Why the hype about subtype? Bipolar I,bipolar II – It’s simply bipolar,through and through!

Bipolar disorder (BD) is frequently misdiagnosed and mismanaged because the ‘concept’ of BD II lacks definition, and prognostic value.

Hypomania was first described and introduced into practice by Falret and Mendel in the 19th century. Despite long-standing recognition of the syndrome, a precise definition of hypomania has remained elusive. In 1976, upon observing less severe symptoms of mania that could be treated without hospitalization, Dunner et al. (1976) conceived BD II. Two decades later, BD II was formally installed in Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition (DSM-IV), with the diagnosis contingent on the occurrence of hypomania in the context of a previous major depressive episode.

In Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition (DSM-5), increased activity/energy was added as a core feature of mania/hypomania, alongside elevated or irritable mood. In hypomania, these symptoms must last for a minimum of 4 consecutive days, contrasting with the 7-day criterion for a manic episode. In addition, hypomania requires an unequivocal change in functioning that is out of character for the individual. An interesting and critical distinction with mania, however, is that this change in function, while observable to others, is not severe enough in hypomania to cause marked impairment, necessitate hospitalization or present with psychotic symptoms. Thus, DSM-5 defines and distinguishes hypomania and mania on the basis of duration of symptoms and severity of illness (associated impairment). However, in practice, it is extremely difficult to distinguish between no, some or marked impairment, which necessarily cascades as a function of severity. In sum, the criteria for hypomania are fuzzy and, clinically, its distinction serves no purpose other than perhaps to obscure the inception of mania.

‘The lines around the states’

The beginning of hypomania (the lower boundary)

In clinical practice, delineating hypomania as ‘abnormal’ is difficult because patients often cannot recall such experiences accurately, and may not register them as unusual, or be particularly concerned by them. On the contrary, when associated functioning is subjectively (and in some cases objectively) improved, hypomania may be experienced/recalled as a return to normality, or a feeling of performing ‘better than usual’. Therefore, in practice, an accurate appraisal of such periods is often only possible by carefully piecing together both the history of the illness and behavioural observations made by others, an exacting process with variable success.

Melding into mania (the upper boundary)

Differentiating the upper boundary of hypomania is equally problematic as it relies predominantly on severity and duration of symptoms. The boundary is therefore arbitrary and lacks correlation to illness course and therapeutic specificity.

A matter of how much and how long for?

Symptom severity

In DSM-5, the symptomatic criteria for hypomania and mania are essentially the same – with the exception of psychotic symptoms. Diagnostic separation of BD I and BD II is determined significantly by impact on social and occupational functioning – along with the peculiar criterion of hospitalization, which is usually determined by local practices and availability of beds rather than symptom severity per se, although it is indicated for the prevention of self-harm or harm to others. Thus, technically, the severity of functional impairment associated with manic symptoms is a key distinguishing feature that should reliably and satisfactorily separate BD I and BD II. Clinically, however, the bulk of functional impairment associated with BD II stems largely from its depressive phase and, during mania/hypomania functioning per se, is loosely calibrated and only given modest attention. Instead, the severity of manic symptoms is inferred predominantly from their duration, and the severity of an episode (mania/hypomania) is gauged largely on the basis of the number of manic symptoms, and the urgency with which treatment is sought. Consequently, findings from many studies have shown potential differences between BD I and BD II across a broad range of clinical characteristics, including patterns of illness, treatment responsiveness and suicidality. But, in actuality, none of these correlations with BD subtype are exclusive or have meaningful prognostic value because most of the associations largely reflect the severity differences that partition mania and hypomania. In other words if, by definition, BD II has less severe symptoms of mania, then this will be borne out when comparing samples that have been grouped accordingly. These types of linkages do not further ‘define’ bipolar subtypes or help to separate BD I from BD II. For example, Frankland et al. (2015) found that BD I depressive episodes are more likely to feature psychomotor retardation, while BD II depressions more commonly exhibit mixed features. While these syndromal tendencies were statistically significant, the authors note that they could not be used to reliably distinguish BD I and BD II.

