Diurnal stability and long-term repeatability of neurometabolites using single voxel 1H magnetic resonance spectroscopy

Purpose: This study was designed to evaluate the diurnal stability and long-term repeatability and reliability of one-dimensional (1D) hydrogen magnetic resonance spectroscopy (1H-MRS) in vitro and in vivo at 3 T. Material and method: A standard brain phantom was used for in vitro study. In vivo diurnal evaluation involved ten healthy subjects, while repeatability study involved six subjects. MRS was acquired from posterior cingulate gyrus (PCG), and processed with LCModel. Diurnal effects were assessed with repeated measures ANOVAs, repeatability was evaluated using coefficient of variation (CV), while reliability was assessed with standard error measurement (SEM) and intra-class correlation coefficient (ICC). Results: Diurnal metabolic changes in vitro were non-significant. The intra/inter-in vitro CVs for the major metabolites; N-acetylaspartate (NAA), creatine (Cr), myo-inositol (mI), glutamate+ glutamine (Glx) and total choline (tCho) were 1-3%/2-6%, respectively. Statistically significant in vivo diurnal effects were only seen for glycerophosphocholine (GPC, + 10%, F= 10.6, p = 0.001) and Glx (+6%, F = 5.1, p = 0.018). The intra/inter-subject CVs for the major metabolites ranged from 2-5%/5-9%, respectively. The major metabolites displayed ICC ranging from 0.5-0.7 and low SEM (0.001-0.078) reflecting high reliability in detecting neurometabolites. The inter-week interval for in vivo measurements had minimal effect on metabolite ratios (F = 1.4, p = 0.09). Conclusion: In vitro MRS showed no diurnal effects and minimal variation in metabolite levels. Most PCG metabolites are not altered diurnally. The low in vivo variability of metabolite concentration supports the use of localised MRS on clinical 3 T scanners for reliable neurometabolic profiling of the brain.