To determine whether the reproducibility of the first-phase insulin response (FPIR) measured during an intravenous glucose tolerance test is improved by the use of a lower glucose dose or retrograde sampling from an arterialized hand vein.


Previous studies have suggested that the high within-subject variation of FPIR measurement of up to 110% could be reduced by sampling from a retrograde cannulated and arterialized hand vein opposite to the cubital fossa vein through which the glucose was injected or by the use of a lower dose of glucose. Two low-dose (glucose, 5 g/m2 injected over 30 s) and two standard Islet Cell Antibody Registry Users Study (ICARUS) (glucose, 0.5 g/kg injected over 3 min) tests were performed on seven normal subjects at 2-week intervals. Samples were collected simultaneously from the cubital fossa vein, through which the glucose was injected, and from a retrograde cannulated, contralateral hand vein that was arterialized by heating. FPIR was expressed as the sum of the insulin measurements 1 and 3 min after the completion of the glucose injection and as the area under the insulin curve between 0 and 10 min.


Responses to the mean sum of serum insulin concentrations at 1 and 3 min after intravenous glucose were significantly lower for the low-dose test (mean 94 mU/l) than for the high-dose test (mean 184 mU/l) for samples taken from the arm (P < 0.05); mean 0- to 10-min insulin areas were 367 and 596 mU/l for low- and high-dose tests, respectively (P < 0.05). Within-subject coefficients of variation for samples from the hand or the arm ranged from 0.33 to 17.5% and 1.3 to 38% for successive ICARUS and low-dose tests, respectively. Reproducibility, measured by the coefficient of variation between successive tests for each protocol, was not significantly different using samples taken from the arm or the contralateral hand.


The intravenous glucose tolerance test is reproducible when performed by the same operator over a short time span. Reproducibility is not significantly improved by sampling from an arterialized, retrograde cannulated, contralateral hand vein. There is no case for changing the present ICARUS protocol to incorporate retrograde cannulation or low-dose (5 g/m2) glucose.