REVIEW OF: McMain S et al, Psychother Psychosom 2022;91(6):382–397

STUDY TYPE: Randomized single-blind controlled trial

Most effective psychotherapies for borderline personality disorder (BPD) are long term, including dialectical behavior therapy (DBT). Many programs require patients to agree to a full year of DBT, which is a big commitment. If DBT could be provided in fewer sessions, we could help more patients with BPD—but would a briefer course of DBT be effective? This study attempted to answer this question by comparing six versus 12 months of DBT for patients with BPD.

The researchers randomized 240 patients with BPD from two sites in Canada to receive either six months of DBT (DBT-6) or the standard treatment of 12 months (DBT-12). Patients were mostly women (79%) with a mean age of 28 years and had to have a recent history of suicidal or nonsuicidal self-injurious episodes to be enrolled. The two therapies were identical except for length of treatment, with DBT-6 concluding after six months in half of the number of sessions. Both therapies involved the standard components of comprehensive DBT, including weekly individual therapy, weekly group skills training, 24/7 phone coaching, and consultation meetings for therapists. The primary outcome was total frequency of self-harm episodes during the previous three-month period, including suicidal and nonsuicidal self-injurious episodes, which was assessed by the Suicide Attempt Self-Injury Interview at baseline and every three months for up to 24 months. 

The two treatments did not differ significantly on the primary outcome of self-harm events. At the beginning of the study, the mean number of self-harm events per patient in the preceding three months was 7.39. By month 12 of treatment (end of treatment for DBT-12 and six months post-treatment for DBT-6), mean self-harm events in the preceding three months had dropped to 0.30 (DBT-12) and 0.26 (DBT-6) (no significant difference). Mean self-harm events continued to decrease for both groups and were 0.22 (DBT-12) and 0.10 (DBT-6) at month 24. Dropout rates were similar for both groups. Secondary analyses were also encouraging but somewhat mixed, with DBT-12 patients having more improvement in measures of interpersonal functioning and anger expression at month 24, but similar improvements in most other measures.

The study had several limitations. For one, there was no control arm such as a waitlist group to control for the passage of time and other independent variables. The authors considered including a control arm but ruled against it for ethical reasons. Additionally, concomitant treatments such as pharmacotherapy were not controlled for, and the primary outcome was subject to bias by relying on self-reporting.

CARLAT TAKE

A shorter, six-month course of DBT may be as effective as a full year in reducing self-harm in BPD.