Introduction
Perhaps one of the most notable changes in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) was the recognition of the possibility of mixed symptoms in major depression and related disorders (MDD). 1 For the first time in American Psychiatric Association nosology, the addition of a mixed features specifier to MDD represents a structural bridge between bipolar and major depression disorders. While debate continues as to the optimal definition for the mixed features specifier, there has been increased recognition of the possibility of a mix of hypomania and depressive symptoms in someone whom has never experienced hypomania or mania.
The historical context for the development of the concept of mixed states, as well as differences in approach, particularly in the European psychiatric literature, is an important background to our consideration of the “with mixed features” diagnosis. The evolving concept of mixed symptoms in the setting of MDD, and also whether these symptoms should be viewed as a harbinger to the development of bipolar disorder, have yet to be fully explored. The implications of this are clinically significant for treatment and management.
Historical Notes
Historically, there has been a long tradition of discussion and recognition of what may be broadly termed “mixed states.” Marenos and Angst have written important overviews of the development of the concept prior to the modern era, in particular detailing with both the ancient recognition and the gradual refinement in definition through the work of Kraepelin and others in the 19th century.Reference Marneros 2 , Reference Angst and Marneros 3
In the older literature, the concepts of melancholia and mania were clearly recognized in the 1800s when mixed states was conceived as “mid-forms” or “mixtures of exaltation and depression.”Reference Marneros 2 There was then greater focus on this clinical presentation toward the end of the 19th century. In this period, recognition developed of the range of disease states termed manic depression illness (which included recurrent unipolar illness), which was naturalistically observed by such individuals as Kraepelin and Weygandt.Reference Marneros 2 , Reference Angst and Marneros 3 The 6 proposed mixed states as Kraepelin defined them are well-known: depressive or anxious mania, excited depression, mania with thought poverty, mania with stupor, depression with flight of ideas, and inhibited mania.
While there was a limited focus on mixed states in the mid-20th century, perhaps in part due to greater focus on psychoanalytic versus biologic considerations, work by Himmelhoch and others in the 1970s and 1980s increased awareness that mania and depression did not exist in only polar isolation, and raised the importance of assessing mixed states not only as an important modifier of disease course and treatment response, but as a distinct subtype of bipolar disorder.Reference Himmelhoch, Mulla, Neil, Detre and Kupfer 4 , Reference Akiskal, Rosenthal, Rosenthal, Kashgarian, Khani and Puzantian 5 While the DSM-5 recently modified the definition of mixed states, it was recognized early on that the DSM-IV definition of a mixed state, as only when a full episode of mania and depression co-occurred, had limitations that did not adequately capture clinical experience.Reference McElroy, Keck, Pope, Hudson, Faedda and Swann 6 Debate continues on the best definition of mixed states, as well as which definitions should rule the day and are the most clinically relevant.Reference Faedda, Marangoni and Rginaldi 7 – Reference Angst, Cui and Swendsen 10
DSM-5 and Differential Diagnosis
While DSM-IV made advances in characterizing the mood disorder range of presentation, including bipolar II disorder for example, the definition of mixed states was limited to full mixed episodes, requiring a full manic episode and a full depressive episode to be simultaneously present. There was no consideration of the possibility of more complex and less well-delineated presentations, such as mixed hypomania or mixed depression in bipolar disorder—presentations which are quite familiar to clinicians but are not formally codified in the nomenclature. No consideration was formally given to the possibility of MDD mixed depression, though the European literature has already included discussion of these more complex presentations.Reference Kukopulos, Caliari and Tundo 11 – Reference Akiskal, Benazzi, Perugi and Rihmer 13
More recent publications both before and during the development of DSM-5 provided a range of evidence supporting the more frequent presentation of mixed symptoms over pure euphoria mania in bipolar disorder and mixed symptoms during depression in patients with bipolar disorder.Reference Suppes, Mintz and McElroy 14 – Reference Miller, Suppes and Mintz 16 These and other studies clarified not only that euphoric presentations may be the minority, but also that in bipolar disorder a mixed presentation may herald a more severe course of illness.Reference Angst, Azorin and Bowden 17
