In preterm infants, protein-turnover rates obtained by [15N]glycine as a tracer are known to be overestimated. This may reflect the insufficient supply of dietary glycine. In this randomized study, the influence of a glycine-rich diet on whole body protein turnover rates in eight male preterm infants (29-32 weeks, 1,200-2,540 g birthweight) using the 15N-tracer technique on days 21 and 28 of life was investigated to evaluate the necessity of supplementing preterm infant formulas with proteins rich in glycine. Before and during the study, the infants were alternately fed with a commercial available preterm infant formula (I, 2% protein, 40 mg glycine/dl) and a variety of this formulation with glycine-rich proteins (II, 2% protein, 130 mg glycine/dl). The protein-turnover rates were computed after 15N-single-pulse labeling with the help of the three-compartment model (TCM) and the urinary ammonia end-product method (AEPM). The tracer used was [15N]glycine (dosage: 2 mg 15N/kg). For the determination of 15N-excess-excretion kinetics, fractionated urine specimens were collected over a 36-h period. The protein-turnover rate calculated by TCM was 8.8 +/- 1.6 g/kg/day (formula I) and 7.7 +/- 2.0 g/kg/day (formula II); using AEPM, the rate was 8.7 +/- 2.5 g/kg/day and 7.5 +/- 1.5 g/kg/day, respectively. We conclude that the presaturation of the precursor pool by an adequate glycine intake minimizes drawbacks that may arise when using [15N]-glycine as a tracer in preterm infants, and a protein concentration of 2%, as in formula I, and consequently, a 170% glycine content when compared with the same human milk volume, meets the glycine requirement.