Dorsey MJ, Cohen LE, Phipatanakul W, Denufrio D, Schneider LC. Ann Allergy Asthma Immunol. 2006;97:182–186 PURPOSE OF THE STUDY. To evaluate dehydroepiandrosterone sulfate (DHEA-S), a corticotropin-dependent adrenal androgen precursor, as a possible marker for adrenal function and hypothalamic-pituitary-adrenal axis suppression in children treated with inhaled corticosteroids (ICSs) compared with low-dose (0.5 μg/m2 up to 1.0 μg) and standard-dose (250 μg) cosyntropin-stimulation testing. STUDY POPULATION. Twenty-two patients with moderate-to-severe–persistent asthma receiving a medium-to-high dose of ICSs for at least 6 months were enrolled (definition of median-to-high dose of ICS: budesonide >400 μg/day or fluticasone >176 μg/day for children <6 years old or >200 μg/day for those ≥6 years). Patients had received no more than 2 courses of systemic corticosteroid exposure of <10 days’ duration in the previous 6 months and no systemic corticosteroid in the 1 month before enrollment. The average age of the patients was 8.6 years (range: 2–12 years). METHODS. After a 12-hour fast, morning cortisol, corticotropin, DHEA-S, and fasting blood sugar levels were measured. Cortisol was measured after the stimulation tests. A cortisol level of ≤18 μg/dL was considered abnormal (adrenal suppression). RESULTS. Of 22 patients, 13 (59%) had an abnormal response to low-dose cosyntropin. One patient had an abnormal standard-dose cosyntropin test result. The normal and abnormal low-dose cosyntropin responders did not differ in age, height, BMI, predicted forced expiratory volume in 1 second, or morning cortisol, corticotropin, or fasting blood sugar levels. There was no difference between normal and abnormal low-dose cosyntropin responders in relation to type of ICS (fluticasone, fluticasone-salmeterol, or budesonide) or dose. DHEA-S levels were significantly lower in abnormal low-dose cosyntropin responders compared with normal responders (31 vs 91 μg/dL; P = .004). Age- and gender-specific mean DHEA-S z scores were significantly lower in abnormal low-dose cosyntropin responders. Receiver-operating-characteristic (ROC) curves for DHEA-S z scores were calculated to obtain optimal cutoff values for DHEA-S. The ROC curve for DHEA-S z scores reached 100% sensitivity with a DHEA-S z score of less than −1.5966 and 100% specificity with a DHEA-S z score >0.0225. CONCLUSIONS. Fifty-nine percent of the children on a medium-to-high dose of ICS had biochemical evidence of adrenal suppression according to the low-dose cosyntropin-stimulation test. Low DHEA-S levels can be used as a screening test to identify children who need more formal testing of the hypothalamic-pituitary-adrenal axis. The half-life of DHEA-S (10–20 hours) is substantially longer than that of cortisol (<2 hour) and, therefore, has less diurnal variation or fluctuating concentration depending on exogenous stress and time of day. REVIEWER COMMENTS. The cosyntropin-stimulation test is cumbersome, labor intensive, and not practical as a screening test, so the DHEA-S test may be more useful. The number of patients on a moderate dose of ICS found to have biochemical adrenal suppression by low-dose cosyntropin tests (59%) is higher than seen in most other studies. The authors speculated that the higher rates of adrenal suppression may be associated with “real-world” ICS use versus study-protocol use. The clinical significance of this biochemical suppression is not clear.