The Distribution of Major Brain Metabolites in Normal Adults: Short Echo Time Whole-Brain MR Spectroscopic Imaging Findings

Abstract

This prospective study aimed to evaluate the variation in magnetic resonance spectroscopic imaging (MRSI)-observed brain metabolite concentrations according to anatomical location, sex, and age, and the relationships among regional metabolite distributions, using short echo time (TE) whole-brain MRSI (WB-MRSI). Thirty-eight healthy participants underwent short TE WB-MRSI. The major metabolite ratios, i.e., N-acetyl aspartate (NAA)/creatine (Cr), choline (Cho)/Cr, glutamate + glutamine (Glx)/Cr, and myoinositol (mI)/Cr, were calculated voxel-by-voxel. Their variations according to anatomical regions, sex, and age, and their relationship to each other were evaluated by using repeated-measures analysis of variance, t-tests, and Pearson’s product-moment correlation analyses. All four metabolite ratios exhibited widespread regional variation across the cerebral hemispheres (corrected p < 0.05). Laterality between the two sides and sex-related variation were also shown (p < 0.05). In several regions, NAA/Cr and Glx/Cr decreased and mI/Cr increased with age (corrected p < 0.05). There was a moderate positive correlation between NAA/Cr and mI/Cr in the insular lobe and thalamus and between Glx/Cr and mI/Cr in the parietal lobe (r ≥ 0.348, corrected p ≤ 0.025). These observations demand age- and sex- specific regional reference values in interpreting these metabolites, and they may facilitate the understanding of glial-neuronal interactions in maintaining homeostasis.

Keywords: echo-planar; magnetic resonance spectroscopic imaging; metabolite; whole-brain.

Figure 1

The representative single-voxel spectra obtained from the left (A) temporal lobe, (B) occipital lobe, (C) thalamus, and (D) sublobar white matter of WB-MRSI of a 25-year-old male. The voxel locations are shown on the T1-weighted 3D magnetization-prepared rapid acquisition gradient echo (MPRAGE) images.

Figure 2

Box and whisker plots showing the region-specific mean (cross mark), median, interquartile range, minimum, and maximum for (A) N-acetyl acetylaspartate (NAA)/creatine (Cr), (B) choline (Cho)/Cr, (C) glutamate + glutamine (Glx)/Cr, and (D) myoinositol (mI)/Cr (Abbreviations: frontal = frontal lobe, parietal = parietal lobe, temporal = temporal lobe, occipital = occipital lobe, insula = insular lobe, limbic = limbic lobe, lentiform = lentiform nucleus, sublobar = sublobar white matter). Each symbol indicates pairs with statistical significance (i.e., p < 0.05 following correction for multiple comparisons). For example, as indicated by ▶, NAA/Cr is different among the frontal lobe, limbic lobe, and lentiform nucleus.

Figure 3

Scatterplots showing the correlation between (A) glutamate + glutamine (Glx)/creatine (Cr) and myoinositol (mI)/Cr of the parietal lobe (r = 0.348), (B) N-acetylaspartate (NAA)/Cr and mI/Cr of the insular lobe (r = 0.490). The straight line indicates the mean and the dotted lines 95% confidence interval. These correlations are statistically significant at corrected p < 0.05 (Pearson’s product-moment correlation analyses).

Figure 4

Scheme summarizing the processing steps (Abbreviations: CRLB = Cramer-Rao lower bound, MRSI = Magnetic resonance spectroscopic imaging, MPRAGE = T1-weighted 3D magnetization-prepared rapid acquisition gradient echo, NAA = N-acetylaspartate, Cho = choline, Cr = creatine, Glx = glutamate + glutamine, mI = myoinositol).

Figure 5

The outline of the regions of interest (ROIs) overlaid on the axial sections of normalized T1-weighted 3D magnetization-prepared rapid acquisition gradient echo (MPRAGE) images of a participant. Each ROI is placed on the either side of the frontal lobe (1); parietal lobe (2); temporal lobe (3); occipital lobe (4); insular lobe (5); limbic lobe (6); caudate (7) and lentiform nuclei (8); claustrum (9); thalamus (10); sublobar white matter (11) containing corpus callosum, external and internal capsules, and periventricular white matter; midbrain (12); pons (13); and cerebellar hemispheres (14).