Objective: The objective of this study was to test the hypothesis that the polarity of the first mood episode may be a marker for suicidal behavior, particularly the violent subtype.
Method: One hundred and sixty-eight patients diagnosed with bipolar disorder (DSM-IV) were grouped according to type of first episode: depression or manic/hypomanic. Groups were compared for demographic and clinical variables. We performed logistic regression in order to test the association between first episode polarity and suicidal behavior.
Results: We found that depressed patients have a lifetime history of more suicide attempts. However, univariate analysis of number of suicide attempts showed that the best model fits the bipolar II subtype (mean square = 15.022; p = 0.010) and lifetime history of psychotic episodes (mean square = 17.359; p = 0.021). Subgrouping the suicide attempts by subtype (violent or non-violent) revealed that manic/hypomanic patients had a greater tendency toward attempting violent suicide (21.2 vs. 14.7%, X(2) = 7.028, p = 0.03). Multiple logistic regression analysis confirmed this result.
Conclusion: Depressed patients had more suicide attempts over time, which could be explained by the higher prevalence of bipolar II subtype in this group, whereas manic/hypomanic patients had a lifelong history of more frequent violent suicide attempts, not explained by any of the variables studied. Our results support the evidence that non-violent suicide attempters and violent suicide attempters tend to belong to different phenotypic groups.