During the development of DSM-5, the consideration was raised that mixed symptoms might also be present during major depressive episodes in those with MDD but lacking any history of hypomanic or manic episodes. Active discussion led to new analyses of older data sets, such as the National Institute of Mental Health (NIMH) Depression Collaborative to identify those with MDD and some degree of hypomanic symptoms.Reference Fiedorowicz, Endicott, Leon, Solomon, Keller and Coryell 18 These large data sets, in concert with new European datasets including the Bipolar Disorder: Improving Diagnosis, Guidance and Education (BRIDGE) study, supported the possibility of MDD depression episodes presented with mixed symptoms. Further studies as well as the BRIDGE study, such as the National Comorbidity Study - Replication (NCS-R) and the Munich study, and a reanalysis of baseline NIMH Depression Collaborative Study data, opened the way to consideration of mixed symptoms during a MDD episode of depression.Reference Angst, Cui and Swendsen 10 , Reference Angst, Azorin and Bowden 17 – Reference Zimmermann, Brückl and Nocon 19 The Munich study provided strong evidence in a somewhat younger population that subthreshold hypomanic symptoms were present in a substantial number of MDD episodes without the subjects necessarily converting to a full bipolar diagnosis over a 10-year observation period.Reference Zimmermann, Brückl and Nocon 19 In this longitudinal study, depending on how symptom count was done, based on 1 hypomanic symptom, 40% of the population observed from an earlier age reported MDEs with some degree of mixity. While re-examination of the NIMH Depression Collaborative found that those subjects presenting at baseline with a depression episode and 3 or more hypomanic symptoms had a significant likelihood of converting to bipolar disorder, evidence also suggested that not all subjects with a degree of subthreshold hypomania went on to develop bipolar disorder.Reference Fiedorowicz, Endicott, Leon, Solomon, Keller and Coryell 18 Both the NCS-R, an epidemiologic replication study, and the multinational BRIDGE study demonstrated a broad range of presentations, including subthreshold bipolarity among many patients with MDD, as well as frank bipolar disorder often missed.Reference Angst, Cui and Swendsen 10 , Reference Angst, Azorin and Bowden 17 It should be noted these studies specifically tested a hypomanic checklist in consideration of the idea that MDD and BD exist on a continuum.
The confluence of these findings suggests that the clinical features that distinguish bipolar disorder from “pure” MDD are very similar to the distinguishing features for those patients with MDD and a depressive episode with mixed features. Cross-cultural studies reinforce findings of increased suicidality, increased family history of bipolar disorder, and earlier onset of illness noted for those with MDD in a depressive episode with mixed features versus a depressive episode without mixed features.Reference Liu and Jiang 20 Despite the different definitions of mixed features, these cross-cultural and cross-national findings overall, given the numbers of patients described, draw a picture of a potential intermediate phenotype. This phenotype has yet to be fully defined, given that a significant percentage fitting this profile may develop bipolar disorder illness over time.
There has been much discussion differentiating MDD and bipolar disorder (BD), with mixed features patients particularly highlighted as those potentially on the road to developing BD. However, such features as early onset, increased recurrences, family history, and temperament have been suggested as distinguishing features for MDD with and without mixed features.Reference Perlis, Uher and Ostacher 21 , Reference Serra, Koukopoulos and De Chiara 22 In particular, the notion that MDD with mixed features may be less responsive to antidepressants has not been firmly established. In one retrospective study (STAR*D), those patients evidencing at least 2 mixed symptoms actually were more responsive to antidepressants.Reference Perlis, Cusin and Fava 23 This, however, continues to be an active area of debate as to whether patients exhibiting MDD with mixed features are more severely ill and less likely to be responsive to usual antidepressant treatments.Reference McIntyre, Soczynska and Cha 24 – Reference Balázs, Benazzi, Rihmer, Rihmer, Akiskal and Akiskal 26
Importantly, the clinical observation of increased suicidality with mixed symptoms during depressive episodes has been reported across a range of study designs.Reference Perugi, Angst and Azorin 9 , Reference Angst, Cui and Swendsen 10 , Reference McIntyre, Soczynska and Cha 24 , Reference Swann, Lafer and Perugi 27 While increased suicidality is certainly not only observed during periods of mixed symptoms, given the periodic increased energy sometimes seen during mixed depression and the consequent lower threshold to action in many cases, caution and careful observation are warranted.
