Is the "pool-fraction" paradigm a valid model for assessment of in vivo turnover in non-steady state?

Quantification of in vivo turnover of endogenous substances in nonsteady state is of fundamental importance for understanding a variety of physiological and clinical metabolic situations. Toward this end, a pool-fraction model has become a paradigm in the glucose and ketone body areas. We discuss the basic assumptions on which the pool-fraction model is based and the criteria on which it has been validated. Specific comments are then made on its current and potential use for quantifying the non-steady-state turnover of glucose, ketone bodies, and insulin. We conclude that the quantitative reliability of predictions provided by the pool-fraction model is quite poor and that new developments are needed for quantifying the non-steady-state situation.