231 - Cortisol Measurements in Very Low Birthweight Infants: Characteristics of those who have Cortisol Measurements Taken and Associations with Neonatal Acute Kidney Injury

Little information exists informing adrenal insufficiency and the utility of cortisol levels in neonates. In addition, it is unclear if associations exist between adrenal insufficiency and neonatal acute kidney injury (AKI).

Objective:

We aimed to: (1) identify characteristics that are associated with cortisol levels measured in very low birthweight (VLBW, i.e., < 1,500 grams) infants and (2) investigate associations between cortisol checks and AKI.

Design/Methods:

We conducted a single-center retrospective chart review of VLBW infants admitted between 1/1/18-12/31/20. AKI was diagnosed using the neonatal modified Kidney Disease: Improving Global Outcomes (KDIGO) serum creatinine criteria.

Comparisons between those with and without cortisol levels drawn were made using chi- square, Fisher’s exact, Student’s t, and Wilcoxon rank sum test as appropriate. Bivariate and multivariate generalized linear mixed modeling was used to examine associations between cortisol and AKI.

Results:

Of 567 infants, 154 (27.2%) had cortisol levels drawn in the NICU. Compared to neonates without cortisol checks, those with cortisol levels measured were significantly smaller at birth, had lower gestational age, were more commonly male, and had lower 1- and 5-minute APGAR scores. Neonates with cortisol checks were born to mothers who experienced less pre-eclampsia but more chorioamnionitis. Intrauterine growth restriction was less common among neonates with cortisol checks, but preterm labor was more common compared to infants without checks (all p< 0.05, Table 1).

Among those with cortisol checks, 286 cortisol levels were drawn, with a median cortisol of 6 mg/dL [IQR 2.8,11.9] and a mean 2.0 ± 1.7 levels drawn per patient. Cortisol levels were drawn at median 10 days of life [IQR 1.5,21.5].

AKI occurred in 130 (22.9%). When examining associations between cortisol checks and AKI, neonates who had a cortisol level checked were at increased risk for AKI, recurrent AKI, and increased length of stay (LOS) after adjusting for potential confounders (all p< 0.01, Table 2). No association was detected between cortisol and AKI severity by KDIGO stage.

 

Conclusion(s):

There were significant differences in baseline/demographic characteristics in our cohort when comparing VLBW infants who did and did not have cortisol levels drawn in the NICU. Those with cortisol levels drawn were at increased risk for AKI, recurrent AKI, and mortality and had LOS nearly 2.5x higher than those without cortisol levels checked. Further investigations into the associations between cortisol measurement and AKI are warranted.