Diagnosis as Defined by DSM-5
In consideration of developing a specifier that would apply to patients diagnosed with either bipolar or major depression and related disorders, what to include or exclude was reviewed by the DSM-5 committee. In particular, in the interests of clarity, discussion included whether to use all of specific symptoms at either pole and to exclude only those symptoms that often overlap in depression, hypomania, and mania. These overlapping symptoms included irritability, agitation, or distractibility. Given that irritability is a common symptom regardless of the presence of mixed features, it was viewed as more nonspecific and therefore more difficult to ascertain if the irritability related to a mixed picture or simply one of illness generally. Similar discussion led to the removal of these overlapping symptoms. This conclusion was also reached in the recently released position paper on mixed states by the International Society of Bipolar Disorders.Reference Swann, Lafer and Perugi 27
While the DSM-5 committee decided to exclude overlapping symptoms from the definition of mixed specifiers, these excluded symptoms are important gateways to suggest to the clinician that further evaluation for mixed symptoms should be considered for both the diagnosis and treatment choices.Reference Angst, Cui and Swendsen 10 , Reference Miller, Suppes and Mintz 16 , Reference Angst, Azorin and Bowden 17 , Reference Koukopoulos and Sani 28 – Reference Benazzi, Koukopoulous and Akiskal 30 In particular, while these symptoms are present at both poles and across mood disorders in general, they serve as red flags to look deeper; recent analyses suggest that when other mixed symptoms are present, these 3 excluded symptoms, particularly irritability and agitation, are likely to be present.Reference Miller, Suppes and Mintz 16 , Reference Benazzi 31
The current criteria for major depressive episode with mixed features are defined as meeting full episode criteria for a depression episode, and in the last 2 weeks some degree of the following symptoms at least every day and for a noticeable amount of time, though which symptoms present can vary. Three of the following symptoms are required:
-
∙ Elevated or expansive mood
-
∙ Inflated self-esteem or grandiosity
-
∙ More talkative than usual or pressure to keep talking
-
∙ Flight of ideas or subjective sense that thoughts are racing (This symptom must be distinguished from anxiety and anxious thoughts.)
-
∙ Increase in energy or goal-directed activity
-
∙ Increase or excessive involvement in activities that have a high potential for painful consequences
-
∙ Decreased need for sleep
Additional factors would include that the mixed symptoms are observable to others and represent a change from usual behavior, and that mixed symptoms are not better explained by a medical condition or substance use. (For this second qualifying condition, one example could be depression with mixed symptoms attributable to a recent period of using cocaine.)
There is debate as to whether 2 versus 3 symptoms is the correct threshold to consider for meeting the mixed features specifier. This issue was considered in the development of the DSM-5 Mixed Features Specifier, and it was decided that in the absence of more extensive prospective data, a conservative decision of requiring 3 non-overlapping symptoms for the specifier should be made. 1 , Reference McElroy, Keck, Pope, Hudson, Faedda and Swann 6 , Reference Benazzi and Akiskal 12 , Reference Koukopoulos and Sani 28 There is a range of views on this topic; some argue that the bipolar spectrum supports considering 1 symptom during depression as consistent with a mixed profile, though there is concern for the precision and limits of our measurement ability.Reference Faedda, Marangoni and Rginaldi 7 , Reference Perugi, Angst and Azorin 9 , Reference Angst, Cui and Swendsen 10 , Reference Angst, Azorin and Bowden 17 The requirement of 3 symptoms of a non-overlapping nature may indeed be too stringent and bias the likelihood of patients meeting this specifier condition with MDD more likely to be those who later go on to develop a full hypomanic or manic episode, thus changing the diagnosis to bipolar disorder.Reference Fiedorowicz, Endicott, Leon, Solomon, Keller and Coryell 18 This debate speaks to the limits of our notions around mixity, and further prospective, biologic, and clinical treatment data are needed to fully inform this debate.
