The relationship between morning serum cortisol and the short ACTH test in the evaluation of adrenal insufficiency.

Background: Morning serum cortisol examinations and short corticotropin (ACTH) tests are popular screening tests for unstressed patients suspected of having adrenal insufficiency. However, the correlation between morning serum cortisol and the peak cortisol response in the short ACTH test after intravenous injection of ACTH has not been studied before. This retrospective study examined the relationship among the mean basal morning cortisol level, the single random morning cortisol level and the peak cortisol level in the short ACTH test in evaluation of adrenal insufficiency.

Methods: In this retrospective study, we examined the relationship among the mean basal morning cortisol level and the peak cortisol level stimulated by intravenous injection of 250 microg synthetic ACTH in 106 unstressed patients with proven or suspected hypothalamic-pituitary-adrenal disease. Plasma cortisol levels were determined by radioimmunoassay. The correlation of the basal morning cortisol to the short ACTH test was determined by linear regression analysis. The ACTH test was analyzed using the receiver operating characteristic (ROC) curve method, and the cut-off points for various sensitivity and specificity were calculated.

Results: The mean basal is highly correlated to peak cortisol response in the ACTH test (r = 0.7724, p < 0.0001). Iatrogenic Cushing's syndrome, caused by ingestion of herbs with illegal steroid additives, was the most common cause (60%) of adrenal insufficiency in the 52 patients with subnormal result. A mean basal morning serum cortisol of > or = 300 nmol/L excluded the possibility of adrenal insufficiency, and a level of < 110 nmol/L suggested adrenal insufficiency. A mean basal cortisol level of > or = 234 nmol/L predicted a normal cortisol response in the ACTH test with optimal sensitivity (80.6%) and specificity (91.4%).

Conclusions: The mean morning cortisol level is a cost-effective screening test in predicting the results of the ACTH test. A mean morning cortisol level > or = 300 nmol/L in suspicious cases of adrenal insufficiency which are not due to acute central etiologies excludes the possibility of subnormal cortisol response to ACTH; whereas a level < 110 nmol/L is likely to have subnormal cortisol response to ACTH. The short ACTH test or insulin tolerance test should be performed in patients with a morning cortisol level 300 nmol/L or with clinical symptoms highly suggestive of adrenal insufficiency.