Normalization of metabolic flux data during clamp studies in humans

Abstract

BackgroundThere is no consensus in the field regarding the optimal method for the expression of metabolic flux data, such as glucose disposal rates during hyperinsulinemic-euglycemic clamp experiments. Several normalization methods are in use today, but their impact on study outcomes is rarely discussed. MethodsWe illustrate this issue using clamp data from 92 lean and 66 obese subjects. Glucose kinetics and insulin sensitivity were determined during hyperinsulinemic-euglycemic clamp studies using [6,6-2H2]glucose. From this single dataset, we calculated 21 expression methods for the glucose disposal rate during hyperinsulinemic conditions. Results and discussionWith most normalization methods, the obese subjects demonstrated reduced insulin-stimulated glucose disposal as compared to the lean subjects. However, depending on the normalization method, glucose disposal rates in obese subjects ranged from 26 ± 1% to 207 ± 10% of glucose disposal rates in lean subjects. We conclude that data normalization methods greatly impacted metabolic flux outcomes in our dataset of lean and obese subjects. There is no compelling evidence to select one method over the other, but we encourage authors in the metabolic arena to think about, and provide a rationale for, the best normalization method for their specific research questions.