What Do Treatment Reports Suggest About Mixed Depression?
At this time there are no medications approved by the US Food and Drug Administration (FDA) for treatment of MDD depression episode with mixed features. There is some concern as to the specificity of this diagnosis, and further studies will be needed to resolve this issue. The question of whether the medications that patients respond to during a MDE with mixed features are different than those for a MDE without mixed features will await future studies. Such studies are needed to directly address both the question of specificity of this subcategory of MDD and to posit a possible approach to distinguishing differences from BD II or “other specified.” While some studies suggest no negative impact of antidepressants on those with MDE with mixed features, others suggest the opposite. In particular, the question remains whether greater treatment resistance is conferred by mixed features as defined by DSM-5.Reference Faedda, Marangoni and Rginaldi 7 – Reference Perugi, Angst and Azorin 9 , Reference Angst, Azorin and Bowden 17 , Reference Perlis, Cusin and Fava 23 , Reference Balázs, Benazzi, Rihmer, Rihmer, Akiskal and Akiskal 26
There is a range of definitions of mixed depression, including the elements that were viewed as “overlapping” in DSM-5, but that are viewed as essential in the conceptualization and assessment of frequency during any mixed depression.Reference Sani, Napoletano and Vohringer 8 , Reference Akiskal, Benazzi, Perugi and Rihmer 13 , Reference Smith, Forty and Russell 25 , Reference Suppes, Silva and Cucchiaro 29 , Reference Benazzi, Koukopoulous and Akiskal 30 , Reference Benazzi and Akiskal 32 It is of interest to consider the 1 study to date that examined the treatment response of patients with MDD currently in an episode with mixed features.Reference Suppes, Silva and Cucchiaro 29
In this study of a newer atypical antipsychotic, lurasidone, for MDD with mixed symptoms, inclusion criteria were modified to allow entry and randomization in this monotherapy, placebo-controlled study based on 2 or 3 symptoms of the mixed specifier, though overlapping symptoms of irritability, distractibility, and agitation were excluded.Reference Suppes, Silva and Cucchiaro 29 (The study was designed and undertaken prior to the final version of DSM-5.) The study found benefit for lurasidone versus placebo for the defined group over the 6-week study; the authors report on which symptoms were most frequently observed. About two-thirds of patients met the entry criteria of 2 mixed symptoms, while the remainder otherwise met DSM-5 criteria for the mixed feature specifier. Of all patients (n=209), about 67% reported flight of ideas, 61% pressured speech, 41% decreased need for sleep, 28% increased energy or activity, 18% elevated or expansive mood periodically, 16% increase in impulsive behavior, and 7% inflated self-esteem or grandiosity. By comparison for overlapping symptoms, 57% reported irritability, 59% distractibility, and 37% agitation.
This study of a newer atypical antipsychotic provides the most thorough prospective treatment response data to date of subjects who approximate the mixed specifier for MDE in MDD. Of particular note is that the overlapping symptoms excluded from the mixed features specifier in DSM-5 occur almost as often as the 2 most frequent symptoms included in DSM-5: flight of ideas and pressured speech. Thus, while excluded to bring greater clarity to the newly proposed specifier bridging MDD and BD, it is important to note that these symptoms may be key to alerting the clinician and the patient that further treatment and assessment are needed beyond the usual approach to a major depressive episode.
