Predictors of treatment non-response for depression in a sample of female adolescents with risk-taking and self-injurious behavior - A one year follow-up

Objective: Depressive disorders (DD) are highly prevalent among adolescents. While up to 60 % do not respond to treatment, evidence on predictors of treatment non-response in this age group is mixed, impeding meaningful clinical implications. Drawing on a consecutive clinical cohort of adolescents with risk taking and self-injurious behavior, the present study aimed to identify predictors of treatment non-response for female DD in a naturalistic one year follow-up.

Methods: The sample comprised female adolescents with verified DD (n = 152). Patients underwent assessments at baseline (T0) and follow-up (T1). Sociodemographic factors (e.g., age), clinical measures (e.g., symptom severity, trauma), and treatment variables (e.g. number of psychotherapy sessions), were analyzed as potential predictors of treatment non-response in unadjusted analyses and analyses adjusting for treatment intensity and age at baseline. Treatment response was defined based on not fulfilling formal diagnosis for DD at follow-up (52.3 %; n = 80) or the 50 % decrease in self-reported depressive symptoms (21.1 %; n = 32) from T0 to T1.

Results: Greater depressive and overall symptom severity, greater frequency of self-injuries, history of suicide attempts and history of childhood trauma at T0 were robustly associated with treatment non-response based on diagnostic interviews. Only a lower number of siblings was robustly associated with treatment non-response based on self-reports.

Limitations: Findings may not generalize to other treatment settings.

Conclusion: Collectively, our results highlight overall symptom severity as significant predictor of treatment non-response in female adolescents with depression. Methodological differences (interviews versus self-reports) and potential implications from these findings for clinical practice are discussed.