The use of amino acids labeled with stable isotopes represents a relatively new approach for determining kinetic parameters of apolipoprotein metabolism; thus, several aspects of experimental protocols need to be defined. The aims of the present study were to determine whether a) different amino acid tracers or b) different methods of tracer administration affected apolipoprotein (apo) B kinetic parameters obtained by multicompartmental modeling, and c) to compare very low density lipoprotein (VLDL)-apoB metabolic parameters determined by multicompartmental modeling with those estimated by linear regression or by monoexponential analysis. [1-13C]leucine and [15N]glycine were given either as bolus injections or as primed constant infusions. A bolus of one amino acid was administered simultaneously with a primed constant infusion (8 h) of the other amino acid into four healthy normolipidemic subjects (age 23.0 +/- 1.4 yr; BMI 20.9 +/- 0.9 kg.m-2). VLDL-, intermediate density lipoprotein (IDL)-, and low density lipoprotein (LDL)-apoB enrichments were followed over 110 h. For subsequent analysis these values were converted to tracer/tracee ratios. Using the multicompartmental model, the fractional catabolic rate (FCR) for VLDL-apoB was estimated to be 0.36 +/- 0.09 h-1 after the administration of the tracer as a primed constant infusion and 0.35 +/- 0.07 h-1 when the tracer was administered as a bolus. The values for VLDL-apoB production were 14.6 +/- 6.5 mg.kg-1.d-1 and 14.1 +/- 5.4 mg.kg-1.d-1, respectively. The corresponding values for LDL-apoB were 0.027 +/- 0.016 h-1 (0.026 +/- 0.018 h-1) for the FCR and 10.5 +/- 3.7 mg.kg-1.d-1 (10.4 +/- 3.8 mg.kg-1.d-1) for the production following administration of the tracer as a primed constant infusion and a bolus, respectively. Approximately 47% of VLDL-apoB ultimately reached the LDL fraction via the VLDL-IDL-LDL pathway. Thirty-five percent of LDL-apoB did not originate from this cascade pathway, but was shunted from a rapidly turning over VLDL compartment directly into the LDL fraction. While there was some variation between individuals, VLDL-apoB and LDL-apoB parameters derived from the bolus and the primed constant infusions showed no significant differences and were closely correlated. Metabolic parameters were also independent of the two amino acids tested. Although values for FCRs of VLDL-apoB obtained from linear regression (0.36 +/- 0.19 h-1) or monoexponential analysis (0.50 +/- 0.36 h-1) did not differ significantly from those obtained by the multicompartmental model, there was considerable variation and no significant correlation in a given individual.(ABSTRACT TRUNCATED AT 400 WORDS)