RESUMO
OBJECTIVE: The aim of this study was to identify predisposing factors and clinical features at baseline that might help predict diagnosis of bipolar disorder vs schizophrenia in a first-episode psychosis (FEP) cohort. METHODS: In this prospective, naturalistic study, we evaluated a cohort of 335 subjects with FEP recruited from April 2009 to April 2012. Baseline features were compared between subjects with a final DSM-IV diagnosis of bipolar disorder and schizophrenia at 12-month follow-up. A binary logistic regression model was used to assess predictors of diagnosis of bipolar disorder at follow-up. RESULTS: At 12-month follow-up, 47 of the 335 subjects included in the study received the diagnosis of bipolar disorder and 105, of schizophrenia. Subjects with a final diagnosis of bipolar disorder had a higher prevalence of family history of mood disorders (38.2% vs 18.0%, P = .02), better baseline premorbid adjustment (Premorbid Adjustment Scale [PAS]: 38.4 vs 50.6, P < .01) and psychosocial functioning (Functional Assessment Short Test [FAST]: 23.6 vs 33.7, P = .001), better cognitive flexibility (number of perseverative errors on the Wisconsin Card Sorting Test [WCST]: 14.2 vs 19.7, P = .01), more manic symptoms (Young Mania Rating Scale [YMRS]: 14.1 vs 7.3, P < .01), lesser negative symptoms (Positive and Negative Syndrome Scale negative scale [PANSS-N]: 15.0 vs 22.3, P < .001), and shorter duration of untreated psychosis (144.2 vs 194.7 days, P < .01) than subjects with a schizophrenia diagnosis. Binary logistic regression model revealed that lower FAST scores (odds ratio [OR] = 0.956; P = .015), lower PANSS-N scores (OR = 0.93; P = .048), and lower number of perseverative errors on the WCST (OR = 0.946; P = .035) were significantly related to diagnosis of bipolar disorder at follow-up. CONCLUSIONS: In our FEP cohort, better psychosocial functioning, lesser negative symptoms, and better cognitive flexibility were related to diagnosis of bipolar disorder at 12-month follow-up.