicating cannabinoid in cannabis (Cannabis sativa, Cannabaceae), has promise in somatic and psychiatric disorders. Cannabis research and clinical use have been limited by concerns about its effects on the central nervous system (CNS), abuse potential, and legal status. CBD is well tolerated, has low abuse potential, and is being studied actively. In vitro and in vivo studies suggest that it has analgesic, anti-epileptic, anti-inflammatory, anti-psychotic, and immunomodulatory effects.

Despite regulatory issues, CBD is sold in many products that are used for various indications. It is unclear what formulations and doses should be used in trials. However, dosages in many over-the-counter (OTC) products (typically 20-25 mg) are far less than dosages used in clinical trials. This systematic review summarizes evidence on CBD administration, dosage, efficacy, and safety in humans.

A search of electronic databases revealed, after removal of duplicates, 362 studies in English published from 2000 to November 5, 2019. Included studies had at least 10 participants each and evaluated CBD alone. Pediatric trials were excluded, which eliminated many studies on epilepsy.

Of the studies retrieved, 25 were included in the review. Of these, 22 were controlled clinical trials and three were observational studies. Of the controlled trials, 20 were randomized (RCTs) and two were not. Of RCTs, 14 were individually randomized parallel trials, five were individually randomized crossover trials, and one was a cluster randomized crossover trial. Studies were conducted in five nations, mostly the United Kingdom. The risk of bias in RCTs was assessed via the Cochrane tool. Heterogeneity, assessed by grouping studies by indication, outcomes, and mode of administration, was too great for data to be pooled. Grading of Recommendations Assessment, Development, and Evaluation (GRADE) criteria were used to rate quality of evidence for given outcomes.

Most studies used pure CBD, one used cannabis seed CBD oil, and one used a CBD-rich botanical extract. Most used oral capsules. Vaporization and sublingual oil also were used. Placebo was the most-used comparator, and only one study had an active control. Among the RCTs, three had low risk of bias, eight had high risk, and nine had uncertain risk. Major potential sources of bias were in randomization and selection of reported results. Some studies did not report outcomes specified in their methods or changed the main outcome after registering the trial.

Cannabis sativa ©2021 Steven FosterAnxiety, assessed in 11 studies (358 participants), was the main outcome in eight studies. Two studies were limited to patients with anxiety disorders, and one was limited to those with non-clinical paranoia. Four individually randomized parallel controlled trials studied CBD’s effects on subjective anxiety. In two of those trials, people with social anxiety disorder (SAD) taking 600 mg and healthy individuals taking 300 mg reported reduced anxiety as assessed by a visual analog scale (VAS) vs. placebo. In a third trial, 150, 600, and 900 mg doses failed to affect anxiety as assessed physiologically. The fourth trial found that 600 mg CBD reduced anxiety according to the Beck Anxiety Inventory compared to placebo. Two crossover RCTs using 400 and 600 mg CBD reported reduced anxiety and increased mental sedation. Two non-RCTs gave 600 mg to healthy individuals. One study found no difference in anxiety and the other found less anxiety compared to placebo. In two studies, 16 or 32 mg CBD, administered in one study via inhalation, enhanced fear extinction (a reduction of previously acquired conditioned fear responses). The evidence supporting the anxiolytic effect of acute CBD intake is of low quality according to GRADE criteria.

Four parallel-group RCTs (196 participants) assessed the effects of CBD on cognitive impairment and psychotic symptoms in people diagnosed with schizophrenia. CBD at 600-1,000 mg per day for up to six weeks was compared to placebo and active controls. One RCT found no differences between CBD and placebo at six weeks. One found that CBD was as effective as amisulpride at four weeks in improving symptoms, with “marked tolerability and safety.” One RCT found 1,000 mg per day CBD an effective adjunct to antipsychotic drugs, significantly improving symptoms and clinical impressions. In a study of single doses (300 or 600 mg), no improvement was seen in selective attention on the Stroop Color and Word Test. Evidence of benefits for this indication is of moderate quality by GRADE criteria.

Three studies assessed the use of CBD in substance use disorders. Single oral doses (200, 400, and 800 mg) of CBD did not reduce reinforcing effects of cannabis smoking. Both inhalation of CBD for one week and single oral doses (800 mg) had possible benefits in reducing tobacco (Nicotiana tabacum, Solanaceae) use. In single RCTs in type 2 diabetes, Crohn’s disease, and ulcerative colitis, no benefits were seen. Single arm trials assessed oral CBD as an adjunct to anti-epileptic treatment and reported fewer seizures (P = 0.01); a topical cream for serious skin inflammation, with good results; and sublingual oil drops as an adjunct treatment for adverse effects (AEs) of human papilloma virus vaccine, again with positive results.

CBD AEs, reported in 10 studies, were generally mild and brief. Higher-quality and consistently reported RCTs, as well as more study of single cannabinoids vs. whole-plant extracts and of different modes of administration, are needed.

Image credit:

Cannabis sativa. ©2021 Steven Foster