Mixed depression: a clinical marker of bipolar-II disorder

Abstract

Background Recent studies have found that mixed depression [i.e., a major depressive episode (MDE) plus intra-MDE hypomanic symptoms] is common in bipolar-II disorder (BP-II), and not uncommon in major depressive disorder (MDD) depressed outpatients. Study aim was to test the predictive power for the diagnosis of BP-II of several dimensional definitions of mixed depression, searching for a clinical marker which could reduce the current underdiagnosis of BP-II. Methods Consecutive 348 BP-II and 254 MDD depressed outpatients were interviewed by the Structured Clinical Interview for DSM-IV, the Hypomania Interview Guide, and the Family History Screen, by a senior psychiatrist in a private practice. Intra-MDE hypomanic symptoms were systematically assessed. Mixed depression was defined as an MDE plus intra-MDE hypomanic symptoms. Results Dimensional definitions of mixed depression (at least 2, 3, 4, 5 or more intra-MDE hypomanic symptoms) were tested for predicting BP-II. A definition requiring 2 or more hypomanic symptoms had the highest sensitivity, the lowest specificity, and the lowest positive predictive value. A definition requiring 5 or more hypomanic symptoms had the highest specificity, the lowest sensitivity, and the highest positive predictive value. The most balanced combination of sensitivity and specificity was found for a definition requiring 3 or more hypomanic symptoms. This definition had the highest positive predictive value, and the highest ROC area (i.e., the best global performance). This definition had also the most balanced combination of sensitivity and specificity for predicting bipolar family history. In order to validate this definition as a clinical marker of BP-II, as bipolar validators were used BP-II, young onset, many recurrences, atypical depression features, and bipolar family history (the most important one). Univariate logistic regression found that this definition was associated with most bipolar validators, especially bipolar family history. Multiple logistic regression found that bipolar family history was its strongest predictor. Conclusions Findings suggest that a definition of mixed depression requiring 3 or more intra-MDE hypomanic symptoms may be a useful clinical marker for predicting the diagnosis of BP-II. Presence of mixed depression should lead to skillful probing for history of hypomania, which would probably reduce the BP-II misdiagnosed as MDD. Findings may also impact treatment of BP-II, as intra-MDE hypomanic symptoms may become more severe by antidepressants alone, and mood stabilising agents may be required before (or concurrently with) antidepressants.