Are leucine turnover measurements valid in the intravenously fed state?

We tested whether the primary or reciprocal pool models for whole-body leucine kinetics accurately depict human adaptation to protein deficiency and repletion in the fed state by comparing model-derived leucine oxidation with urea appearance calculated from urinary urea excretion and changes in the body urea pool. Five normal men consumed a control diet providing maintenance energy and 80 g protein/d for 5 days; this was followed by 7 days of an isoenergetic protein-free diet, and finally by a return to the original control diet for 5 days. At the end of each dietary period, urea appearance and leucine oxidation were measured during a 4-hour intravenous infusion of crystalline amino acids providing a total N to leucine N ratio similar to that in mixed body proteins. Primary pool-derived fed-state leucine oxidation decreased after adaptation to protein deficiency and remained low during refeeding (18.6 +/- 1.2, 13.2 +/- 1.1, and 15.0 +/- 1.8 mumol.kg-1.h-1, respectively, P < .01), in agreement with the physiologic prediction. A similar pattern occurred with the reciprocal pool model (24.4 +/- 2.7, 17.3 +/- 2.0, and 20.3 +/- 2.7 mumol.kg-1.h-1, P < .01) as well as with urea N appearance (3.06 +/- 0.32, 2.34 +/- 0.24, and 2.38 +/- 0.26 mmol N.kg-1.4 h-1, P < .05). Despite these relative agreements, absolute rates of whole-body amino acid oxidation were 25% (reciprocal pool model) to 43% (primary pool model) lower than when estimated from urea N appearance.(ABSTRACT TRUNCATED AT 250 WORDS)