Dexamethasone (DEX) is a glucocorticoid steroid used in varying dosages for the treatment of a wide variety of diseases such as altitude sickness [1], inflammatory and autoimmune conditions [2], and as a direct chemotherapy drug in leukaemia [3,4]. Previous pre-clinical and clinical studies indicated a decrease in cognitive and executive function as well as psychological changes after the introduction of DEX in disease treatment [5-7]. While these previous studies were done with cognitive testing and patient questionnaires, this work aims to evaluate the unknown pathophysiology behind the neurotoxic effects of DEX by studying structural and ¹H spectroscopy data of healthy subjects at three time-points throughout DEX administration and its washout. We hypothesize that the central action of DEX in the brain will induce variations in neurotransmission in the hippocampus, which has high concentrations of glucocorticoid receptors. Here, we show a significant reduction in brain structure volumes and GABA/water ratio in the hippocampal area from start to end of DEX administration, which is recovered at washout. Twenty healthy human subjects aged 22-28 years old were recruited and orally administered 12mg/day of DEX for 5 days. For each subject, MR data was acquired at three time points, on day 1 before any DEX administration, on day 5 after the last dose of DEX administration, and on day 16 after a washout period of 11 days, which represents 5 half-lives of DEX. Structural and spectroscopy data were obtained at each of the three time points. A structural sagittal T1-weighted MPRAGE sequence was acquired with the following parameters: 256mm FOV, TE/TR/TI = 1.9/2300/900 ms, 192 contiguous slices, with a resolution of 1 x 1 x 1 mm³. Single-voxel edited ¹H MR spectra were acquired with MEGA-PRESS on a 40 x 25 x 15 mm³ voxel-of-interest positioned along the midline of the right hippocampus, shown in Fig. 1. Water suppression pulses were applied at 35 Hz bandwidth and MEGA-editing was done with Gaussian editing pulses of bandwidth 44 Hz applied at 1.9ppm and 7.5ppm alternately, with TR/TE = 1500/68ms. The whole spectrum had a bandwidth of 1000 Hz. 240 averages were done with no phase-cycling for a scan time of 12:00 minutes. Fig.2. shows a representative MEGA-PRESS spectrum acquired. One additional average of a MEGA-PRESS spectrum was acquired without water suppression. Brain volume and thickness estimates were obtained with Freesurfer [8], and images were processed with the longitudinal stream with an un-biased within-subject template space [9]. Spectra were processed with locally written fitting software in Matlab, and the GABA signal at 3 ppm and unsuppressed water signal were integrated with a linear fit of the spectral baseline. GABA values are expressed with reference to the unsuppressed water signal to give GABA/water metabolite ratios. Metabolite ratios and brain volume estimates from Day 5 and Day 16 were compared with Day 1, using a two-tailed paired t-test.Total brain (without ventricles), total gray, total white, hippocampus, cerebellum and thalamus volumes were compared to assess volume changes throughout DEX administration and washout. There was no significant change in white matter volume at day 5 of DEX, but a mean decrease in white matter volume reached significance at Day 16. There were significant decrease in total brain (p<<0.05), gray matter (p<<0.05), hippocampus (p< 0.013), cerebellum (p<<0.05) and thalamus (p<<0.05) volumes, which recovered on day 16. Bonferroni correction for multiple comparisons was applied in all statistical testing. Fig. 3 summarizes the changes in volumes observed for Day 5 and Day 16, compared to the baseline provided by Day 1. Fig 4. shows that a similar pattern was observed in GABA/water metabolite ratios. GABA/water was significantly reduced in the hippocampus from start (Day 1) to end (Day 5) of DEX administration (p < 0.02). This reduction in GABA/water is recovered at Day 16, so that there is no significant difference in metabolite ratio values before DEX administration and washout. DEX is commonly used to decrease brain edema, by reducing CSF and decreasing capillary permeability [10]. Our findings of volume decrease in various brain regions and subsequent reversibility in several of these volumes after DEX administration is consistent with literature as it could be due to the reduction of CSF and fluids. A decrease in white matter was observed at washout, and we postulate that recovery was not observed due to the duration of the study. All other volumes trended towards recovery at Day 16. The reduction in GABA/water supports our hypothesis of changes in the hippocampus, and indicates a physiological reason for previous observations of psychological changes during DEX administrations.