Effects of cannabidiol on symptoms in people at clinical high risk for psychosis

Cannabidiol (CBD), a non-intoxicating constituent of cannabis, has potential anxiolytic and antipsychotic properties² and a good safety profile. In two out of three clinical trials in patients with established psychosis, evidence of its antipsychotic efficacy has been reported^3–5. However, there have not been trials of a period of treatment with CBD in CHR individuals. We assessed the clinical effects of a course of CBD treatment in people with a CHR state following a protocol approved by the National Research Ethics Service Committee London (Camberwell, St. Giles) (ISRCTN46322781).

The study was conducted on antipsychotic-naïve subjects attending early detection services in the UK who met one or more criteria for CHR state for psychosis: a) attenuated psychotic symptoms; b) brief limited intermittent psychosis (i.e., a psychotic episode lasting <1 week which remitted without treatment); c) recent functional decline and either schizotypal personality disorder or first-degree relative with psychosis. Key exclusion criteria were history of previous psychotic disorder or manic episode, neurological disorder, or current DSM-IV diagnosis of substance dependence.

Thirty-three subjects were recruited after they provided written informed consent. They were advised to refrain from using cannabis for 96 hours, alcohol for a minimum of 24 hours, nicotine for 6 hours, and any other recreational drugs for 2 weeks before entering the study, and to continue to refrain from using cannabis or other recreational drugs during the course of the study. Baseline assessments included the Comprehensive Assessment of At-Risk Mental States (CAARMS)⁶; the Spielberger State-Trait Anxiety Inventory, State Subscale (STAI-S)⁷; and the Positive and Negative Syndrome Scale (PANSS)⁸.

Using a parallel group, double-blind, placebo-controlled design, participants were randomly allocated to either CBD (N=16) or placebo (N=17). They received either a CBD capsule or an identical-looking placebo capsule as a single daily oral dose, which they continued for 21 days. The dose of CBD (99.9% pure) was 600 mg/day, found to be effective and well-tolerated previously^4, 9. All clinical assessments were repeated after 7 and 21 days of treatment, except for the CAARMS, which was administered at baseline and at the end of treatment. Blood samples were collected before and after taking the study drug on days 1 and 21 to assay CBD plasma levels. The effects of treatment on symptoms were examined using analyses of variance with treatment (CBD vs. placebo) as the between-subject factor after controlling for baseline scores.

At baseline, the two treatment groups were comparable in demographic and clinical variables (see supplementary information). None of the participants received any psychotropic medication other than CBD or placebo during the course of the study. Two participants dropped out from the placebo arm. Following 21-day treatment (intention-to-treat, last observation carried forward analysis), CBD-treated participants had a lower total CAARMS score (F_1,30=7.168, p=0.012) than those receiving placebo, after controlling for baseline score. There were no significant differences between the treatment groups in the incidence of treatment-emergent side effects (see also supplementary information).

The CBD group also reported less distress associated with psychotic symptoms (F_1,30=4.66, p=0.039) and had a lower PANSS total score (p=0.042), after controlling for the respective baseline values. There was a greater reduction in the CAARMS negative symptoms (p=0.045), but not in the CAARMS positive symptoms (p=0.144), in CBD-treated patients. State anxiety levels following treatment were not different between the two groups (p=0.862).

When the analyses were restricted to participants with complete data for the respective measures, the CBD-treated group again had a lower total CAARMS score (p=0.033), with a trend for less distress associated with psychotic symptoms (p=0.072) and a lower total PANSS score (p=0.056). There were no group differences in the mean number of pills missed (p=0.85) or the proportion of patients who missed at least one pill (p=1.00). CBD levels were detectable in all except one out of 15 CBD participants with available data (see also supplementary information).

These data provide the first evidence that CBD can ameliorate symptoms in the CHR population. We found that treatment with CBD for three weeks was accompanied by a reduction in the severity of  CHR symptoms and the distress associated with psychotic ex-periences. As we did not find an effect of CBD on the STAI-S score, these effects were unlikely to be driven by a reduction in state anxiety.

Recent meta-analyses suggest that existing pharmacological and psychological treatments have little effect on symptoms or the incidence of psychosis in CHR subjects¹, and there is currently no licensed treatment for this population. Consistent with previous evidence^3–5, the incidence of adverse effects in the CBD-treated group was not different from the placebo-treated group, making CBD a good candidate treatment for CHR subjects. No subject dropped out of the CBD arm.

A limitation of the study is the small sample size, underscoring the preliminary nature of the evidence. Because treatment was limited to 21 days, we were not able to examine the effect of treatment on the risk of later transition to psychosis. As some clinical outcome data were missing, we analyzed effects on clinical outcomes using a last observation carried forward method of imputation. However, when we subsequently repeated the analyses but restricted inclusion to participants with complete data for the respective measures, the main results remained unchanged.

Our findings indicate that short-term treatment with CBD can ameliorate the symptoms of CHR state for psychosis, and is well tolerated. These results highlight the potential of CBD as a novel treatment for psychosis, and the need for large-scale efficacy studies to further evaluate its clinical utility.