Defining mixed depression

Abstract

Background Mixed depression, i.e. a major depressive episode plus co-occurring manic/hypomanic symptoms, has recently become the focus of research. However, it is still unclear if its definition should be based on specific manic symptoms or on a number/score of manic symptoms. Different definitions may have different diagnostic utility, such as treatment impacts. Study aim Study aim was to test which definition of mixed depression was more supported, by using, as diagnostic validator, early age at onset on the basis of previous studies supporting its bipolar nature. Methods Consecutive 336 Bipolar II Disorder (BP-II), and 224 Major Depressive Disorder (MDD) outpatients were cross-sectionally assessed for major depressive episode (MDE) and concurrent DSM-IV hypomanic symptoms when presenting for treatment of depression, by a mood disorder specialist psychiatrist using the Structured Clinical Interview for DSM-IV as modified by Akiskal and Benazzi (J Clin Psychiatry, 2005) and the Hypomania Interview Guide (HIG), in a private practice. Mixed depression was defined as co-occurrence of MDE and hypomanic symptoms. Early age at onset (EO) below 21 years was used as diagnostic validator. Results Multivariable logistic regression of EO versus all within-MDE hypomanic symptoms, controlled for BP-II, showed that no specific symptom was independently associated with EO. By ROC analysis versus EO, the best combination of sensitivity and specificity, and the highest figure of correctly classified, were shown by a cutoff number > = 3 symptoms, and by a cutoff HIG score > = 8. Both cutoffs had similar strength of association with EO. Mixed depression defined by > = 3 within-MDE hypomanic symptoms (A), or by a HIG score > = 8 (B), were more frequent in EO group versus LO group (A: 70.5% versus 49.8%; B: 60.7% versus 40.9%; p < 0.001), and in BP-II versus MDD (A: 72.3% versus 39.7%; p < 0.001; positive predictive value for BP-II = 73.1%; B: 63.9% versus 29.0%; p < 0.001; positive predictive value for BP-II = 76.7%). Discussion Findings could support the diagnostic validity of a definition of mixed depression based on a cutoff number/score of within-depression hypomanic symptoms versus one based on specific symptoms, complementing and supporting previous studies using bipolar family history as validator. Diagnosing mixed depression has treatment impacts, such as careful use of antidepressants added to mood stabilising agents or no use of antidepressants, as recently shown by large naturalistic and controlled studies.