The realiability of rates of glucose appearance in vivo calculated from constant tracer infusions.

The rate of appearance of unlabelled glucose was calculated from tracer data and compared with the actual rate of infusion of unlabelled glucose into a anaesthetized dog with all sources of endogenous glucose production surgically removed. The mean steady-state rate of appearance of unlabelled glucose calculated from the equilibrium specific radioactivity was insignificantly higher (0.3%) than the actual rate of infusion of unlabelled glucose (n = 6). During non-steady states, a time-variable volume of distribution of glucose (V) was necessary to predict the rate of appearance of unlabelled glucose correctly from the pool-dependent equation described by Steele [(1959) Ann. N.Y. Acad. Sci. 82, 420--430]. Rapid fluctuations in the rate of appearance of glucose could be predicted reasonably well by using a fixed value of V for 40ml/kg, but by using larger fixed values for V (100--160ml/kg) the rates were inaccurate. The pool-dependent two-radiactive-isotope technique described by Issekutz, Issekutz & Elahi [(1974) Can. J. Physiol. Pharmacol. 52, 215--224] predicted single-step increases in the rate of infusion of glucose reasonably accurately, but the Steele (1959) equation was better at predicting sequential changes in the rate of infusion of unlabelled glucose.