Borderline personality disorder among individuals with bipolar disorder: Prevalence, correlates, and experiences of the comorbid diagnosis.

Background: Borderline personality disorder (BPD) is prevalent among individuals with bipolar disorder and is associated with severe illness course. Research comparing the prevalence and correlates of BPD between bipolar I and II disorders is limited and no existing research explores the experience of living with both diagnoses.
Aim. To examine the prevalence, clinical correlates, and experiences of BPD among individuals with bipolar I or II disorder in the UK.
Methods: Lifetime prevalence of BPD diagnosis was examined among 1157 individuals with best-estimate main lifetime DSM-IV diagnosis of bipolar I (n=808) or II (n=349) disorder from the UK Bipolar Disorder Research Network (BDRN). Association of BPD with sociodemographic and lifetime clinical variables, including history of childhood abuse, was examined in bipolar I and II disorder. Thematic analysis was conducted on interviews with 15 participants reporting both diagnoses.
Results: 16.4% (95% CI 14.4-18.6%) reported BPD. BPD was significantly more common in bipolar II disorder (24.4%; 95% CI 20.1-29.1%) than bipolar I disorder (13.0%, 95% CI 10.8-15.4%) (p<.001; OR 2.16). Bipolar disorder was diagnosed first in most participants (59.4%, 95% CI 51.1-67.4%) and more participants found BPD unhelpful to their treatment (45.2%, 95% CI 37.0-53.6%) than helpful (27.4%, 95% CI 20.6-35.5%). Significant predictors of BPD in bipolar I disorder included history of suicide attempt (OR 1.86), presence of an anxiety disorder (OR 2.18) and heavier alcohol use (OR 1.01); in bipolar II disorder, the significant predictor was heavier alcohol use (OR 1.01). Themes from interviews highlighted differences in how participants viewed the two diagnoses: “an unknown diagnosis – a known diagnosis”; “feeling dismissed – feeling treated”; “keeping the diagnosis a secret – disclosing the diagnosis”; and “not identifying as BPD – identifying as bipolar disorder.”
Conclusion: A clinical diagnosis of BPD was common among those with bipolar disorder (one in six) and twice as likely in bipolar II disorder compared to bipolar I disorder. BPD was associated with more severe illness outcomes and other comorbidities and was perceived by the participants as more stigmatising than bipolar disorder and a barrier to treatment. Participants felt much better informed about bipolar disorder than BPD and felt better able to discuss bipolar disorder with others. Clinicians should consider screening for BPD particularly among individuals with bipolar II disorder and other comorbidities. Improvements to education and information about BPD may help reduce stigma and improve outcomes for those with both diagnoses.