7 Tesla magnetic resonance spectroscopy estimates of GABA concentration relate to physiological measures of tonic inhibition in the human motor cortex

GABAergic neurotransmission within the cortex plays a key role in learning and is altered in several brain diseases. Quantification of bulk GABA in the human brain is typically obtained by magnetic resonance spectroscopy (MRS). However, the interpretation of MRS-GABA is still debated. A recent mathematical simulation contends that MRS detects extrasynaptic GABA, mediating tonic inhibition. Nevertheless, no empirical data have yet confirmed this hypothesis. Here we collected ultra-high-field 7 Tesla MRS and transcranial magnetic stimulation coupled with high-density electroencephalography (TMS-hdEEG) from the motor cortex of 20 healthy participants (age 23.95 ± 6.4 years), while they were at rest. We first applied a neural mass model (NMM) to TMS-evoked potentials to disentangle the contribution of different GABAergic pools. We then assessed to which of these different pools MRS-GABA was related to by means of parametric empirical Bayesian (PEB) analysis. We found that MRS-GABA was mostly positively related to the NMM-derived measures of tonic inhibition and overall functionality of the GABAergic synapse. This relationship was reliable enough to predict MRS-GABA from NMM-GABA. These findings clarify the mesoscopic underpinnings of GABA levels measured by MRS. Our work will help fulfil the promises of MRS-GABA, enhancing our understanding of human behaviour, brain physiology and pathophysiology. KEY POINTS: GABA neurotransmission is essential for synaptic plasticity and learning (especially motor learning) and is altered in several brain disorders, such as epilepsy and stroke. Quantification of GABA in the human brain is typically obtained by magnetic resonance spectroscopy (MRS). However, the interpretation of MRS-GABA is still debated. By using a biophysical neural mass model, here we show that MRS-GABA relates to physiological measures of tonic inhibition in the human cortex.