Prediction of relapse in melancholic depressive patients in a 2-year follow-up study with corticotropin releasing factor test

Purpose: To study the power of CRF stimulation test to predict relapse in a sample of melancholic depressive patients in depressed phase, followed-up over a two-year period from the moment they achieved complete remission of depressive symptoms. Methods: Fifty-one outpatients diagnosed with unipolar depressive disorder with melancholic features according to DSM-IV were assessed with the CRF test. The Structured Clinical Interview for DSM-IV (SCID-IV) was used for diagnosis. Monthly follow-up visits were held over a two-year period after remission; relapse was established using HDRS according to Frank's criteria [Frank E, Prien RF, Jarret RB, Keller MB, Kupfer DJ, Lavori PW, et al. Conceptualization and rationale for consensus definitions of terms in major depressive disorder: remission, recovery, relapse, and recurrence. Arch Gen Psychiatry 1991;48:851-5]. Forty-three patients completed the study. Non-controlled antidepressant treatment protocols were used. Predictive statistical analysis was performed through logistic regression. Findings: The final predictive model included three variables: net area under cortisol curve (NAUCC), previous suicide attempt, and stress during follow-up. Sensitivity was of 89%, and specificity was of 92%. NAUCC has shown a predictive power of 80%, with an optimal cut-off point of 251.24 mu g/ml/min. Conclusions: Cortisol is the hormone of the HPA axis which shows the highest power to predict relapse. NAUCC is the most relevant variable. The complete predictive model is a complex combination of biological, clinical and psychoenvironmental variables (NAUCC, previous suicide attempts, and stress during follow-up). Further studies with better control of the psychoenvironmental variables are required to obtain more precise neuroendocrine findings. (C) 2009 Elsevier Inc. All rights reserved.