A theoretical model to predict the behavior of glycosylated hemoglobin levels

Abstract

The measurement of glycosylated hemoglobin as a percentage of total hemoglobin is rapidly becoming the standard method of monitoring the average blood sugar level in diabetics for research purposes and may soon become the standard for clinical care and diagnosis. Much speculation exists in the literature about the nature of the glycosylation reaction. Most experimenters expect a linear relationship between the plasma glucose level and percent glycosylated hemoglobin in whole blood; however, a curve of decreasing slope with increasing glucose concentration is found. Here, a reaction model including simple first order kinetics between glucose and hemoglobin and a finite erythrocyte life of 120 days is considered. By carrying out the integration for each erythrocyte cohort followed by an integration combining all cohorts, a curve corresponding to the experimental result is found. In addition, results on expected glycosylated hemoglobin percent as a function of erythrocyte age and plasma glucose concentration are presented as well as a plot of glucose concentration versus glycosylated hemoglobin percent for the 40-day erythrocyte life in mice. All of the results correlate with experimental values in the literature if a rate constant of k = 1·0 × 10 −5 dl mg −1 day is used. The evaluation of a radioactive iron-transferrin experiment in the literature reveals the possibility that the glycosylation reaction begins during erythropoiesis. Finally, a curve is displayed which shows the expected 120-day decay during normoglycemia, of an elevated glycosylated hemoglobin level resulting from a preceding period of constant hyperglycemia.