γ-Aminobutyric acid (GABA) is the chief inhibitory neurotransmitter of the human brain, and GABA-ergic dysfunction has been implicated in a variety of neuropsychiatric disorders. Recent MRS techniques have allowed the quantification of GABA concentrations in vivo, and could therefore provide biologically relevant information. Few reports have formally characterized the reproducibility of these techniques, and differences in field strength, acquisition and processing parameters may result in large differences in measured GABA values. Here, we used a J-edited, single-voxel spectroscopy method of measurement of GABA + macromolecules (GABA + ) in the anterior cingulate cortex (ACC) and right frontal white matter (rFWM) at 3 T. We measured the coefficient of variation within subjects (CVw) and intra-class correlation coefficients on two repeated scans obtained from 10 healthy volunteers with processing procedures developed in-house for the quantification of GABA + and other major metabolites. In addition, by segmenting the spectroscopic voxel into cerebrospinal fluid, gray matter and white matter, and employing a linear regression technique to extrapolate metabolite values to pure gray and white matter, we determined metabolite differences between gray and white matter in ACC and rFWM. CVw values for GABA + /creatine, GABA + /H(2) O, GABA + , creatine, partially co-edited glutamate + glutamine (Glx)/creatine, partially co-edited Glx and N-acetylaspartic acid (NAA)/creatine were all below 12% in both ACC and rFWM. After extrapolation to pure gray and pure white matter, CVw values for all metabolites were below 16%. We found metabolite ratios between gray and white matter for GABA + /creatine, GABA + , creatine, partially co-edited Glx and NAA/creatine to be 0.88 ± 0.21 (standard deviation), 1.52 ± 0.32, 1.77 ± 0.4, 2.69 ± 0.74 and 0.70 ± 0.05, respectively. This study validates a reproducible method for the quantification of brain metabolites, and provides information on gray/white matter differences that may be important in the interpretation of results in clinical populations.
Published in 2011 by John Wiley & Sons, Ltd.