“…All MR scan protocols and setups for this longitudinal study were identical to our initial study 6 performed on a 3-T MR system (Allegra; Siemens, Erlangen, Germany) using a custom-built partial volume transmit/receive coil, a quadrature radiofrequency (RF) helmet coil. 18 The participants lay supine in the MR scanner, and a series of MR scans were performed identical to their initial MR scans at Time 1: (1) three-plane scout MR images to locate the volume of interest (VOI), a 3-cm axial slab positioned above the corpus callosum including the frontal to parietal regions; (2) localized automated shimming to optimize static magnetic field homogeneity in the CSI volume by adjusting all first- and second-order shim currents based on measured field maps; (3) conventional CSI for creatine measurement (slice thickness = 2 cm, matrix size = 16 × 16, field of view (FOV) = 20 × 20 cm 2 , VOI = 10 × 12 cm 2 , time to echo (TE)/repetition time (TR) = 30/2000 ms, and number of averages = 2); and 4) the selective multiple quantum CSI of GSH acquired from the selected CSI slice (slice thickness = 3 cm, matrix size = 8 × 8, FOV = 20 × 20 cm 2 , TE/TR = 115/1500 ms, spectral width = 2 kHz, and number of averages = 12). The CSI of GSH is based on a two-echo scheme, allowing simultaneous acquisition of the creatine signal, which serves as an internal concentration reference.…”