Endocrine Abstracts

Background: Adrenal insufficiency in neonates presents a diagnostic challenge due to non-specific clinical signs, physiological immaturity of the hypothalamic-pituitary-adrenal axis, and lack of established gestation-specific reference intervals or cutoff values for random unstimulated cortisol levels. Unstimulated (basal) cortisol is frequently measured in neonates as a preliminary screen for adrenal insufficiency. However, its discriminatory value in predicting adrenal response on Short Synacthen Testing (SST) remains poorly established.

Objective: To evaluate the discriminatory performance of unstimulated serum cortisol in predicting SST outcomes in a neonatal cohort.

Methods: This was a retrospective cohort study which included neonates admitted to a tertiary neonatal intensive care unit over a five-year period (2020–2024) who underwent both basal cortisol testing and SST. SST “pass” was defined as peak serum cortisol ≥450 nmol/l at both 30 and 60 minutes post-ACTH stimulation. Receiver Operating Characteristic (ROC) curve analysis was performed to assess discriminatory performance, and optimal thresholds derived using the Youden Index.

Results: A total of 47 neonates were included, with a mean gestational age of 32.8±0.8 weeks and a mean birth weight of 1846±166 g. Sixty-six percent were male, and the cohort included 36.2% small-for-gestational-age (SGA), 59.6% appropriate-for-gestational-age (AGA), and 4.3% large-for-gestational-age (LGA) infants. The most common clinical indication for cortisol testing was hypoglycaemia (42.3%), followed by hyperbilirubinaemia (17.9%), clinical suspicion of adrenal insufficiency (16.7%), differences in sex development (6.8%), concerns regarding pituitary abnormalities (6.4%), and other causes (9.8%). The area under the ROC curve was 0.640, indicating limited discriminatory ability of basal cortisol to predict an adequate cortisol response following ACTH stimulation. A cortisol threshold of 183.5 nmol/l achieved high specificity (90%) but low sensitivity (33%), with a Youden Index of 0.233.

Conclusion: Basal cortisol demonstrated limited utility in predicting adrenal response to ACTH stimulation in our neonatal cohort, with poor sensitivity. These findings underscore the inadequacy of relying on basal cortisol alone in the evaluation of neonatal adrenal function. Future research should focus on establishing gestation-specific reference intervals for cortisol to improve the interpretation of basal values and support more accurate and timely diagnosis of adrenal insufficiency in this vulnerable population.