Convergence of Cannabis and Psychosis on the Dopamine System

Abstract

ImportanceDespite evidence that individuals with a cannabis use disorder (CUD) are at elevated risk of psychosis and that the neurotransmitter dopamine has a role in psychosis, the mechanism linking cannabis use and psychosis remains unclear.ObjectiveTo use neuromelanin-sensitive magnetic resonance imaging (MRI), referred to as the neuromelanin-MRI signal, a practical, proxy measure of dopamine function, to assess whether a common alteration in the dopamine system may be implicated in CUD and psychosis and whether this alteration can be observed in those with a CUD whether or not they have a diagnosis of first-episode schizophrenia (FES).Design, Setting, and ParticipantsThis longitudinal observational cohort study recruited individuals from 2019 to 2023 from an early psychosis service and the surrounding communities in London, Ontario. The sample included individuals with and without CUD, with some in each group also diagnosed with FES.ExposuresFES and CUD diagnoses from the Structured Clinical Interview for DSM-5.Main Outcomes and MeasuresNeuromelanin-MRI signals within the midbrain (substantia nigra [SN]/ventral tegmental area [VTA]) including a subregion previously linked to the severity of untreated psychosis (a priori region of interest). Linear mixed-effects analyses were performed relating neuromelanin-MRI signals to clinical measures.ResultsA total of 36 individuals without CUD (mean [SD] age, 22.3 [3.2] years; 29 male [81%]; 12 with FES) and 25 individuals with CUD (mean [SD] age, 24.3 [4.7] years; 22 male [88%]; 16 with FES) participated in the study. One-year follow-up was completed for 12 individuals with CUD and 25 without CUD. CUD was associated with elevated neuromelanin-MRI signal in a set of ventral SN/VTA voxels (387 of 2060 SN/VTA voxels, corrected P = .03, permutation test). CUD was also associated with elevated neuromelanin-MRI signal in the psychosis-related region of interest (t92 = 2.12, P = .04) with a significant dose-dependent association (higher burden of CUD symptoms associated with higher neuromelanin-MRI signal, F1, 96 = 4.89; P = .03). In contrast, participants with FES did not exhibit a significant elevation in neuromelanin-MRI signal (241 SN/VTA voxels had elevated signal, corrected P = .09). There was no association between time and neuromelanin-MRI signal.Conclusions and RelevanceElevated dopamine function in a critical SN/VTA subregion may be associated with psychosis risk in people with CUD. Cannabis was associated with the hypothesized final common pathway for the clinical expression of psychotic symptoms.