The intravenous glucose tolerance test (IVGTT) is a tool for assessment of glucose disposal in diabetes research and clinical investigations. The test includes a bolus injection of hypertonic glucose and repeated measurement of the plasma glucose and insulin concentrations over up to 3 hours, sometimes with administration of insulin after 20 min [
An overlooked aspect of IVGTTs is the cardiovascular strain caused by osmotic translocation of intracellular fluid caused by hypertonic glucose. Only 50–80 mL of fluid are administered, but how much and how fast the plasma volume becomes expanded does not seem to have been previously studied. This is of interest, as the test might be administered to diabetic patients with limited cardiovascular reserves.
Mathematical methods for analysis and simulation of the disposition of hypertonic crystalloid [
Twenty volunteers, 8 females and 12 males, aged between 18 and 51 years (mean, 28) and with a body weight of 49–88 kg (mean, 68) were studied. The study was approved by the Regional Ethics Committee of Stockholm, and informed consent was obtained from the participants. Mean BMI was 23.4 (SD 2.3) kg/m2, and the serum concentrations of electrolytes and of HbA1c were normal in all cases. None of the subjects used daily medication.
The experiments were conducted at the Department of Endocrinology at Södersjukhuset and began at approximately 8 am. After an overnight fast, each volunteer was placed comfortably on a bed. In each subject, a cannula was inserted into the antecubital vein of each arm, one for sampling blood and the other for infusing fluid. Monitoring consisted of electrocardiography, pulse oximetry, and noninvasive blood pressure.
After a 30 min equilibration period to obtain hemodynamic steady state, a short regular IVGTT was performed by administrating 0.3 g/kg of glucose in a 30% solution over 1 min. Blood was sampled at 0, 2, 4, 6, 8, 10, 20, 30, 40, 50, 60, and 75 min for assessment of the plasma glucose and blood hemoglobin (Hb) concentrations. Plasma glucose was measured by the glucose oxidase method used by the hospital’s usual laboratory, and the Hb concentration was measured by colorimetry (Technicon Advia, Bayer, Tarrytown, NY, USA).
Plasma concentration
The half-life (
Hypertonic glucose causes an osmotic shift that draws water from the intracellular (40% of the body weight (
The optimal estimates for the unknown parameters in the glucose and fluid models were calculated for each of the 20 experiments individually by nonlinear least-squares regression. No weights were used. The software used was Matlab 4.2 (Math Works Inc., Natick, Mass, USA).
The half-life (
The results were presented as mean and standard deviation (SD) and, when there was a skewed distribution, as the median (25th–75th percentile range). All reported correlations were statistically significant by
Baseline plasma glucose was 4.8 (0.5) mmol/L, and the blood Hb concentration was 13.5 (11.7–14.0) g/dL.
All 20 experiments could be analyzed with the proposed equations for plasma glucose and insulin kinetics. The modeled rise in plasma glucose was 8 mmol/L (Figure
(a) Change in plasma glucose concentration. (b) Dilution of venous plasma calculated from changes in the blood hemoglobin concentration in response to an intravenous injection of 0.3 g/kg of glucose over 1 min. Each experiment is represented by a thin line and the modeled average by a thick line.
(a) Simulation showing the plasma volume expansion that would ensue if the injection time was increased to 5 min (left) from 1 min (right) for an IVGTT dose of 0.2 g/kg and 0.3 g/kg of glucose, respectively. (b) Simulation of the plasma glucose level if the glucose load was decreased to 0.2 g/kg (left) from 0.3 g/kg (right) in surgical patients (solid line) and in volunteers (broken line). The kinetic data for volunteers were taken from the present study, and those for surgery (
The
Calculation of the osmotic fluid shift resulting from the injection of hypertonic glucose showed that each mmol of exogenous glucose translocated 1.83 mL of fluid from the intra- to the extracellular fluid space. The translocated plus injected fluid volume expanded a body fluid space of 3.0 (2.5–4.0) L in size, and the
Recalculating the fluid kinetics on the assumption that 1.83 mL of fluid was eliminated for each mmol of eliminated glucose yielded practically the same size
Computer simulations were performed based on the average parameters obtained by the kinetic analysis of the glucose and fluid kinetics.
The simulations suggested that the brisk peak in plasma dilution at the end of the injection of hypertonic glucose could be reduced to 60% by lowering the dose of glucose from 0.3 to 0.2 g/kg and extending the time of injection from 1 to 5 min (Figure
Modification of the IVGTT could also be justified in clinical situations associated with altered glucose kinetics. Simulations using
The hypertonic glucose caused a virtually instant plasma volume expansion of 10%, which corresponded to approximately 300 mL in our volunteers. Most of this volume (75%) was allocated from the intracellular fluid space by virtue of osmosis. The return of the plasma volume to normal was also governed by osmosis, as
Volume expansion was much greater than would be expected if assuming that fluid translocated by osmosis is distributed throughout the extracellular fluid space. If that had been the case, the plasma volume expansion of 10% would have represented an extracellular volume expansion of almost 1.4 L, which is unreasonable considering the limited amount of injected osmotically active glucose molecules. Rather, mass balance calculations indicate that the translocated plus injected fluid volume was in the range of 300 mL. Hence, the size of the body fluid space expanded by the infused volume could not have been larger than the plasma volume. This finding is not intuitive but is in agreement with previous kinetic studies based on IV infusion of isotonic glucose solutions in healthy volunteers [
Plasma dilution of 10% is not completely trivial, as it corresponds to that measured at the end of a 45 min infusion of 12.5 mL/kg (approximately 1 L) of glucose 2.5% with electrolytes [
The degree and speed of the onset of plasma volume expansion would be far less dramatic by decreasing the dose of hypertonic glucose from 0.3 to 0.2 g/kg and by increasing the injection from 1-2 minutes to 5 minutes. The simulations we performed based on the kinetic parameters suggest that the plasma volume expansion would then be cut in half and also develop more slowly.
The kinetic analyses also argue against the use of a conventional IVGTT in the presence of surgery. A smaller dose, 0.2 g/kg, would smooth the hyperglycemia enough to prevent prolonged glycosuria (
A limitation of the present study is that no invasive hemodynamic monitoring was performed. However, one would hardly expect problems in the cardiovascular adaptation to the rapid onset of plasma volume expansion in the group of healthy volunteers used in this study. Hemodynamic measurements would be of greater interest in elderly subjects with reduced cardiovascular reserves, such as diabetic patients. Insulin-resistant persons, such as those with obesity, would also be likely to experience a more long-lasting plasma volume expansion with an ensuing risk of hemodynamic stress. The reason is that the fluid clearance is governed almost completely by the rate of uptake of glucose to the cells, which occurs more slowly in insulin-resistant states.
In conclusion, the plasma volume expansion resulting from an IVGTT is greater than commonly believed (10%) because the injected and osmotically translocated fluid have a volume of distribution that corresponds only to the plasma volume instead of to the entire extracellular fluid space. A smaller dose and slower injection time might be appropriate if an IVGTT is applied in debilitated or surgical patients.
Drs. Filip Larsen and Stefan Ljunggren assisted during the experiments. Financial support was received from the Stockholm County Council (Grant no. 2009-0433) and Stiftelsen Olle Engkvist Byggmästare.