RESUMO
BACKGROUND: Insulin-induced hypoglycemia (IHT) is considered the gold standard test for evaluating the HPA axis. Serum free cortisol or its surrogate, salivary cortisol as opposed to total cortisol concentrations, offers a better reflection of the activation of HPA axis. Our study aimed to derive reference ranges for the normal salivary cortisol levels in healthy patients and patients with adrenal insufficiency. DESIGN AND METHODS: Serum cortisol concentrations, using the gold standard of IHT, and salivary cortisol were obtained. 36 patients referred to our outpatient endocrine testing unit for evaluation of adrenal function were included in the study. Most subjects had a history of suspected hypothalamic/pituitary disease causing adrenal insufficiency. RESULTS: We found a strong linear correlation between the serum and salivary cortisol concentrations in simultaneously collected samples (r=0.81, 95% CI 0.74-0.86, p<0.0001). The corresponding salivary cortisol equivalent to a serum cortisol of 500 nmol/L, using a linear-regression equation, was 16.7 nmol/L (95% CI 13.3-20.1 nmol/L, p=0.0001). A salivary cortisol of 13.3 nmol/L has a specificity of 89.3% to detect abnormal HPA function. Using the upper 95% CI result of salivary cortisol 20.1 yields a sensitivity of 87.5%. CONCLUSION: With the present assay, adrenal insufficiency may be diagnosed with reasonable confidence if a random salivary cortisol is lower than 13.3 nmol/L and excluded if a random salivary cortisol is higher than 20.1 nmol/L. Future studies should correlate these thresholds with clinical outcomes.