Psychopathological correlates of plasma cortisol after dexamethasone suppression: a polydiagnostic approach

Seventy-seven consecutively admitted inpatients with depressive syndromes were examined with the Present State Examination and classified according to eight different operational diagnoses of endogenous depression. All patients received a 1.5 mg dexamethasone suppression test (DST). Sensitivity, specificity and the corrected predictive values of DST nonsuppression (50 or more ng/ml at 0800 hr, 1600 hr, or 2300 hr), adjusted to a 50% prevalence of endogenous and nonendogenous depression, varied considerably depending on the diagnostic definition used. The highest predictive value (89.9%) was found with the Taylor-Abrams criteria (sensitivity = 43.9%, specificity = 95.0%), and the lowest predictive value (53.3%) with DSM-III (sensitivity = 37.7%, specificity = 68.1%). Eliminating the patients with dexamethasone levels of less than 2000 pg/ml improved the diagnostic specificity of the DST for most of the eight definitions of endogenous depression. This further indicates that plasma dexamethasone levels should be analyzed in studies designed to explore the diagnostic utility of the DST. A significant, chance-corrected association between DST nonsuppression and the diagnosis of endogenous depression was found with clinical diagnosis (according to the International Classification of Diseases), and for four out of eight operational diagnoses (Newcastle Scale I, Newcastle Scale II, Taylor-Abrams Criteria, and Vienna Research Criteria). For the other diagnoses (Research Diagnostic Criteria, DSM-III, Michigan Discriminant Index, and Hamilton Endogenomorphy Index), no significant association was found. The RDC criterion "early or intermittent awakening" was the only one out of 28 diagnostic criteria tested which was significantly associated with DST nonsuppression.