Minimum duration threshold

Another key distinction in DSM-5 between BD I and II is the duration of symptoms needed to meet criteria for a manic or hypomanic episode – presently 7 and 4 days, respectively. In the transition from DSM-IV, the minimum duration criterion for hypomania did not change – a decision out of step with accumulating research findings. For instance, recent studies have shown that individuals with even brief periods of manic symptoms lasting 1–3 days separate from those with major depressive disorder (MDD) – supporting distinction based on even 1 day of manic symptoms (based upon data following diagnosis). However, a minimum duration of 2 days (as opposed to 4) is clinically valid (Angst et al., 2012) and applicable in practice (Benazzi and Akiskal, 2006).

Revision of the duration criteria, perhaps along these lines, was likely given some consideration when developing DSM-5 because of the introduction of ‘Short-duration hypomanic episodes (2-3 days) and major depressive episodes’. This category captures those individuals who fall short of the duration criteria for hypomania, as it stands. It also mirrors increasing clinical recognition that many depressed individuals experience ‘subthreshold bipolar symptoms’. Presently, and problematically, these individuals are ‘lumped’ in with other presentations, such as those who have an insufficient number of hypomanic symptoms in the context of previous depression, and those with short-duration cyclothymia (see Figure 1). This creates a mixed bag of overlapping presentations that are difficult to disentangle and explore, and for which there is no clear management strategy. One possible solution to reduce the misdiagnosis of these disorders is to lower the current duration criterion for mania to 2 days (effectively removing hypomania altogether) and, in doing so, to reclaim those shorter periods of manic symptomatology that ‘belong’ to mania (Angst et al., 2012; Benazzi and Akiskal, 2006). This will then leave a residual group of those with manic symptoms lasting less than 2 days. This group can then be investigated and characterized by drawing on longitudinal patterning of symptoms and illness course, treatment effects and interactions with personality. A further advantage of aligning the cut-off for mania with clinical experience is that clinicians are likely to apply the distinction more strictly because it has greater face validity. This will mean that the clinical category of mania will also be more representative – reducing misdiagnosis and increasing the likelihood of identifying neurobiological substrates.

Figure 1.

Reclassifying duration of mania.

Why reclassify?

In practice, it is important that our diagnoses accurately capture disorders, which in turn meaningfully inform management and ideally reflect underlying pathophysiology. Eliminating the upper boundary between hypomania and mania, and effectively removing BD II as a separate diagnostic entity, would provide consistency for clinicians in that this distinction rarely, if ever, governs treatment. Mood stabilizers, second-generation antipsychotics and antidepressant medications can be, and are, used liberally in both ‘kinds’ of BD. The myriad patterns of BD are also not captured by the pseudo-specificity of BD II, in which manic symptoms of varying intensity may last 4 or nearly 7 days with different symptomatic profiles. Instead, the focus should be on capturing the precise phenomenology and the duration and severity of specific symptoms so as to provide a detailed picture of mania as it evolves and dissolves both naturally and in response to treatment. Similarly, in the current classification, there are significant difficulties in defining the lower boundary of hypomania that invariably abuts subsyndromal manic symptoms and also encroaches on mixed manic features. Lowering the duration threshold while more strictly adhering to the criteria may provide clearer guidelines, particularly when considering differential diagnosis between BD and disorders with similar clinical features (e.g. Borderline Personality Disorder).

Thus, in our proposed classification (see Figure 1), mania, hypomania and short-duration hypomania are no longer distinct entities. Instead, ‘Mania’ is any period of significant manic symptoms that exceeds 2 days (i.e. lasting 2 full days or more). The symptoms used to define Mania remain the same but they need to be characterized carefully and specified exactly in terms of severity and duration. Clinically, the treatment options also remain the same, but the focus is shifted to more accurate description rather than indiscriminate grouping. From a research perspective, changing the criteria in this manner is likely to capture more accurately, and with much needed granularity, the variability of symptoms inherent in BD and to provide a better testing ground for the development of new treatments. Similarly, individuals falling below the 2-day duration threshold are likely to be a more homogeneous group, which will also facilitate investigation and advance our understanding of these presentations. For example, by removing the ill-defined ‘category’ of hypomania (BD II) and distilling manic symptoms from 2 days and more, research into presentations on both sides of the boundary can be better focused on the influence of continuous variables, such as the severity of individual symptoms on therapeutic response.

Finally, it is important to note that this is not a purely theoretical consideration and that phenomenological clarity is necessary if we are to advance our understanding of mood disorders and to develop effective treatments with specific actions. Hence, our proposal is that there is no bipolar II and that, instead, it is all bipolar, through and through.