ABSTRACT
OBJECTIVE: There is currently no optimal test to screen for endogenous Cushing's syndrome (CS) in children. Traditional 24-hour urine or midnight serum cortisol values may be difficult to obtain or elevated by venipuncture stress. We hypothesized that salivary cortisol measurement is a reliable way to screen for CS in children. STUDY DESIGN: Sixty-seven children (5-17 years) were studied: 24 obese volunteers, 29 non-obese volunteers, and 14 children with CS. Saliva was obtained at 7:30 AM, bedtime, and midnight for measurement of free cortisol by radioimmunoassay. RESULTS: Salivary cortisol was detectable in all morning and evening samples from patients with CS but was frequently undetectable in healthy children at bedtime (66%) and at midnight (90%). With cut points that excluded healthy children, a midnight salivary cortisol value of 7.5 nmol/L (0.27 microg/dL) identified 13 of 14 patients with CS, whereas a bedtime value >27.6 nmol/L (1 microg/dL) detected CS in 5 of 6 patients. The diagnostic accuracies of midnight salivary cortisol and urinary free cortisol per square meter were the same (93%). CONCLUSION: Salivary cortisol measurement at bedtime or midnight rules out CS in nearly all cases. Nighttime salivary cortisol sampling is thus a simple, accurate way to screen for hypercortisolism in children.