Temporal relationship of first-episode non-affective psychosis with cannabis use: A clinical verification of an epidemiological hypothesis

Background: We analyzed the association of age at onset of psychosis treatment (AOPT) with having a history of cannabis use in patients with a first episode of non-affective psychosis. We also investigated the impact on the AOPT of exposure to cannabis in adolescence, compared with young adulthood, and of the additional exposure to cocaine. Method: We recruited 112 consecutive patients (66 men and 46 women; age range, 18-57 years) with a first psychotic episode. The composite international diagnostic interview (CIDI) was used to assess drug use and to define the age at onset of heaviest use (AOHU) of a drug, defined as the age when drug was used the most for each patient. The effect of cannabis and cocaine AOHU on AOPT was explored through Kruskal-Wallis and Mann-Whitney tests, and logistic regression. Sex-adjusted cumulative hazard curves and Cox regression models were used to compare the AOPT of patients with and without a history of cannabis use, or associated cocaine use. Results: We found that the AOPT was significantly associated with the use of cannabis, independently of sex, use of cocaine, tobacco smoking or excessive alcohol consumption. There was a dose-response relationship between cannabis AOHU and AOPT: the earlier the AOHU the earlier the AOPT. Hazard curves showed that patients with a history of cannabis use had a higher hazard of having a first-episode psychosis than the rest of the patients (sex-adjusted log-rank X-2 = 23.43, df = 1, p < 0.001). Their respective median AOPT (25th, 75th percentiles) were 23.5 (21, 28) and 33.5 years (27, 45) (for log-transformed AOPT, t = 5.6, df = 110, p < 0.001). The sex-adjusted hazard ratio of psychosis onset comparing both groups was 2.66 (95% CI, 1.74-4.05). Conclusions: Our results are in favor of a catalytic role for cannabis use in the onset of psychosis. (C) 2009 Elsevier Ltd. All rights reserved.