Background Inadequate cortisol production in response to critical illness in extremely preterm infants may exacerbate poor outcomes. Despite commonly measuring cortisol concentration and administering hydrocortisone for presumed adrenal insufficiency, the relationship between serum cortisol concentration and illness severity remains unclear in this unique population.

Objective To determine the relationship between cortisol concentrations and illness severity as measured by the Score for Neonatal Acute Physiology II, neonatal Sequential Organ Failure Assessment and Vasoactive-Inotropic Score in premature infants.

Design/Methods This retrospective, single-center cohort study included preterm infants born <30 weeks gestational age admitted to a level IV neonatal intensive care unit (NICU) between June 2011 and July 2018, who had a serum cortisol obtained for clinical indications before 36 weeks PMA. Demographic data were collected on infants and mothers. Nine clinical variables were identified a priori that could potentially modify cortisol concentration including critical illness. Univariate and multivariable analyses determined the relationship between cortisol concentration and each of these variables.

Results A total of 224 preterm infants with pretreatment serum cortisol concentration met criteria for inclusion. The median (IQR) gestational age at birth was 25 weeks (24, 26) and at cortisol measurement was 26 weeks (25, 28). The median cortisol was 13.3 ug/dL. Non-survivors had the highest values. Cortisol concentration did not correlate with any of the selected illness severity scores.

Conclusion(s) Cortisol concentrations in extremely preterm infants did not correlate with illness severity regardless of gestational age. Further studies are needed to identify clinically useful mediators of adrenal dysfunction and to guide clinical management.