Abstract
Background: There is a negative glucose gradient between the capillary and venous systems, produced by glucose uptake into peripheral tissues. This gradient is augmented by oral glucose ingestion in healthy volunteers; thus prandial status may impact on capillary glucose meter performance. Our primary aim was to investigate whether the (capillary-venous plasma) glucose difference changed in relation to prandial status, in healthy volunteers.
Methods: Glucose was measured fasting and also one hour after an ad libitum breakfast, in 103 healthy volunteers. Duplicate capillary (finger stick) measurements were undertaken at both time points, using both the FreeStyle Lite and AccuChek Performa meters. Simultaneous venous (antecubital fossa) samples were centrifuged immediately after collection and plasma glucose was measured using the laboratory hexokinase method.
Results were compared by Bland-Altman difference analysis. Results: The mean (95% CI) pre- and postprandial (capillary-plasma) glucose differences (mmol/L) were calculated for each meter. For the Freestyle Lite, the preprandial difference was -0.51 (-0.58 to -0.45) and postprandial difference was 0.81 (0.69-0.94). Corresponding differences for the Performa were -0.13 (-0.20 to -0.06) and 1.19 (1.07-1.31), respectively. T-test comparison of participants' paired pre- and postprandial (capillary-plasma) glucose differences confirmed a significant meal-related change in glucose estimation for both meters (P < 0.0001). Also, both meters read highest at lower glucose concentrations.
Conclusions: In healthy volunteers, both glucose meters showed a systematic positive bias one hour after breakfast. The significance of this finding in diabetes remains to be determined.