Method: Fourteen adolescents diagnosed with bipolar disorder (mean age = 15.7; sd = 1.48; 11 females) participated in a 12-month open pilot study to assess the feasibility and efficacy of an adaptation of DBT (Miller et al., 1997; Miller, 1999). Eligible adolescents met DSM-V criteria for bipolar disorder via semi-structured interview. The DBT intervention consisted of individual therapy, phone coaching, and family skills training designed to reduce core symptoms of bipolar disorder such as impulsivity and affect lability. Over 12 months, youth were assessed at baseline, 3-, 6-, 9-, and 12 months using self-report measures including the Children's Affective Lability Scale- Child report (CALSC; Gerson et al., 1996) and the Borderline Personality Inventory (BPI; Chabrol et al., 2004; Leichsenring, 1999).
The CALSC includes 20 items reflecting self-rated mood lability using a 5-point likert scale ranging from ‘0 never to ‘5 more than once a day'. A total score reflecting affective lability and two subscales measured disinhibition and angry/depressed mood were used. The BPI includes 53 items rated “true or false” that correspond to the presence of symptoms including affect instability and impulsivity. Total number of items endorsed were calculate for the overall score and six subscales - affectivity, impulsivity, dissociative/psychotic, interpersonal instability, narcissism, and substance abuse.
Key Results: Repeated measures analysis of variance revealed significant reductions in both impulsivity and affective lability across the duration of the study (baseline to 12 months). These included significant reductions in disinhibition, F(1, 4) = 5.47, p < .01; anger/depression, F (1, 4) = 3.67, p < .05; and total affective lability, F(1, 4) = 5.80, p < .05. Effects were also consistent on the BPI measure with significant reductions in impulsivity, F(1, 4) = 8.34, p < .05; affectivity, F (1, 4) = 7.11, p < .05; and total borderline symptomatology, F(1, 4) = 17.39, p < .001), along with significant reductions in dissociation/psychotic symptoms, F(1, 4) = 7.68, p < .05. Post-hoc comparisons revealed significantly lower levels of symptom endorsement from baseline to post-treatment for all significant findings. No significant effects were observed for BPI interpersonal instability, narcissism, and substance abuse.
Conclusions: These findings demonstrate that a family–based adaptation of DBT for adolescents with bipolar disorder is feasible and also effective in reducing key symptoms of bipolar disorder including impulsivity and affective lability. Based on these promising feasibility findings, future research is needed utilizing rigorous randomized controlled designs, as well as implementation in community-based settings.