Psychiatric symptoms caused by cannabis constituents: a systematic review and meta-analysis

Abstract

Background: Approximately 188 million people use cannabis yearly worldwide, and it has recently been legalised in 11 US states, Canada, and Uruguay for recreational use. The potential for increased cannabis use highlights the need to better understand its risks, including the acute induction of psychotic and other psychiatric symptoms. We aimed to investigate the effect of the cannabis constituent Δ⁹-tetrahydrocannabinol (THC) alone and in combination with cannabidiol (CBD) compared with placebo on psychiatric symptoms in healthy people.

Methods: In this systematic review and meta-analysis, we searched MEDLINE, Embase, and PsycINFO for studies published in English between database inception and May 21, 2019, with a within-person, crossover design. Inclusion criteria were studies reporting symptoms using psychiatric scales (the Brief Psychiatric Rating Scale [BPRS] and the Positive and Negative Syndrome Scale [PANSS]) following the acute administration of intravenous, oral, or nasal THC, CBD, and placebo in healthy participants, and presenting data that allowed calculation of standardised mean change (SMC) scores for positive (including delusions and hallucinations), negative (such as blunted affect and amotivation), and general (including depression and anxiety) symptoms. We did a random-effects meta-analysis to assess the main outcomes of the effect sizes for total, positive, and negative PANSS and BPRS scores measured in healthy participants following THC administration versus placebo. Because the number of studies to do a meta-analysis on CBD's moderating effects was insufficient, this outcome was only systematically reviewed. This study is registered with PROSPERO, CRD42019136674.

Findings: 15 eligible studies involving the acute administration of THC and four studies on CBD plus THC administration were identified. Compared with placebo, THC significantly increased total symptom severity with a large effect size (assessed in nine studies, with ten independent samples, involving 196 participants: SMC 1·10 [95% CI 0·92-1·28], p<0·0001); positive symptom severity (assessed in 14 studies, with 15 independent samples, involving 324 participants: SMC 0·91 [95% CI 0·68-1·14], p<0·0001); and negative symptom severity with a large effect size (assessed in 12 studies, with 13 independent samples, involving 267 participants: SMC 0·78 [95% CI 0·59-0·97], p<0·0001). In the systematic review, of the four studies evaluating CBD's effects on THC-induced symptoms, only one identified a significant reduction in symptoms.

Interpretation: A single THC administration induces psychotic, negative, and other psychiatric symptoms with large effect sizes. There is no consistent evidence that CBD induces symptoms or moderates the effects of THC. These findings highlight the potential risks associated with the use of cannabis and other cannabinoids that contain THC for recreational or therapeutic purposes.

Funding: UK Medical Research Council, Maudsley Charity, Brain and Behavior Research Foundation, Wellcome Trust, and the UK National Institute for Health Research.

Figure 1

Study selection BPRS=Brief Psychiatric Rating Scale. CBD=cannabidiol. PANSS=Positive and Negative Syndrome Scale. THC=Δ⁹-tetrahydrocannabinol.

Figure 2

Forest plot of total psychiatric symptom severity following THC relative to placebo The size of the squares reflects the weight attributed to each study. Exact study weights are presented in the appendix (p 13). The diamond denotes the summary effect size for the random-effects model for all studies, and the width of the diamond depicts the overall 95% CI. THC=Δ⁹-tetrahydrocannabinol. *Low cannabis use sample. †High cannabis use sample.

Figure 3

Forest plot of positive symptom severity following THC relative to placebo The size of the squares reflects the weight attributed to each study. Exact study weights are presented in the appendix (p 13). The diamond denotes the summary effect size for the random-effects model for all studies, and the width of the diamond depicts the overall 95% CI. THC=Δ⁹-tetrahydrocannabinol. *Low cannabis use sample. †High cannabis use sample.

Figure 4

Forest plot of negative symptom severity following THC relative to placebo The size of the squares reflects the weight attributed to each study. Exact study weights are presented in the appendix (p 13). The diamond denotes the summary effect size for the random-effects model for all studies, and the width of the diamond depicts the overall 95% CI. THC=Δ⁹-tetrahydrocannabinol. *Low cannabis use sample. †High cannabis use sample.

Figure 5

Forest plot of general psychiatric symptom severity following THC relative to placebo The size of the squares reflects the weight attributed to each study. Exact study weights are presented in the appendix (p 13). The diamond denotes the summary effect size for the random-effects model for all studies, and the width of the diamond depicts the overall 95% CI. THC=Δ⁹-tetrahydrocannabinol. *Low cannabis use sample. †High cannabis use sample.