No Bipolar II–for me,or for you!

We were delighted to read commentary (Fletcher et al., 2018) in the journal regarding bipolar II disorder and, in particular, the observation that it has been used as a catch-all term for all manner of diagnoses. There is no doubt that this has hindered, to some extent, a deeper understanding of bipolar II, but we feel there is little to be gained by further examining this ‘disorder’, largely because it has become evident that the basis upon which it was constructed was wholly arbitrary (Malhi et al., 2016) and that, in effect, it simply serves to sample a loosely categorised sub-population within a spectrum (Phillips and Kupfer, 2013). We therefore disagree with the authors of this article (Fletcher et al., 2018) that further research is warranted in attempting to identify the neurobiological underpinnings of bipolar II, given that the primary problem of poor delineation, both phenomenologically and within the clinical arena, remains. Indeed, we believe this would be counterproductive as, no doubt, there will be differences of sorts that can be identified by comparing any two populations along a spectrum, particularly if that spectrum is multi-dimensional (e.g. cognition and activity alongside mood), but such differences have little clinical value and cannot serve to define a disorder.

When first defined, bipolar II had to be clearly separated from bipolar I, and this was putatively achieved by necessitating hospitalisation and a greater degree of impairment for bipolar I disorder to be assigned. In addition, any incidence of psychotic phenomena elevated the diagnosis to bipolar I disorder immediately. Alongside these differences, it was assumed that the same symptoms occurred in both hypomania and mania and that, to all intents and purposes, bipolar II was simply a less severe variant of bipolar I disorder. At the same time, to ensure that bipolar II did not meld into normal perturbations of health, a relatively high bar was set for the duration of criteria for hypomania (namely, 4 days).

In the four decades since its inception, there have naturally been many studies that have examined and compared attributes of bipolar I and II disorders. Interestingly, for bipolar II disorder in which there was, by definition, a bias towards depression, there is a higher likelihood of suicide, which is unsurprising given that this is most likely to occur in the depressive phase of the illness. But the preponderance of depression in bipolar II has been interpreted as warranting it equivalence in terms of impact on quality of life, because it confers the same degree of disability as bipolar I and a comparable, if not greater, risk of suicide. These and other such findings have meant that bipolar II has been increasingly viewed as a distinct disorder, and reified to some extent as a disease, potentially with its own separate pathophysiology. But, there is scant evidence to support any such view.

Clearly, patients currently labelled with bipolar II disorder experience significant impairment and do require targeted treatment; however, bipolar II cannot always be reliably differentiated from bipolar I in clinical management. Indeed, the diagnosis of bipolar II rarely governs treatment and the same treatment options (e.g. mood stabilisers, second-generation antipsychotics, antidepressants) are used across both bipolar I and II diagnoses (Malhi et al., 2016). Since guiding treatment is one of the main purposes of giving a diagnosis, this in itself demonstrates that bipolar II has failed to achieve the primary requirement of a diagnostic label. Rather, the spectrum approach to bipolar disorder (advocated by the authors themselves) should range from unipolar depression to bipolar I disorder, and the arbitrary stepping stone of bipolar II should be dispensed with.

Therefore, instead of further pursuing the quest for pristine cleavage between bipolar I and bipolar II, efforts should be focused on defining the more practical and meaningful cut-off between normalcy and the syndrome of manic symptoms, which most likely occurs at a duration of two days (Angst et al., 2012), and identifying symptoms, if any exist, that are characteristic of mania and bipolar depression. In addition, bipolar disorder research should move to examining the domains of cognition and activity (Ghaemi and Dalley, 2014), alongside that of mood, and adopt a multi-dimensional approach to understanding not only bipolar disorder but mood disorders as a whole. Increasingly, it is accepted that the mood disorders exist along a spectrum, which is mostly demarcated by boundaries that are ephemeral and varying, depending on context. A good example of this is the transition of major depression itself into bipolar disorder through the occurrence of manic symptoms and/or its transmogrification into a mixed state. These are the key and important questions that face the field at present and warrant our utmost attention. The debate as to whether bipolar II disorder exists, whether it has a separate and distinct substratum, and whether it should be better clinically defined, is well and truly over.

Final thought:

Hooray, hurrah, we are changing our mind,

it’s all bipolar, no subtype or kind.

It’s less confusing,

so let’s start using,

simply ‘bipolar’, and see what we find!