[Biological hepatic abnormalities, cholestatic jaundice and hospital artificial nutrition. A comparative study in adults with cyclic total parenteral nutrition and enteral nutrition].

Two hundred and fourteen hospitalized nonsurgical, nonacutely stressed adult patients, who were receiving, either prolonged (greater than 21 days) cyclic total parenteral nutrition (n = 106) or enteral nutrition (n = 108), including lipids, for gastrointestinal disorders, and who were free of prenutrition liver function test abnormalities were prospectively investigated for hepatic dysfunction: 41 p. 100 and 17 p. 100 of parenteral and enteral nutrition patients, respectively, developed abnormalities of liver function tests after 25 +/- 5 and 21 +/- 4 days (p less than 0.01). This prevalence was unrelated to catheter sepsis, blood transfusion, caloric load, site of digestive disease or initial nutritional status, and decreased by 50 and 40 p. 100 before the end of total parenteral nutrition and enteral nutrition respectively. In the parenteral nutrition group, separate infusion of lipids, as compared with mixed infusion, was associated with a lower rate of hepatic dysfunction (29 vs 54 p. 100; p less than 0.02). Nine (8.5 p. 100) parenteral nutrition but no enteral nutrition patients developed severe cholestatic jaundice, 47 +/- 39 days after the onset of liver function test abnormalities; no identifiable cause excepting parenteral nutrition per se was found in 3 patients with diffuse obstructive small intestinal disease; jaundice led to death from noncirrhotic hepatic failure in the 2 patients who remained totally parenteral nutrition-dependent.(ABSTRACT TRUNCATED AT 250 WORDS)