Interactions of borderline personality disorder and anxiety disorders over 10 years.

OBJECTIVE: This report examines the relationship of DSM-IV borderline personality disorder (BPD) to anxiety disorders using data on the reciprocal effects of improvement or worsening of BPD and anxiety disorders over the course of 10 years.

METHOD: We reliably and prospectively assessed borderline patients (n = 164) with DSM-IV-defined co-occurring generalized anxiety disorder (GAD; n = 42), panic disorder with agoraphobia (n = 39), panic disorder without agoraphobia (n= 36), social phobia (n = 48), obsessive-compulsive disorder (OCD; n = 36), and posttraumatic stress disorder (PTSD; n = 88) annually over a period of 10 years between 1997 and 2009. We used proportional hazards regression analyses to assess the effects of monthly improvement or worsening of BPD and anxiety disorders on each other's remission and relapse the following month.

RESULTS: BPD improvement significantly predicted remission of GAD (hazard ratio [HR] = 0.65, P <.05) and PTSD (HR = 0.57, P < .05), whereas BPD worsening significantly predicted social phobia relapse (HR = 1.87, P < .05). The course of anxiety disorders did not predict BPD remission or relapse, except that worsening PTSD significantly predicted BPD relapse (HR = 1.90, P < .05).

CONCLUSION: BPD negatively affects the course of GAD, social phobia, and PTSD. In contrast, the anxiety disorders, aside from PTSD, had little effect on BPD course. For GAD and social phobia, whose course BPD unidirectionally influences, we suggest prioritizing treatment for BPD, whereas BPD should be treated concurrently with panic disorders, OCD, or PTSD. We discuss state/trait issues in the context of our findings.