Assessment of postprandial glucose metabolism: conventional dual- vs. triple-tracer method

The dual-tracer method has been used conventionally for assessment of postprandial fluxes, i.e., appearance in plasma of ingested glucose (R-a (meal)), endogenous glucose production (EGP), and disposal (R-d). To quantify the magnitude of errors affecting the calculations and their dependence on model assumptions, this method was assessed and compared with the triple-tracer method, which provides model-independent estimates. For this purpose, the dual- tracer protocol was performed twice in eight normal subjects, with [1-C-13] glucose to trace ingested glucose and [6,6-H-2(2)] glucose constantly infused. A third tracer, [6-H-3] glucose, was infused at variable rates to render the calculation of Ra meal and EGP virtually model independent. The dual- tracer method analyzed with a one-compartment model performed poorly, since R-a (meal) peak was significantly lower and delayed compared with triple-tracer reference, resulting in a significantly lower estimation of the amount of absorbed glucose (9,036 +/- 558 vs. 11,316 +/- 823 mu mol/kg, P = 0.0117). EGP showed a paradoxical pattern, with an initial overshoot followed by a rapid decay to negative values, resulting in a significant underestimation of EGP suppression (57 +/- 3 vs. 65 +/- 4%, P = 0.0117). A two-compartment model performed better but did not overcome the limitations of the dual- tracer approach, since the amount of absorbed glucose was still significantly underestimated (10,231 +/- 661 vs. 12,169 +/- 838 mu mol/kg, P = 0.0117) and EGP still showed a paradoxical behavior. R-d, estimated from R-a (meal) and EGP, was significantly underestimated with the dual- tracer method, irrespective of adopted model. We conclude that three suitably infused tracers are required for accurate assessment of postprandial R-a (meal), EGP, and R-d.