The relationship between depression and mania remains poorly understood and is responsible for much of the confusion about mixed states. The difficulty in conceptualizing opposite states such as euphoric and depressive moods during the same episode may account for the considerable differences in reported frequencies of mixed states, among acutely manic patients. It is possible that the fundamental mood characteristic of mania is not tonality of mood (e.g. euphoric, irritable or depressed mood), but rather the intensity of emotions.

We interviewed 30 patients hospitalized for a manic episode, asking about their symptoms during the episode, using the list of symptoms for manic and depressive episode of the DSM-IV criteria. Emotional hyper-reactivity, defined as an increase in the intensity of all emotions, was assessed using the Hardy Scale. Manic symptoms were also assessed by a clinician using the Beck–Rafaelsen Mania Scale.

This study showed that most of the manic episodes presented many dysphoric symptoms, more particularly depressive mood (33%), irritability (53%), anxiety (76%), and recurrent thoughts of death or suicidal ideation (33%). However, only 10% of our sample met the criteria for mixed state. The other symptoms reported by patients and included in the DSM-IV criteria for depressive mood are common between depressive and manic episodes. All patients (100%) reported that they felt all their emotions with an unusual intensity.

We suggest that the most appropriate way to define mood in manic states is as a function of intensity, and not as a function of tonality. This definition circumvents the arbitrary dichotomy between mania and mixed state. With this definition, manic episodes can be described as being more or less dysphoric, with the actual characteristics of dysphoria encompassing irritability, anxiety, or depressive affect. This point could be extremely helpful in discriminating mixed state or dysphoric mania from depression.

Draft DSM-5 criteria for a mixed major depressive episode have been proposed, but their predictive validity has not yet been established. We hypothesized that such symptoms would be associated with poorer antidepressant treatment outcomes.

We examined outcomes among individuals with major depressive disorder participating in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study, an effectiveness study conducted at primary and specialty care centers in the USA. Mixed features were derived from the six self-report items of the mania subscale of the Psychiatric Diagnosis Screening Questionnaire. Primary analyses examined the association between the presence of at least two of these in the 6 months before study entry, and remission across up to four sequential treatment trials, as well as adverse outcomes.

Of the 2397 subjects with a major depressive episode of at least 6 months' duration, 449 (18.7%) reported at least two mixed symptoms. The presence of such symptoms was associated with a greater likelihood of remission across up to four sequential treatments, which persisted after adjustment for potential confounding clinical and demographic variables (adjusted hazard ratio 1.16, 95% confidence interval 1.03–1.28). Two individual items, expansive mood and cheerfulness, were strongly associated with a greater likelihood of remission.

Proposed DSM-5 mixed state features were associated with a greater rather than a lesser likelihood of remission. While unexpected, this result suggests the potential utility of further investigation of depressive mixed states in major depression.