Prospective controlled study of parenteral nutrition-associated cholestatic jaundice: Effect of protein intake

The development of transient cholestatic jaundice is a well-recognized complication of prolonged parenteral nutrition. The etiology and long-term significance have not been established. This report describes a prospective controlled study designed to compare the hepatic effects of two different parenteral protein intakes, a lower protein regimen of 2.3 gm/kg/day and a higher protein regimen of 3.6 gm/kg/day. The minimum criterion for ChJ was a direct bilirubin ≥2.0 mg/dl. Eighty-two consecutive infants completed the study. No ChJ occurred in the 39 infants who received PN for less than two weeks. In 43 infants requiring PN for at least two weeks, the incidence of ChJ was similar in the two groups, six infants in the LP and seven in the HP group. However, infants prospectively assigned to the HP group developed ChJ earlier than the LP group, 27±4 versus 47±6 days (P<0.01), and achieved a significantly greater peak direct bilirubin level, 8.4±1.6 versus 3.2±0.3 mg/dl (P<0.001). Retrospective comparison of the 13 ChJ infants to the 30 long-term PN infants without ChJ revealed a significantly greater daily dextrose intake in those developing ChJ, 16.2±1.1 gm/kg/day versus 13.4±0.9 in the group without ChJ (P=0.025). Our results indicate that an increased parenteral protein intake is associated with an earlier onset and increased magnitude of ChJ and suggest that an increased dextrose intake may be associated with an increased frequency of ChJ.