Changes in personality functioning following psychotherapy: Utilizing machine learning to identify predictors in a psychodynamic psychotherapy sample.

INTRODUCTION Although personality functioning has a long psychodynamic tradition and has received renewed interest in psychotherapy research with the DSM-5 and ICD-11, little is known about its course and influencing factors following psychotherapy.

METHODS In a sample of 1208 completed outpatient psychodynamic psychotherapies, we examined changes in personality functioning in the 1-year follow-up. Machine learning was used to filter out potential predictors from all 227 possible predictors for changes in personality functioning following psychotherapy.

RESULTS For most patients, improvement in personality functioning remained stable following psychotherapy. Fourteen potential predictors were identified, including the length of psychotherapy, adverse childhood experiences, diagnosis of PTSD or recurrent depressive disorder, the improvement in interpersonal problems and symptoms as well as the stability of the bond between psychotherapist and patient. Furthermore, it was found that strong improvements can be considered a potential indicator for subsequent loss of personality functioning following psychotherapy.

DISCUSSION Our results suggest that in most cases improvement in personality functioning achieved during psychotherapy remains stable. In addition, the change in personality functioning following psychotherapy appears to be influenced by a multitude of small effects. It is then discussed that most of the identified predictors are associated with interpersonal relationships and experiences.