Importance: Cannabidiol (CBD) has antipsychotic effects in humans, but how these are mediated in the brain remains unclear.
Objective: To investigate the neurocognitive mechanisms that underlie the therapeutic effects of CBD in psychosis.
Design, setting, and participants: In this parallel-group, double-blind, placebo-controlled randomized clinical trial conducted at the South London and Maudsley NHS Foundation Trust in London, United Kingdom, 33 antipsychotic medication-naive participants at clinical high risk (CHR) of psychosis and 19 healthy control participants were studied. Data were collected from July 2013 to October 2016 and analyzed from November 2016 to October 2017.
Interventions: A total of 16 participants at CHR of psychosis received a single oral dose of 600 mg of CBD, and 17 participants at CHR received a placebo. Control participants were not given any drug. All participants were then studied using functional magnetic resonance imaging (fMRI) while performing a verbal learning task.
Main outcomes and measures: Brain activation during verbal encoding and recall, indexed using the blood oxygen level-dependent hemodynamic response fMRI signal.
Results: Of the 16 participants in the CBD group, 6 (38%) were female, and the mean (SD) age was 22.43 (4.95) years; of 17 in the placebo group, 10 (59%) were female, and the mean (SD) age was 25.35 (5.24) years; and of 19 in the control group, 8 (42%) were female, and the mean (SD) age was 23.89 (4.14) years. Brain activation (indexed using the median sum of squares ratio of the blood oxygen level-dependent hemodynamic response effects model component to the residual sum of squares) was analyzed in 15 participants in the CBD group, 16 in the placebo group, and 19 in the control group. Participants receiving placebo had reduced activation relative to controls in the right caudate during encoding (placebo: median, -0.027; interquartile range [IQR], -0.041 to -0.016; control: median, 0.020; IQR, -0.022 to 0.056; P < .001) and in the parahippocampal gyrus and midbrain during recall (placebo: median, 0.002; IQR, -0.016 to 0.010; control: median, 0.035; IQR, 0.015 to 0.039; P < .001). Within these 3 regions, activation in the CBD group was greater than in the placebo group but lower than in the control group (parahippocampal gyrus/midbrain: CBD: median, -0.013; IQR, -0.027 to 0.002; placebo: median, -0.007; IQR, -0.019 to 0.008; control: median, 0.034; IQR, 0.005 to 0.059); the level of activation in the CBD group was thus intermediate to that in the other 2 groups. There were no significant group differences in task performance.
Conclusions and relevance: Cannabidiol may partially normalize alterations in parahippocampal, striatal, and midbrain function associated with the CHR state. As these regions are critical to the pathophysiology of psychosis, the influence of CBD at these sites could underlie its therapeutic effects on psychotic symptoms.
Trial registration: isrctn.org Identifier: ISRCTN46322781.