Chronic pain is in many cases associated with a poor quality of life and high health care costs. Due to the multimodal disease origin, therapeutic methods are often rather unspecific. By applying modern functional imaging methods, neuronal dysfunction has been recently identified in patients and has been hypothesized to reflect the disease related central sensitization of neuronal pain processing¹. However, the underlying neurochemical processes, especially changes in neurotransmitter turnover, remain widely unexplored although their understanding might be essential to improve current therapeutic approaches². ¹H-MR spectroscopy allows in vivo quantitation of the most important neurotransmitters glutamate and GABA, and provides direct insight into chronic pain related neurochemical changes³. In the present study, GABA and glutamate were quantified in three different pain processing regions of the brain of patients with unspecific chronic low back pain and compared to healthy subjects in order to investigate potentially altered regional and interregional neurochemical interrelations.13 patients (12f/1m) and 13 age and gender matched healthy subjects (54 ± 8 years) participated in this study. Patients underwent detailed clinical evaluation to assess their chronic pain state and accompanying effects. All measurements were performed on a whole-body 3T MR scanner (Siemens, Germany). High resolution anatomic 3D MRI data (MP-RAGE, TR/TE/TI = 2300/3.03/900 ms; isotrop 1 mm³ voxels with 256×256 matrix) were used to select spectroscopic voxels (Fig. 1) in the anterior cingulate cortex (aCC, 13 ml), left insula (Ins, 13 ml) and posterior cortex (PC, 10 ml). In each voxel, conventional PRESS (TR/TE: 1800/30 ms, TA = 1 min) and spectrally edited MEGA-PRESS (TR/TE: 1800/68 ms, TA = 11 – 13 min) ¹H-MR spectra were collected (Fig. 2). Sum intensities of glutamate and glutamine (Glx) were quantified from conventional spectra by using the LCModel⁴, whereas the GABA intensities were determined from the edited spectra with the jMRUI package⁵. Subsequently, all intensities were normalized with the corresponding creatine intensity. Regional and interregional interrelations were determined by means of Pearson correlations between GABA and Glx within as well as between the examined brain regions and compared between the two groups.Fig. 3 shows the correlation matrix (a) with significant (p < 0.05) regional and interregional correlations of neurotransmitters within and between the brain regions in healthy controls as well as selected data distributions (b–f) with significant regional and interregional Glx and GABA correlations obtained in healthy controls (in black) as well as corresponding data distributions in patients (in red). Interestingly, in contrast to patients, who did not show any significant associations between investigated metabolites, healthy subjects had a negative correlation between Glx and GABA in the aCC (rho = -0.6). Furthermore, Glx in Ins and PC of controls was positively associated with Glx in aCC (rho = 0.65). Finally, GABA in the aCC of healthy persons was negatively correlated with Glx in Ins (rho = -0.72) as it was also with Glx in the PC (rho = -0.58).To our knowledge, this study shows for the first time the differences in interregional neurotransmitter associations between healthy subjects and patients suffering from chronic pain. The identified regional and interregional interrelations in healthy controls are expected, since they are related to the neurochemical regularization of functional pathways within and between the examined brain regions⁶. On the other hand, the lack of these correlations in chronic pain patients may reflect disordered functional connections and thus may be associated with the progressing central sensitization of the cerebral pain processing. However, it remains an open question, which has to be directly investigated in the future studies with combination of ¹H-MRS and functional MRI in chronic pain patients.