Relations between feeding intolerance and stress biomarkers in preterm infants.

OBJECTIVES Feeding intolerance (FI) in preterm infants is common but the etiology remains unclear. This study examined FI as a stress-related disease involving brain-gut interactions and tested the model of allostatic load and complications of prematurity. Specific aims were to describe demographic/medical variables and biomarker levels at each time and over time for the sample; describe/compare variables and biomarker levels at each time for infants with/without FI; and compare biomarker interquartile/interpercentile distributions between infants with/without FI. METHODS Preterm infants <32 weeks' gestation were recruited. The primary outcome was FI by day 7 defined as a feeding withheld, discontinued, or decreased because the infant was not tolerating enteral feedings. Allostatic load was operationalized using cortisol and 8-hydroxydeoxyguanosine (8-OHdG) from cord blood and from saliva and urine on days 1, 7, and 14. Descriptive statistics and comparative analyses were performed. RESULTS Seven of 31 infants enrolled met criteria for FI. Infants with FI had lower median urinary cortisol on day 1 (P = 0.007) and trended to have lower cortisol in the cord blood (P = 0.056). Interquartile distributions were significantly different between infants with/without FI for urinary cortisol on day 1 (P = 0.034) and trended for differences in 8-OHdG on day 14 (P = 0.087). Interpercentile distributions were significantly different in salivary cortisol on day 14 (P = 0.034) and trended for differences in 8-OHdG on day 1 (P = 0.079). CONCLUSIONS Results support further testing of the model in a larger sample; investigation of the cellular mechanisms associated with the stress and the free radical/antioxidant systems; and inclusion of prenatal factors.