Holding the treatment frame: borderline features, personality functioning, and psychotherapy dropout in a youth mental health service

Baseline personality pathology may shape both everyday functioning and engagement in psychotherapy. This study tested whether maladaptive personality trait domains and borderline symptom severity were associated with psychosocial functioning at intake and whether these variables were related to subsequent psychotherapy dropout within the same clinical pathway. Participants were 124 young adults (M age = 20.81, SD=2.30; 51.6% male and 48.4% female) assessed at intake in a youth mental health early intervention service in Milan, Italy (2024-2025). Measures included the Personality Inventory for the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (PID-5), the Borderline Symptom List-23 (BSL-23), and the Social and Occupational Functioning Assessment Scale (SOFAS). SOFAS was examined using stepwise multiple linear regression including PID-5 domains, BSL-23, sex, and age. Dropout was examined using Bayesian logistic regression and model comparison in the psychotherapy-only subsample (n=63; 13 dropouts, 20.6%). In the full sample, the final regression model retained only BSL-23, F(1, 122)=5.57, p=.020, R2=.04, indicating that higher borderline symptom severity was associated with lower functioning (B=-2.86, β=-.21, 95% CI [-5.27, -0.46]). For dropout, inclusion evidence was strongest for disinhibition (Bayes factor for inclusion [BF_inclusion] = 3.36), followed by BSL-23 (2.39) and SOFAS (2.23), although model-averaged 95% credible intervals (CrIs) included zero for all predictors. Clinically, these findings suggest that disinhibition and borderline severity may be better understood as markers of relational intensity requiring a structured yet reflective therapeutic stance rather than as straightforward dropout risks.