Multisteroid analysis after DST in depressed patients--a controlled study.

111 consecutively admitted in-patients with a depressive syndrome received a dexamethasone suppression test (DST) after all known factors which might confound the test results had been carefully excluded. Plasma concentrations of cortisol, corticosterone and dexamethasone were compared with several diagnostic evaluations (RDC, DSM-III, ICD-9) in a controlled study. The positive predictive value of nonsuppressed corticosteroid levels was only moderate for each diagnostic category. Diagnostic specificities were 84.6% for major depression, endogenous subtype (RDC), 71.2% for melancholia (DSM-III) and 86.8% for endogenous depression (IDC-9) when using a post-DST cortisol value above 50 ng/ml (5 micrograms/dl) as the referent value to define DST nonsuppression. Combined determination of cortisol and corticosterone as an index of dexamethasone-induced suppression raised the sensitivity rate considerably at the cost of the predictive value for major diagnostic categories. Dexamethasone plasma levels were reciprocally correlated with cortisol levels and neglect of samples with low plasma dexamethasone contents improved the diagnostic performance for endogenous depression according to RDC and ICD-9, but not for DSM-III melancholia. Although it would be speculative to suppose that the observed low dexamethasone levels are involved in DST nonsuppression, the present findings emphasize that multisteroid analysis which includes dexamethasone is important in future studies designed to explore the clinical utility of the DST.