The limited role of midnight salivary cortisol levels in the diagnosis of subclinical hypercortisolism in patients with adrenal incidentaloma.

OBJECTIVE The criteria for defining subclinical hypercortisolism (SH) are debated and a real gold standard test or combination of tests is lacking. Recently, late-night salivary cortisol (MSC) has been described as a sensitive and easy-to-perform marker for diagnosing overt hypercortisolism. No data are available on the role of MSC in the diagnosis of SH. The aim of this study was to evaluate the sensitivity and specificity of MSC levels in the diagnosis of SH in patients with adrenal incidentalomas (AI). METHODS In 103 (females/males, 69/34) patients with AI, MSC levels were studied. One milligram overnight dexamethasone suppression test (DST), urinary-free cortisol (UFC), and ACTH plasma levels were also evaluated. Patients were defined as affected by SH if they showed two of the following criteria: DST>83 nmol/l, ACTH <2.2 pmol/l, and UFC >193 nmol/24 h. RESULTS No difference in MSC levels in patients with SH (3.1+/-3.1 nmol/l) compared with patients without SH (2.2+/-2.8 nmol/l) was observed. In patients with SH, MSC levels were significantly correlated with DST (r=0.4, P<0.05). Using the cut-off of 5.1 nmol/l, the sensitivity and specificity of MSC levels for diagnosis of SH is 22.7 and 87.7% respectively. CONCLUSION In patients with AI, normal levels of MSC do not exclude SH, whereas high levels may suggest the presence of SH identified by conventional tests. Thus, MSC is not suitable as a screening test, although it may be used in conjunction with other tests as the confirming test in selected patients.