• mixed states
• major depression
• borderline personality disorder

Objectives: Studies that have considered the relationship between mood disorders and borderline personality disorder (BPD) have reached different and sometimes contrasting conclusions. The aim of the present study was firstly, to estimate the prevalence of mixed features in a large sample of major depressive patients with and without comorbid BPD and, secondly, to compare the clinical profile of major depressive patients with comorbid BPD with and without mixed features.
Method: The sample comprised 2811 patients with a Major Depressive Episode (MDE), enrolled in the multicenter, multinational, transcultural BRIDGE-II study, designed to assess mixed symptoms in patients with a MDE. Patients with (BPD+) and without (BPD-) comorbid BPD were compared on socio-demographic, familial and clinical characteristics. Afterwards, 187 subjects diagnosed with BPD according to DSM-IV criteria were selected from the general sample, and the two groups of BPD depressive patients with (MXS+) and without (MXS-) mixed features according to DSM-5 criteria were compared on diagnostic, socio-demographic, familial and clinical characteristics.
Results. 187 (6.7%) MDE patients of the present sample met criteria for BPD. Using DSM-IV-R criteria, the diagnosis of Bipolar Disorder (BD) was significantly more frequent in BPD+ than in BPD- group. BPD+ were more likely to show DSM-5-MXS and atypical features than BPD-. In comparison with BPD-, BPD+ reported significantly lower mean age and mean age of onset. BPD+ patients were significantly more likely to be female than were BPD- patients. All the mental disorders explored were more frequently present in BPD+ than in BPD- patients. BPD+ patients reported higher rates of positive family history for BD, past history of hypo/manic switches with antidepressants, onset of first psychiatric symptoms and first MDE before 30 years of age, duration of the current episode <1 month, number of mood episodes >3, and history of suicide attempts. BPD+ patients were significantly more likely to be currently treated with mood stabilizers and antipsychotics and to be taking more than 3 drugs. The rates of all (hypo)manic symptoms explored during the current MDE were significantly higher in BPD+ than BPD-. The multiple logistic regression demonstrated a significant association of comorbid BPD with the presence of mixed features such as psychomotor agitation, emotional lability, aggression, impulsivity and risky behaviours. The multivariate logistic regression demonstrated specific associations of BPD comorbidity with the presence of SQ5, SQ10, SQ12, SQ15, SQ22, SQ27, SQ29, and the absence of SQ8 at the Self-Questionnaire for mixed symptoms. Among the 187 patients with BPD 27.8% (n=52) were diagnosed as DSM-5-MXS. BPD MXS+ patients reported similar rates of mean age, gender distribution, BD, atypical features, lifetime comorbidity (except for Anorexia Nervosa), family history for BD,and personal history of suicide attempts in comparison with MXS-. Mean age at onset of first psychiatric symptoms was lower in BPD MXS+ patients than in MXS-. In comparison with MXS-, BPD MXS+ reported higher rates of past history of hypo/manic switches with antidepressants, onset of first psychiatric symptoms and first MDE before 30 years of age, duration of the current episode <1 month, and number of mood episodes > 3. Regarding their current drug treatment, BPD MXS+ patients were more frequently treated with mood stabilizers and antipsychotics and more frequently taking 3 or more drugs in comparison with MXS-. All the (hypo)manic symptoms observed during the current MDE were significantly higher in BPD patients with mixed features than those without. Among the 9 DSM-5 criteria for BPD, the presence of criteria 2 (instable interpersonal relationships), 8 (inappropriate and intense anger) and 9 (transient stress related paranoid ideation or dissociative symptoms), and the absence of criterion 6 (affective instability) increased the odds of the DSM-5 diagnosis of BPD+ MDE with mixed features. SQ7, SQ9, SQ10 SQ14, SQ15, SQ22, SQ31, increased the odds of the DSM-5 diagnosis of BPD+ MDE with mixed features.
Limitations: Centers not randomly selected, recall bias, cross-sectional design, and limited training of participating psychiatrists.
Conclusion: The results provides evidence of the high prevalence of mixed features in a large sample of MDE patients with comorbid BPD. These results are consistent with the view that a subset of current BPD criteria identifies a mixed presentation, at least in a substantial minority of patients with mood disorder. The study also confirm that features traditionally considered as reliable external validators of bipolarity such as early age at onset, highly recurrent course, (hypo)manic switches with antidepressant treatments and first-degree family history of BD are highly represented in major depressive patients with comorbid BPD. Finally, our results indicate that the presence of mixed features according to DSM-5 defintion in patients with MDE and BPD are associated with complex course of illness, reduced treatment response and worse outcome.