Alternative Considerations for Diagnosis
What are our assumptions about mood disorder spectrum, and should we consider whether it really is a spectrum? Developing biologic data suggest that our categories may not reflect what the brain is experiencing in terms of how symptoms are clumping together in comparison to the relatively arbitrary categories imposed by DSM-5 and other similar systems of diagnosis.Reference Williams 34 We hope that the developing biological literature will inform the discussions on mood spectrum from the purely observational approach currently utilized to address these questions. 1
One of our usual assumptions diagnostically is that the development of bipolar disorder is a one-way road. The implication of this is that once a patient has experienced a hypomanic (or manic) episode, the lifetime diagnosis changes to bipolar II disorder regardless of future course of illness, or similarly once a patient has a manic episode, his or her diagnosis changes to bipolar I and never falls back to MDD or BDII regardless of future course. The limitation to this one-way road is that our approach to treatment changes from “treatment for MDD” to “treatment for BD.” It has been observed that patients experiencing MDD who have a MDE with mixed features respond poorly to antidepressants, though more formal testing of this hypothesis is needed. On the other end of this spectrum, no one would advocate more monotherapy antidepressants in BDI (although it is unclear what the right treatment is, for example, for a currently depressed 60-year-old patient whose last manic episode was at age 22). What, however, is the correct approach to depression in BPII? There is indication that some patients with BDII with or without rapid cycling do nearly as well on antidepressants as on lithium.Reference Altshuler, Sugar and McElroy 35 –37 Do these results imply that one group of patients with mixed features with only to-date subsyndromal hypomania, such as those with MDD, may not do well on antidepressants, but those potentially seen as further along this one-way road in fact do fine on antidepressants? What would these results imply about the concept of the “spectrum” that is currently the predominant approach to mood disorders?
There are other unconsidered paradigms besides a simple spectrum or continuous gradient of severity. The assumption of a one-way road to modeling mood disorders probably should be questioned. If all those with MDD with mixed features do not in fact develop bipolar disorder over time, and some of these individuals do less well on antidepressants than some patients with bipolar II disorder, this argues for discrete symptoms making up a set of disorders versus a continuum. Rather than the ongoing assumption that the only paradigm to model bipolar and major depression and related disorders is a spectrum, it is to be hoped that the developing biologic information will more appropriately define whether this is a continuous or a discontinuous process requiring other approaches than a spectrum.
Conclusion
While the concept of mixed states is old, the specific criteria as set out in DSM-5 are new, and treatment of patients with MDD with mixed features is clearly lacking, so there is a critical need to study it as a construct. Even with such studies, critical questions may be left unanswered. The current data from epidemiological samples, observational samples, and subgroup analyses of treatment studies that may have included patients with mixed features only suggest that these patients have differential outcomes compared to patients without such features. These data do not explain whether prospective studies, randomized for mixed features, will show the same differential outcomes. Perhaps standard antidepressants alone would be effective for them; perhaps other treatments such as monotherapy or adjunctive antipsychotic medications, or even novel treatments such as ketamine, would be more effective.
In spite of bipolar disorders and depressive disorders being in different chapters in DSM-5, the mixed features specifier remains a natural bridge between them. Depression in bipolar II disorder is nearly as inadequately studied in a systematic way as is MDD with mixed features. However, it is likely that the (apparently low) risk of mania from antidepressants in bipolar II disorder will not be a barrier to the treatment of MDD with mixed features, since presumably many such patients have already safely been exposed to antidepressants in the dozens of existing studies of MDD.
Further consideration should be given to the nature of the presentations across mood disorders. Should we simply consider MDD with mixed features a presentation as inevitably on the path to developing bipolar disorder? Or is it in fact a legitimate and delineated category distinct from bipolar II and bipolar disorder other specified? Certainly a proportion of patients with MDD with mixed features will go on to develop full-criteria bipolar disorder (as will patients with MDD alone, as is well-described),Reference Fiedorowicz, Endicott, Leon, Solomon, Keller and Coryell 18 but only with prospective studies of patients with well-defined MDD with mixed features—both short- and intermediate-term in-treatment studies and longer-term observational ones—can this answer be truly clarified. Might there be future biomarkers, be they functional (as in neuroimaging or through cognitive testing), genetic (such as risk genes or epigenetic phenomena), or others, taken in combination with studies of treatment response, that can build a risk model for the identification of and treatment predictions for subgroups of patients with MDD? The assiduous use of the mixed features specifier as this is explored may provide an additional framework in which to find such answers.
