Alexithymia and outcome in psychotherapy

Grabe, H. J., Frommer, J., Ankerhold, A., Ulrich, C., Gröger, R., Franke, G. H., . . . Spitzer, C. (2008). Alexithymia and outcome in psychotherapy. Psychotherapy and Psychosomatics, 77, 189–194.


The construct of alexithymia focuses on difficulties in describing and expressing feelings, on the paucity of fantasies. Recent studies have associated alexithymia with dissociation, depression, anxiety disorders, pathological gamblin and a broad range of psychopathologic features. Given the relative temporal stability, the pattern of correlations with traits of personality models like the NEO-FFI and the temperament and character model, alexithymia is considered to be a unique and distinct personality construct. However, there is an ongoing debate on the changeability of alexithymic traits by psychotherapy in the light of lacking absolute stability. The impact of alexithymia itself on outcome in psychotherapy is less clear. First, subjects with alexithymia are often socially avoidant, cold, less emotionally at- tached to others. This could lead to a reduced adherence to psychotherapy despite of severe mental distress. Second, the lack of imagination, psychological mindedness and awareness to emotional cues may significantly reduce the ability to be successfully engaged in psychotherapy. Third, early observations of Sifneos and others described alexithymic patients to respond poorly to dynamic psychotherapy. However, there has been little empirical research to investigate whether alexithymia predicts psychotherapy outcome. Some treatment studies found alexithymia to be associated with persistent somatization in somatoform disorders and with a negative outcome in med- ical treatment of functional gastrointestinal disorders. In short-term group therapy for outpatients with complicated grief and in short-term individual therapy for outpatients with mixed diagnoses, alexithymia predicted a negative outcome as well as in a naturalistic follow-up of outpatients with major depression. However, alexithymia did not interfere with the response to multimodal cognitive behavioral therapy in patients with obsessive-compulsive disorder.


We evaluated a large sample of inpatients undergoing intensive psychotherapeutic treatment to investigate the following hypotheses: Assuming higher levels of interpersonal stress and


The first hypothesis was not confirmed by our data. Patients who stopped treatment within the first four weeks were not more alexithymic than patients who continued the treatment program. Although unexpected, this finding is in line with one study that found alexithymia not to interfere with the compliance to psychotherapy in patients referred to a psychiatric consultationliaison service. Additionally, one experimental study provided evidence that verbalized empathic response from the physician may be especially crucial for the alexithymic patients’ postconsultation satisfaction and may thereby become the basis for a solid treatment alliance. The second hypothesis was fully confirmed by significantly higher levels of psychopathological distress in alexithymic patients at the beginning of the therapy. In contrast to our third hypothesis, the psychotherapeutic ‘high-care’ inpatient setting yielded a significant symptom reduction in alexithymics which was comparable to the relative symptom reduction in the nonalexithymic group. Still, the alexithymics had mean GSI scores at the end of the treatment that were almost identical to GSI scores of the nonalexithymic group in the beginning of the therapy. This corresponds to the finding of residual symptoms in depressed alexithymic patients after short-term psychotherapy. There were modest reductions of TAS-20 scores in the nonalexithymic group. Unexpectedly large reductions of TAS-20 scores were found in the baseline-alexithymic group, indicating a lack of absolute stability of alexithymia during treatment. In contrast to Rufer et al., all three TAS factors decreased significantly during the treatment. However, we found evidence for a high degree of relative stability of TAS-20 scores between t0 and t2 in the total sample which is in line with a large body of evidence. Only 13–16% of the variance in the changes of TAS-20 scores was explained by the changes in GSI scores from baseline to t2. Therefore, besides the changes in psychopathological distress, other unmeasured or unknown factors contributed to the majority of changes in the TAS-20 scores. Acknowledging the significant decrease in TAS-20 scores and the robust symptom reduction of psycho- pathological distress (GSI) at the end of the treatment in the alexithymic group, we assume that the ‘high-care’ in-patient setting was very effective in improving the identification, the differentiation and the verbalization of emotions and feelings. Future studies should investigate the efficacy of different treatments in alleviating alexithymia and should use a recently developed interview for the assessment of alexithymia. Prospective follow-up studies are required to evaluate the impact of persistent alexithymia and residual psychopathological symptoms at discharge on long-term outcome.


Hans Joergen Grabe MD.

Department of Psychiatry. Ernst-Moritz-Arndt University of Greifswald. HANSE-Klinikum Stralsund Rostocker Chaussee 70 D–18437 Stralsund (Germany)

Tel. +49 3831 45 2106.