• major depression
• comorbidity
• BRIDGE
• bipolar disorder
• bipolar spectrum
• bipolarity specifier
• borderline personality disorder

Objectives: The nature and the relevance of the association between borderline personality disorder (BPD) and bipolar disorder (BD) are still unclear and their relationship remains controversial. The aim of this study was, firstly, to explore the relationship between BPD and bipolarity, according to different criteria set, in a world-wide sample of patients with a current major depressive episode (MDE) and, secondly, to characterize the clinical profile of the subgroup of probands affected by both BD and BPD in comorbidity.
Methods: The sample comprised 5635 in and out-patients with an MDE enrolled in the multicenter, multinational, transcultural BRIDGE study, designed to assess varying definition of hypo/mania and familial and clinical variables associated with bipolarity. Patients with (BPD+) and without (BPD-) comorbid BPD were compared on socio-demographic, familial and clinical characteristics. Afterwards, 2658 subjects diagnosed with BD according to Bipolarity Specifier criteria were selected from the general sample, and the two groups of Bipolar Specifier patients with (BipSp/BPD+) and without (Bip/BPD-) comorbid BPD were compared on diagnostic, socio-demographic, familial and clinical characteristics.
Results: 532 patients (9.3%) met criteria for BPD in the general sample. A diagnosis of bipolar disorder was significantly more frequent in the BPD+ than in BPD- group using either DSM-IVTR-modified criteria or the bipolar specifier. BPD+ patients were significantly younger than BPD- depressives in regard to age, age at onset of first psychiatric symptoms and age at first diagnosis of depression. They also showed significantly more hypomania/mania in first-degree relatives in comparison to BPD- as well as more comorbid Alcohol and Substance abuse, Anxiety disorders, Eating Disorder and Attention Deficit Hyperactivity Disorder. BPD+ depressive patients showed significantly more psychotic symptoms, mixed states, atypical features, seasonality of mood episodes and post-partum depression than BPD- depressives. The two groups were also significantly different with respect to history of suicide attempts age at first psychiatric symptoms and depression <20 years, prior mood episodes >2 and >4, antidepressants induced hypo/manic switches, mood lability and irritability, and resistance to antidepressant treatments. All these clinical features were more common in BPD+ than in BPD- depressive patients. As regards the second part of the study, patients fulfilling Bipolarity Specifier criteria were selected from the ovarall sample, and 386 probands (14.5%) belonging to this group met criteria for BPD as well. A diagnosis of BD according to DSM-IV criteria was significantly more frequent in the BipSp/BPD- than in BipSp/BPD+, while similar rates in the two groups occurred using DSM-IV-Modified criteria. A subset of the BD criteria with an atypical connotation, such as irritability, mood instability and reactivity to drugs are significantly related to the presence of BPD in Bipolar specifier patients. BipSp/BPD+ patients were significantly younger than BipSp/BPD- patients in regard to age, age at onset of first psychiatric symptoms and age at first diagnosis of depression. They also showed significantly more comorbid Alcohol and Substance abuse, Anxiety disorders, Eating Disorder and Attention Deficit Hyperactivity Disorder. Among patients selected according to Bipolarity specifier, those who presented comorbid BPD, in comparison with BPD- patients, showed significantly more psychotic symptoms, history of suicide attempts, mixed states, mood reactivity, atypical features, seasonality of mood episodes, antidepressants induced mood lability and irritability, and resistance to antidepressant treatments.
Limitations: Centers not randomly selected, recall bias, cross-sectional design, and limited training of participating psychiatrists.
Conclusions: The study confirms, in a very large world-wide sample, the high prevalence of bipolarity in BPD depressive patients. The strong connection between BPD and BD is confirmed by the significant association with familial and clinical features classically considered as external validators of bipolarity. Furthermore, we confirm the high prevalence of patients that meet DSM-IV criteria for BPD in the subgroup of BD patients with MDE extracted from the original sample. Further prospective researches should clarify whether the mood reactivity and instability captured by BPD DSM-IV criteria are distinguishable from the subjective mood experienced of an instable, irritable, dysphoric manic/hypomanic/mixed state or simply represent a phenotypic variant related to developmental factors that might differentiate BD individuals.