Use of (2)H(2)O for estimating rates of gluconeogenesis: determination and correction of error due to transaldolase exchange.

The use of deuterated water as a method to measure gluconeogenesis has previously been well validated and is reflective of normal human physiology. However, there has been concern since the method was first introduced that transaldolase exchange may lead to the overestimation of gluconeogenesis. We examined the impact of transaldolase exchange on the estimation of gluconenogenesis using the deuterated water method under a variety of physiological conditions in humans by using the gluconeogenic tracer [U-(13)C]propionate, (2)H(2)O, and (2)H/(13)C nuclear magnetic resonance (NMR) spectroscopy. When [U-(13)C]propionate was used, (13)C labeling inequality occurred between the top and bottom halves of glucose in individuals fasted for 12-24 h who were weight stable (n = 18) or had lost weight via calorie restriction (n = 7), consistent with transaldolase exchange. Similar analysis of glucose standards revealed no significant difference in the total (13)C enrichment between the top and bottom halves of glucose, indicating that the differences detected were biological, not analytical, in origin. This labeling inequality was attenuated by extending the fasting period to 48 h (n = 12) as well as by dietary carbohydrate restriction (n = 7), both conditions associated with decreased glycogenolysis. These findings were consistent with a transaldolase effect; however, the resultant overestimation of gluconeogenesis in the overnight-fasted state was modest (7-12%), leading to an error of 14-24% that was easily correctable by using either a simultaneous (13)C gluconeogenic tracer or a correction nomogram generated from data in the present study.