Effect of postprandial hyperglycaemia in non-invasive measurement of cerebral metabolic rate of glucose in non-diabetic subjects.

Abstract

The aim of this study was to determine the effect of postprandial hyperglycaemia (HG) on the non-invasive measurement of cerebral metabolic rate of glucose (CMRGlc). Five patients who had a meal within an hour before a fluorine-18 fluorodeoxyglucose positron emission tomography (FDG-PET) examination were recruited in this study. They underwent intermittent arterial blood sampling (measured input function), and, based on this sampling, CMRGlc was calculated using an autoradiographic method (CMRGlc(real)). Simulated input functions were generated based on standardised input function, body surface area and net injected dose of FDG, and simulated CMRGlc (CMRGlc(sim)) was also calculated. Percent error of the area under the curve (AUC) between measured (AUC(real)) and simulated input function (AUC(IFsim)) and percent error between CMRGlc(real) and CMRGlc(sim) were calculated. These values were compared with those obtained from a previous study conducted under fasting conditions (F). The serum glucose level in the HG group was significantly higher than that in the F group (165+/-69 vs 100+/-9 mg/dl, P=0.0007). Percent errors of AUC and CMRGlc in grey matter and white matter in HG were significantly higher than those in F (12.9%+/-1.3% vs 3.5%+/-2.2% in AUC, P=0.0015; 18.2%+/-2.2% vs 2.9%+/-1.9% in CMRGlc in grey matter, P=0.0028; 24.0%+/-4.6% vs 3.4%+/-2.2% in CMRGlc in white matter, P=0.0028). It is concluded that a non-invasive method of measuring CMRGlc should be applied only in non-diabetic subjects under fasting conditions.