Background: About 25% of all patients seeking psychotherapeutic treatment are considered to be alexithymic. Alexithymia has been assumed to be negatively associated with therapeutic outcome. On the other hand, it is unclear to which extent alexithymia itself may be modified by psychotherapeutic interventions. Methods: From 414 consecutively admitted inpatients, 297 were followed up after 4 weeks (t1) and after 8–12 weeks (t2) upon discharge. Patients were treated with psychodynamic group therapy in a naturalistic setting. The Toronto Alexithymia Scale (TAS-20) and the Symptom Checklist-90 were administered. Results: Twenty-seven percent of the patients were alexithymic (TAS-20 ≧61) at baseline. Multivariate models with repeated measurements indicated significant changes in Global Severity Index of the Symptom Checklist-90 in both alexithymic and nonalexithymic subjects. However, alexithymic subjects had significantly higher Global Severity Index scores than nonalexithymic subjects at t0, t1 and t2 (p < 0.001). The TAS-20 scores demonstrated a high relative stability in the total sample. However, in the alexithymic group, the TAS-20 scores changed considerably from baseline to discharge [66.3 (SD = 4.7) to 55.9 (SD = 9.9); t = 8.69; d.f. = 79; p < 0.001]. Conclusion: The inpatient treatment program including psychodynamic group therapy significantly reduced psychopathological distress and alexithymic features in alexithymic patients. Still, these patients suffered from higher psychopathological distress at discharge than nonalexithymics. Therefore, alexithymic features may negatively affect the long-term outcome.

The BSI-18, an abridged version of the Brief Symptom Inventory of Derogatis, contains the 3 six items scales Somatization, Depression, Anxiety, and the Global Score (GSI). In a sample of N=638 psychotherapeutic patients, reliability and validity were proven. Reliability of the 3 scales was good: Somatization α=0.79, Depression α=0.84, Anxiety α=0.84, and GSI α=0.91. The postulated three-factor structure was proven sufficiently using confirmatory and explorative factor analyses. The questionnaire separated different patients groups. Judgments of the therapists corresponded well with the self-rating behavior of the patients. In conclusion, the psychometric evaluation of the BSI-18 resulted in persuasive evidence for its reliability and validity. The loss of information, as a result of item reduction, is acceptable analyzing large samples; in cases of individual analyses, the SCL-90-R is advised.

No abstract