Clinical expression of bipolar disorder type I as a function of age and polarity at onset: convergent findings in samples from France and the United States

Background: The clinical presentation, course, and comorbidities of bipolar disorder type I are highly heterogeneous, and this variability remains poorly predictable. Certain onset characteristics (eg, age and polarity at onset) may delineate subgroups differing in clinical expression and outcome.

Method: We retrospectively investigated the association between both age and polarity at onset and the clinical characteristics of bipolar I disorder (DSM-IV) in 2 independent adult samples: 480 French patients assessed in 1992-2006 (patients had been recruited from 3 university-affiliated psychiatry departments) and 714 US patients assessed in 1991-2003 (data were extracted from the Bipolar Disorder Phenome Database).

Results: Polarity at onset correlated with subsequent predominance (P < .001). Most patients experienced a depressive onset (57.9% in France vs 71.0% in the United States; P < .001) associated with a higher density of depressive episodes, suicidal behavior, and alcohol misuse. A manic onset was associated with a higher density of manic episodes. Early onset was frequent in both countries (42% in France vs 68% in the United States; P < .001) and was associated with suicidal behavior and cannabis and cocaine/opiate misuse. Sensitivity for the prediction of clinical characteristics was 1%-35% for age at onset and 26%-47% for polarity at onset.

Conclusions: Onset characteristics are associated with subsequent predominant polarity, suicidal behavior, and substance misuse in bipolar I disorder. These findings may facilitate personalized treatment strategies based on type of onset and may also facilitate early focused strategies for preventing comorbidity. Given the relatively low sensitivity and specificity of these onset characteristics for predicting clinical variables, the relevance of age and polarity at onset as specifiers in nosographical classifications will require further studies. However, polarity at onset may be the more relevant specifier, with further investigation required for age at onset.