Diagnostic performance of the thirty-four hour dexamethasone suppression test.

The performance of the dexamethasone suppression test (DST) in the diagnostic confirmation of endogenous depression was compared according to two times of blood collection--1600 hr on day 2 (usual sample) and 0800 hr on day 3 (34 hr after dexamethasone intake)--in 14 endogenous depressives and in a control group of 17 psychiatric inpatients with other diagnoses. For the day 2 (1600 hr) sample, a 5 micrograms/dl cortisol concentration represented the best cut-off score, with sensitivity of 57% specificity of 88%, and diagnostic confidence of 80%. For the day 3 (0800 hr) sample, the best cut-off score was 20 micrograms/dl, with the same sensitivity (57%) but there was a decrease of both specificity (to 76%) and diagnostic confidence (to 67%). The mean cortisol levels were much higher on day 3 than on day 2, suggesting that the inhibitory activity of dexamethasone was no longer present.