Rückfallprävention bipolarer Störungen: ein explorativer, clusteranalytischer Ansatz bei einer randomisierten, kontrollierten Psychotherapiestudie

The aim of this study was to differentiate between types of bipolar disorders and the associated features using explorative analysis. The focus was particularly on the role of bipolar 1 and bipolar 2 disorders as well as the influence of prophylactic interventions for relapse in a randomized, controlled treatment study. A total of 274 of the 305 originally included persons could be investigated in the study. Patients participated in either cognitive behavioral group therapy (SEKT) or supportive, patient-centered group therapy (FEST). Treatment took place over 4 days separated by a 1-month interval (equivalent to 16 double hours). Depressive and manic symptoms were assessed using the longitudinal interval follow-up evaluation (LIFE). The symptoms were retrospectively assessed for the previous 6 months, with respect to each week before and after the intervention phase and for 6‑month and 12-month follow-ups. The results show that the effects of both group therapies were comparable; however, there were statistically significant differences in a multivariate proportional hazards model for the factors bipolar 1 and 2 as well as the interaction of therapy with bipolar 1 and 2. In particular, bipolar 2 patients benefited significantly less from the SEKT intervention than from the FEST intervention. There were three clusters identified that separated bipolar 1 (SEKT, no comorbidity, predominantly no recurrences, younger patients), from bipolar 2 (FEST, no comorbidity, at least 1 often 2 recurrences, older patients) and from a heterogeneous group (SEKT and FEST, comorbidity). The distinction between bipolar 1 and bipolar 2 disorder is important and has so far not received sufficient attention. Bipolar 2 disorders generally have a worse course and respond particularly poorly to cognitive behavioral therapy (SEKT). An open, unstructured, supportive, patient-centered psychotherapy (FEST) is generally effective.