ABSTRACT
Little is known about the neurometabolic effects of cannabis use. Using meta-analytic modeling of proton magnetic resonance spectroscopy (1H-MRS) studies, this study aimed to assess the differences in brain metabolite levels associated with cannabis use (PROSPERO: CRD42020209890) to inform treatment development for cannabis use disorder (CUD). Hedge’s g with random-effects modeling was used, and heterogeneity and publication bias indices were assessed. A complete literature search was conducted, and 15 studies met the inclusion criteria (e.g., 1H-MRS, cannabis group compared to a control group, brain region-specific results, necessary data to complete modeling). There were 29 models across gray matter regions in the brain. All models had between 2 and 5 studies (k), indicating that results should be interpreted with caution due to the limited number of available studies. Compared to the control groups, the cannabis-using groups showed lower levels of GABA and N-acetylaspartate in the anterior cingulate cortex (k = 3); lower glutamate in the basal ganglia/striatum (k = 2); and lower glutamine and myo-inositol in the thalamus (k = 2; although the two effect sizes came from the same sample). This is the first meta-analysis to consolidate the extant 1H-MRS studies focused on the neurometabolic effects of cannabis. Despite the few studies available, the evidence suggests cannabis use may impact important neural processes, including glutamatergic and GABAergic functioning (glutamate, glutamine, and GABA), neural health (N-acetylaspartate), and glial functioning (myo-inositol). The findings should be interpreted with caution considering the small sample size; the inability to test the impact of demographic, substance use, and methodological factors; and the heterogeneity of studies. Understanding the neurobiological effects of cannabis may inspire novel pharmacotherapy and/or psychosocial interventions for CUD.