Send to

Choose Destination
J Clin Endocrinol Metab. 2005 Oct;90(10):5730-6. Epub 2005 Jul 12.

Reproducibility of nighttime salivary cortisol and its use in the diagnosis of hypercortisolism compared with urinary free cortisol and overnight dexamethasone suppression test.

Author information

Clinic for Endocrinology, Diabetes, and Clinical Nutrition, University Hospital Basel, Switzerland.



Nighttime salivary cortisol (NSC) has been suggested to be a useful diagnostic test for Cushing's syndrome (CS). In the absence of published data on its day-to-day variability, we assessed the reproducibility of NSC by repeated measurements in healthy volunteers. Its diagnostic performance was compared with 24-h urinary free cortisol (UFC) and 1 mg overnight dexamethasone suppression test in 12 patients with CS, 20 healthy volunteers, 14 referred patients in which CS was excluded or not firmly established, 16 obese patients, and 20 women in late pregnancy.


NSC showed a superior reproducibility in healthy volunteers with a low day-to-day variability as reflected by an intraclass correlation coefficient of 0.78. The receiver operating characteristic curve-estimated cutoff of 6.1 nmol/liter (0.22 microg/dl) demonstrated a sensitivity and specificity of 100% (area under the receiver operating characteristic curve, 1.0; 95% confidence interval, 0.94-1.0) in the diagnosis of CS. NSC, 24-h UFC [after adjusting the local laboratory cutoff to 504 nmol/d (183 microg/d)], and the urinary cortisol/creatinine ratio showed a tendency to be superior to 1 mg dexamethasone suppression test in correctly identifying CS. In late pregnancy, the preserved diurnal variation at a higher level of salivary cortisol reduced the specificity of NSC to 75%.


Based on its remarkable reproducibility, easy noninvasive nature, and at least similar diagnostic performance, NSC appears to be a preferable alternative to 24-h UFC as a first-line screening test for CS. The cutoff values of NSC, 24-h UFC, and urinary cortisol/creatinine ratio have to be carefully adjusted using assay and center-specific reference ranges of sufficiently large populations.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Silverchair Information Systems
Loading ...
Support Center