Parker examined whether borderline personality disorder (BPD) is a bipolar or unipolar mood condition and concluded by suggesting that it is probably neither.^{Reference Parker1} I would like to offer a supplementary interpretation of the literature; that is BPD is in large part a mood disorder but is not necessarily a bipolar or unipolar mood variant.

Borderline personality disorder is highly comorbid with bipolar disorder^{Reference Mantere, Melartin, Suominen, Rytsälä, Valtonen and Arvilommi2} and depression,^{Reference Grant, Chou, Goldstein, Huang, Stinson and Saha3} and those who develop bipolar disorder have early temperamental markers of emotional dysregulation.^{Reference Luby and Navsaria4} Support that BPD is a mood disorder is also aligned with the fact that affective instability is a core feature of the syndrome. While under-investigated, there is emerging evidence that affect or mood instability, as opposed to mood episodes, might be the core feature of bipolar disorders.^{Reference Strejilevich, Martino, Murru, Teitelbaum, Fassi and Marengo5} The majority of patients with established bipolar disorder, even after symptomatic control continue to experience daily or weekly mood swings.^{Reference Bonsall, Wallace-Hadrill, Geddes, Goodwin and Holmes6} Further, the prevalence of mood instability and cyclothymic temperament is increased in unaffected bipolar probands^{Reference Diler, Birmaher, Axelson, Obreja, Monk and Hickey7} and it predicts functioning in those with bipolar disorder.^{Reference Strejilevich, Martino, Murru, Teitelbaum, Fassi and Marengo5} Mood instability is highly prevalent in unipolar depression^{Reference Marwaha, Parsons, Flanagan and Broome8} and independently links to suicidality and health-service use. Furthermore in BPD, affective instability is the least stable of the ‘trait-like’ features of the syndrome over 2 years.^{Reference Chanen, Jackson, McGorry, Allot, Clarkson and Yuen9} Thus, all three disorders share mood instability as a clinical component and this all points to BPD, at least in part, being a disorder of mood.

However BPD does not exactly fit into the bipolar or depressive affect rubric, given that the affective shifts do not last long enough for either diagnosis. Detailed studies of the nature of affective instability in mood disorders and BPD using the same measurement methods are limited. However, as Parker states, there are differences. Those with bipolar disorder have greater levels of euthymia-elation and affect intensity. In BPD there are more shifts between anxiety, depression and euthymia-anger.^{Reference Reich, Zanarini and Fitzmaurice10} Negative emotionality is a critical feature of BPD but it is changeable, as is obvious to clinicians who have been charged with the care of people with BPD on in-patient wards.

Affect can be studied on the basis of intensity, frequency of shift, rapidity of rise-times and return to baseline, reactivity to psychosocial cues or whether endogenously driven, and the extent to which there is overdramatic expression.^{Reference Koenigsberg11} To this could be added valence. Using this framework, BPD could be conceptualised as a disorder of mood in which affect changes are intense, frequent, rapid to occur, slow to dissipate and in which the valence of the mood state is typically negative incorporating depression, anxiety and anger. This pattern of difficulties although related to mood, do not appear to overlap to a significant extent with how depression or bipolar disorder might be described using the same affective framework. Though it is clear that terms such as ‘intensity’, ‘frequency’ and ‘rapidity of rise’ need to be better specified, experience-sampling methods analysing affective patterns in the three disorders might further illuminate this area and indeed the debate.