Sociodemographic and clinical predictors of depression in children and adolescents at clinical high-risk for psychosis: Results of a two-year follow-up study

Depressive disorders are a main cause of disability already in children and adolescents, in whom the clinical picture somewhat differs from adult-onset depression. Depression is also a frequent comorbidity in somatic and mental disorders and has been considered an actionable outcome for, for example, patients at clinical high-risk for psychoses (CHRsingle bondP). Thus, we studied sociodemographic and clinical predictors of depression/dysthymia in an underage sample with focus on those considered at CHRsingle bondP. Our baseline sample (N = 676) included CHR-P patients (n = 183), inpatients admitted for non-psychotic, non-affective disorders (n = 277), and community participants (n = 216) of age 8.0–17.9 years (43.8 % male). They were interviewed for mental disorders and symptoms with various instruments, including the Mini International Neuropsychiatric Interview for Children and Adolescents, which was also used to assess depression/dysthymia in the CHR-P group at one- and two-year-follow up (n = 117/73). Stepwise logistic regression analyses were used to first identify a cross-sectional baseline model in the complete sample that was then tested prospectively in CHR-P patients. The baseline model included nationality and 13 clinical variables, particularly mild depressive symptoms. Variables contributing significantly to the prediction of persistence or new occurrence of depression/dysthymia varied over time, indicating that depression/dysthymia in CHR-P minors may require different predictors depending on the follow-up time. Furthermore, the prospective accuracy of ruling out depression/dysthymia was superior to the accuracy of ruling it in. This lower positive likelihood ratio might be overcome in future by stepwise approaches that further stratify risk in those CHR-P minors identified as at increased risk of depression/dysthymia.