Cannabis Use Interacts With Psychosis Vulnerability in Borderline Personality Disorder: La consommation de cannabis interagit avec la vulnérabilité à la psychose dans le trouble de la personnalité limite
James Mirabelli, James MacKillop, David Streiner, Michele Korostil
The Canadian Journal of Psychiatry September 11, 2026 DOI: 10.1177/07067437261486962 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Cross-sectional clinical study with moderation analyses Peer reviewed |
|---|---|
| Sample size | 79 |
| Population | Outpatients with borderline personality disorder |
| Topics | Cannabis |
| Key findings | Cannabis use moderated the relationship between psychotic-like experiences and BPD symptom severity: the PLEs-BPD association strengthened as cannabis use increased, with the highest BPD symptoms among those high in both PLEs and cannabis use and the lowest among those low in PLEs and high in cannabis use. Among individuals with low PLEs, cannabis use was negatively associated with BPD symptoms. |
Abstract
Background: Individuals with borderline personality disorder (BPD) frequently use cannabis, often for self-medication. However, cannabis use has also been linked to BPD symptom severity, suggesting heterogeneity in its effects. Individuals with BPD frequently experience psychotic-like symptoms and may demonstrate elevated psychosis vulnerability, which is associated with greater symptom severity and poorer clinical outcomes. Cannabis use may exacerbate psychotic symptoms. The relationship between psychosis vulnerability and BPD symptoms may therefore be influenced by cannabis use. This study examined whether cannabis use interacts with psychotic-like experiences (PLEs) in relation to BPD symptom severity in a clinical sample.
Method: Seventy-nine outpatients with BPD completed questionnaires assessing BPD symptom severity, cannabis use, and PLEs. Moderation analyses tested whether cannabis use influenced the association between PLEs and BPD symptoms.
Results: Cannabis use significantly moderated the relationship between PLEs and BPD symptom severity (ΔR 2 = 0.069, P = 0.006), such that the association between PLEs and BPD symptoms strengthened as cannabis use increased. A crossover interaction was observed: the lowest BPD symptom scores occurred in individuals with low PLEs and high cannabis use, whereas the highest BPD symptom scores occurred in those with high PLEs and high cannabis use. In addition, PLEs moderated the relationship between cannabis use and BPD symptoms, such that cannabis use was negatively associated with BPD symptoms among individuals with low PLEs ( β = −0.358, P = 0.018).
Conclusions: PLEs appear to be an important factor in understanding heterogeneous associations between cannabis use and BPD symptom severity. Individuals with BPD who endorse elevated PLEs may benefit from targeted psychoeducation regarding cannabis-related risks and consideration of treatment for cannabis use disorder when clinically indicated. Conversely, the observed negative association between cannabis use and BPD symptoms in individuals with lower PLEs warrants further study of clinically meaningful heterogeneity and individualized treatment approaches.