Comparison of Salivary Cortisol, Cortisone, and Dexamethasone with Serum Measurements Across Dynamic Adrenal Testing

Abstract Context Salivary cortisol and cortisone provide non-invasive measures of biologically active cortisol, but their performance during dynamic testing remains incompletely defined. The utility of salivary dexamethasone to confirm exposure during dexamethasone suppression testing (DST) is unknown. Objective To compare salivary cortisol, cortisone, and dexamethasone with serum measurements during dynamic adrenal testing and to derive salivary cortisol and cortisone thresholds corresponding to clinically relevant serum cortisol concentrations. Design and Setting Cross-sectional study at a tertiary referral center. Participants Adults undergoing DST (n = 473), cosyntropin stimulation testing (CST; n = 163), or overnight metyrapone testing (OMT; n = 11). Measurements Serum cortisol was measured by immunoassay. Serum and salivary dexamethasone, salivary cortisol and cortisone were quantified by liquid chromatography–tandem mass spectrometry. Results Among 640 participants, 815 paired serum and salivary samples were analyzed. Salivary cortisol and cortisone correlated well with serum cortisol across dynamic testing, with the greatest difference following 1-mg DST (r = 0.86 vs 0.53 for cortisone and cortisol, respectively; both P < 0.001). During CST, correlations ranged from r = 0.75–0.87 for both analytes (all P < 0.001). Salivary dexamethasone correlated with serum dexamethasone (r = 0.61, P < 0.001). Salivary cortisone thresholds corresponding to prespecified clinically relevant serum cortisol concentrations of 1.8, 5, 10, and 18 µg/dL (50, 138, 276, and 497 nmol/L) were 117, 329, 585, and 1262 ng/dL (3.25, 9.13, 16.23, and 35.01 nmol/L), respectively; corresponding salivary cortisol thresholds were 16.5, 38, 80, and 278 ng/dL (0.46, 1.05, 2.21, and 7.67 nmol/L). Conclusions These findings support saliva-based dynamic adrenal testing and identify clinically relevant salivary thresholds.