OBJECTIVE

To evaluate the usefulness of the measurement of midnight salivary cortisol (SC) in the initial diagnosis of Cushing's syndrome (CS).

METHODS

A total of 18 patients with pathologically confirmed CS were recruited in the study. Thirteen patients with pure obesity and 36 healthy people served as controls. Salivary samples at 8:00, 16:00, 24:00 and 8:00 the next day after the 1 mg dexamethasone suppression test(DST) were collected by a commercially-available saliva collection device, and assayed by Electro Chemiluminescence Immuno Assay (ECLIA).

RESULTS

There were highly significant correlations between salivary cortisol and plasma cortisol (PTC) at the 4 time points. The correlation coefficient was 0.73 (0.71, 0.74, 0.78 and 0.61 at the four time points, respectively) (P < 0.05). Midnight salivary cortisol level in Cushing's Syndrome [(10.58 +/- 5.17) nmol/L] was significantly higher than healthy people [(1.55 +/- 0.60) nmol/L]. The receiver operating characteristics curve (ROC) analysis showed a sensitivity of 100.00% and specificity of 84.00% at 8:00 with the cut-off point of 6.21 nmol/L.

CONCLUSION

The measurement of midnight salivary cortisol is a convenience and noninvasive screening test for the initial diagnosis of Cushing's Syndrome.