Parenteral nutrition (PN) has been widely used in premature infants untill enteral feeding can be tolerated. Cholestasis is an important complication of PN. The objective of this study was to evaluate the role of serial measurements of serum amyloid A (SAA) during PN and compare its` effectiveness with C-reactive protein (CRP) and procalcitonin (PCT). We also aimed to determine the risk factors for PN associated cholestasis (PNAC). Premature infants ( < 34 weeks` gestational age) who were started on PN during hospitalization were included in this prospective study. SAA, CRP and PCT levels were measured on days 0, 3, 7, 14, and 21 of PN in all infants. Infants who had PN for less than 2 weeks, who developed sepsis and/or necrotizing enterocolitis were excluded. A total of 85 infants were included. The mean birth weight was 1226±329 g, and the mean gestational age was 29.4±1.8 weeks. The birth weight of infants who developed cholestasis were significantly lower. Enteral nutrition was started significantly later in infants with cholestasis. CRP and PCT did not correlate with conjugated bilirubin levels at any time point. SAA levels on days 7 and 14 showed a significant correlation with conjugated bilirubin levels. SAA levels on day 7 was found to have the highest sensitivity for prediction of PNAC. Low birth weight, late commencement of enteral feeding, and prolonged PN were the main risk factors for PNAC development. This is the first study that shows the predictive value of SAA for PNAC development. We suggest that SAA may be used as an accurate and useful biomarker for prediction of PNAC in high risk premature infants receiving PN.

Keywords: C-reactive protein, parenteral nutrition associated cholestasis, premature infant, procalcitonin, serum amyloid-A