Assessing depression in youth at clinical high risk for psychosis: a comparison of three measures.

Depressive symptoms are prevalent among individuals at clinical high-risk (CHR) for psychosis. Prior studies have used the Beck Depression Inventory (BDI), Hamilton Depression Rating Scale (HDRS), and the "dysphoric mood" item of the Scale of Prodromal Symptoms (SOPS) to assess depressive symptoms in CHR samples. We compared the psychometric properties of these instruments in a CHR cohort, to support the selection of appropriate depressive symptoms measures in future studies and in clinical settings. Internal consistency was assessed using Cronbach's alpha. Construct validity was assessed through correlations with SOPS items that were expected or not expected to be related to depressive symptoms. Criterion validity was assessed by comparing scores between patients with and without a major depressive disorder diagnosis. We hypothesized based on the schizophrenia literature that the BDI would have superior internal consistency and discriminant validity compared to the HDRS, and that all three measures would show convergent validity and criterion validity. The BDI demonstrated superior internal consistency and construct validity in this at-risk sample. The BDI and HDRS differentiated patients with major depressive disorder, but SOPS dysphoria did not. This has implications for the choice of depression measures in future CHR studies and for the interpretation of past findings.