Tobacco use moderates associations between cannabis use, psychiatric symptoms, and cognitive performance in psychosis spectrum disorders.

Nelson A.; Srinivasan A.; Clementz B.; Gershon E.; Keshavan M.; McDowell J.; Pearlson G.; Tamminga C.; Brady R.; Yassin W. et al. · Schizophrenia research · 2026

**Background & Methods** This was a secondary cross-sectional analysis of data from the Bipolar-Schizophrenia Network on Intermediate Phenotypes 2 (B-SNIP 2), a large multi-site study examining psychosis spectrum disorders. The analysis included 884 individuals with psychotic disorders who underwent cognitive and clinical assessments, excluding those with substance abuse or dependence diagnoses in the preceding 30 days to 3 months (except tobacco use). **Key Findings** • Tobacco-only users demonstrated significantly greater symptom burden and worse cognitive impairment compared to other substance use patterns among individuals with psychosis, establishing tobacco use as a marker of disease severity. • Cannabis use demonstrated differential associations depending on tobacco co-use status: greater adult cannabis use was associated with worse general psychopathology among current tobacco users but with fewer general symptoms among never-tobacco users, indicating a moderating effect. • Cognitive performance showed complex patterns where increased cannabis use correlated with better cognitive performance in current tobacco users but demonstrated no significant association with working memory in past tobacco users. **Dosage & Administration** Not reported. The study examined patterns of tobacco and cannabis use (current, past, or never use) without specifying quantity, frequency, or route of administration. **Safety & Adverse Effects** Not reported as a primary focus. The study excluded individuals with active substance abuse or dependence diagnoses, which may have limited assessment of acute adverse effects. However, the findings implicitly suggest that tobacco use associates with greater clinical symptom burden and cognitive impairment in psychotic populations. **Evidence Quality** This secondary analysis represents moderate-quality evidence with several limitations. The cross-sectional design prevents causal inference regarding substance use and outcomes. The sample excluded individuals with recent substance dependence, potentially limiting generalizability to more severely affected populations. The study provides important descriptive data on co-use patterns but cannot establish whether tobacco/cannabis use causes worse outcomes or represents a marker of underlying disease severity. The multi-site design and large sample size (n=884) strengthen internal validity. Results require replication with longitudinal designs to clarify temporal relationships and potential confounding variables.

More recent cannabis research