Functional MRI of the hypothalamic response to an oral glucose load

Abstract

To the Editor: Functional MRI (fMRI) using a blood oxygen level dependent (BOLD) sequence provides a tool to measure the effects of alterations in blood glucose on the patterns and magnitudes of neuronal activation in the human brain [1]. Previous experiments using a glucose load have suggested that it is possible in humans to identify changes in the BOLD signal from areas within the hypothalamus [2–6]; however, the results have been contradictory. This study involved seven healthy participants aged 27– 38 years (three female, four male; mean±SD BMI 23.6± 1.3 kg/m), who gave informed consent. The study was approved by the local research ethics committee. Each participant attended for two studies in random order: one involving ingestion of a glucose load and the other water. After anatomical scans and a 5 min baseline period of fMRI acquisition, participants were either given 75 g glucose diluted with 200 ml water or the same volume of water only. fMRI acquisition continued for a further 25 min after the start of drinking. Blood samples for measurement of plasma glucose were taken 5 min before the start of drinking and 25 min after the start of drinking. There was no difference in mean glucose levels between conditions at baseline, but significantly higher values in the glucose arm at the end of the study (4.41±0.31 mmol/l for the water arm vs 5.36± 0.28 mmol/l for the glucose arm, p00.001). With the use of a six-channel head coil at 1.5 T, fMRI variables were: repetition time04 s (three-shot echo planar imaging, volume acquisition in 12 s); echo time045 ms; 50 sagittal slices; 2 mm slice thickness; 0.2 mm interslice gap; 2×2 mm in-plane resolution. The 30 min fMRI measurement produced a sequence of 150 volume acquisitions, which were aligned to the 80th volume using SPM5 software (Wellcome Trust Centre for Neuroimaging, London, UK). The position of the hypothalamus was identified on a near-midline sagittal slice of the 80th fMRI volume acquisition on the basis of landmarks: the anterior commissure (AC), the optic chiasm (OC) and the mammillary body (MM). The MarsBaR toolbox for SPM5 (available from http://marsbar.sourceforge.net/) was used to specify the overall region of interest (ROI)—a parallelogram with three D. E. Flanagan (*) :B. Krishnan :A. Watt Department of Endocrinology, Level 9, Derriford Hospital, Plymouth PL6 8DH, UK e-mail: danielflanagan@nhs.net