A genome-wide association study of long-term cannabis users identified a significant signal at the CHRM3 gene linked to cannabis-induced hallucinations. The strongest association was found in European Americans, with the lead SNP rs115455482 reaching genome-wide significance. The risk allele was associated with lower CHRM3 expression in the thalamus, and CHRM3 was co-expressed with three psychosis risk genes in brain tissues. Findings did not replicate in an independent sample, though meta-analysis strengthened the association. No significant signals were found in African Americans. The results suggest CHRM3 may contribute to cannabis-induced hallucinations and point to the thalamus's potential role.
After adjusting for preexisting psychiatric conditions, the link between hallucinogen use and psychosis disappears. Among 273,466 people with substance-related hospital admissions, psychosis diagnoses were more common after hallucinogen-related admissions (16.4%) than after other substance admissions (6.6%). However, once clinical characteristics were accounted for, the increased risk became nonsignificant (hazard ratio 0.97). This suggests that observed associations between hallucinogens and psychosis are largely due to underlying mental health vulnerabilities, not a direct causal effect. The findings inform psychedelic policy by indicating that population-level data on hallucinogen safety may reflect preexisting risk factors.