Intervertebral discs (IVDs) are fibrocartilaginous cushions that act to protect both the vertebrae and the brain by working as a shock-absorbing system for the spine. IVDs are composed of two components, the annulus fibrosus (a strong outer layer) that encloses the nucleus pulposus (containing a hydrated gel–like material that is resistant to compression). Proteoglycan molecules (PGs) are an important component of IVDs, found in both the annulus fibrosis and nucleus pulposus. PGs serve to both attract and retain water, contributing to the hydration of the tissue, and also serve the purpose of regulation of matrix assembly. The loss of PG from the disc is the main chemical change observed in intervertebral disc degeneration [1]. Therefore, it is of clinical interest to be able to reliably measure PG on a clinical MR scanner. The purpose of this study was to assess the feasibility of reliably quantifying PG concentration using 1H magnetic resonance spectroscopy (1HMRS) at 1.5T, and to determine how this relates to both the disc height, disc volume and the T2* values for the disc.Scans were performed on a group of 13 healthy volunteers (age = 27.5 ± 3.3 yo; 8F, 5M) using a 1.5T Siemens Avanto with the inbuilt spine coil. All participants gave written informed consent and the study was approved by the local research ethics committee. 1HMRS was performed using the following parameters: TE = 30ms, TR = 1500ms, and 256 averages. Voxels were 25mmx20mm in the axial plane and 8mm in thickness, contained fully within the disc with minimum infiltration of signal from neighbouring vertebrae (figure 1). Saturation bands were placed surrounding the voxel to reduce signal from external tissues. Voxels were all placed within lumbar discs, either between L3 and L4 or L4 and L5, whichever was the thickest. Water unsuppressed data was also collected (16 averages) to allow for eddy current correction and quantification. Shimming was performed manually, to obtain the best possible line-width. The total acquisition time was approximately 7 minutes. Peaks arising from the N-acetyl resonance (2.04ppm) associated with PG were fitted, using the jMRUI processing tool with AMARES [2]. Attempts were also made to fit peaks associated with lipids and macromolecules (1–1.5ppm, LMM13) and carbohydrates (3.54ppm). LMM13 peaks are likely to be due to signal infiltration from outside the voxel, but could interfere with measurement of PG if very large. Additional multi-echo gradient-echo scans were also performed to measure T2* of the selected disc, to determine whether any correlation can be seen with PG. All the discs were measured for both thickness and overall volume; this was also assessed for correlation. All results are quoted as mean±sd.Thirteen participants were scanned as part of this study, all had corresponding volumetric measures and nine had T2* values. The quality was acceptable in all spectra, with an average line width of 8.0±2.0Hz and SNR of 11.2±6.8. All of the spectra showed a clear singlet at 2.04ppm relating to PG (figure 2). Three of the data sets did show LMM13 contamination, however PG was still measureable in these spectra. Although, most of the spectra exhibited a peak for the carbohydrates, this is not reliably measured due to residual saturation from the water suppression. Zuo et al [3] showed that PG could be reliably measured using 1HMRS at 3T, using Pfirrmann grading as a variable. Controls with no discography (Pfirrmann grade 1) showed values for water/PG of 299±62. In the study presented here, we obtained a value of 308.8±59.9. This agreement indicates that it is feasible to reliably measure PG concentrations at 1.5T. The majority of clinical MR scanners are 1.5T, and not 3T. Therefore, in order for it to be of use clinically, it is necessary to prove 1HMRS of IVDs is robust at 1.5T. A robust method of PG quantification at 1.5T has been achieved here. Mean values for the other measure were as follows: Disc thickness = 10.0±1.0mm, Disc volume = 12.4±2.6cm³, T2* = 50.9±7.9. Correlations were sought between PG concentration and: age, disc thickness, disc volume and disc T2*. There was no correlation between disc volume or disc thickness and PG concentration, although a trend was emerging between PG concentration and T2* (R² = 0.57). This was limited by number of data sets available, and the sample size would need to be significantly increased in order to accurately determine whether there is any correlation present.It has been shown that PG can be reliably quantified in IVDs of healthy controls at 1.5T.