Proton magnetic resonance spectroscopy (¹H-MRS) is a noninvasive and efficient neuroimaging technique, which quantitatively measures metabolite changes associated with Alzheimer’s disease (AD)-type pathology (Kantarci et al., Reference Kantarci2008; Ross et al., Reference Ross, Bluml, Cowan, Danielsen, Farrow and Tan1998). ¹H-MRS metabolite changes have been shown to associate with neuropathological changes in AD such as amyloid-β plaque and neurofibrillary tangle pathologies (Kantarci et al., Reference Kantarci2008; Murray et al., Reference Murray2014).

In this issue of International Psychogeriatrics, Motegi et al. (Reference Motegi2019) report the glutamine+glutamate (Glx) and gamma-aminobutyric acid (GABA) concentrations in the medial prefrontal cortex (mPFC) and posterior cingulate cortex (PCC) using MRS in cognitively healthy elderly individuals. The authors choose these specific regions to study with single-voxel spectroscopy based on prior work showing alterations in Glx and GABA in these regions due to aging, mild cognitive impairment (MCI), and AD. Furthermore, they investigate the association of MRS findings with diffusion tensor MRI (DTI)-based mean diffusivity (MD) measurements of cortical microstructural integrity in order to determine the basis of Glx alterations. They report that disruption of microstructural organization in the gray matter determined by the elevation in MD is accompanied by the reduction in Glx levels.

Glutamate is the primary excitatory neurotransmitter of the central nervous system. Excess glutamate is recycled by astrocytes from the synapse after being transformed to glutamine. Glutamine is then sent back to the presynaptic neuron for glutamate synthesis. This process is external energy dependent but acts mostly as a closed circuit with minimal amounts of glutamate or glutamine released into the bloodstream (Nedergaard et al., Reference Nedergaard, Takano and Hansen2002). Therefore, total glutamate and glutamine levels in a brain region would be constant as long as the integrity of the neuron-astrocyte complex and associated blood–CNS barrier is preserved. As a feature of neurodegeneration, in prolonged excitation with suboptimal astrocyte compensation, excitotoxicity and postsynaptic neuronal death can ensue. If the neuron-astrocyte complex integrity is significantly disrupted, the total glutamate levels are expected to decrease ultimately followed by atrophy. Thus, a decrease in glutamate, which makes up most of the Glx signal, is expected to precede macrostructural changes measured with atrophy on structural MRI. However, a decrease in Glx may be associated with the microstructural alterations in the gray matter as demonstrated in the current study.

Previous studies have shown that Glx concentration is reduced in patients with amnestic MCI (Riese et al., Reference Riese2015) and AD (Antuono et al., Reference Antuono, Jones, Wang and Li2001; Hattori et al., Reference Hattori, Abe, Sakoda and Sawada2002; Walecki et al., Reference Walecki, Barcikowska, Cwikla and Gabryelewicz2011). With a newly optimized semi-Localization with Adiabatic Selective Refocusing spectroscopy (sLASER) ¹H-MRS technique (Oz and Tkac, Reference Oz and Tkac2011), it was possible to isolate glutamate and glutamine peaks in the 3T MR spectra. Using sLASER MRS, glutamate concentration was found to be lower in the PCC in patients with amnestic MCI compared to cognitively unimpaired individuals (Zeydan et al., Reference Zeydan2017). Hence, it is likely that the reduced Glx signal identified in the current study is mostly due to a reduction in glutamate.

In the current study, Motegi et al. reported that higher cortical MD on DTI in both mPFC and PCC was associated with lower Glx levels in cognitively unimpaired older individuals (Motegi et al., Reference Motegi2019). Elevated MD on DTI suggests neuronal loss associated with neurodegenerative pathologies in older adults without dementia (Kantarci et al., Reference Kantarci2011). Such pathologies may include AD-related tau neurofibrillary tangle and β-amyloid pathologies, transactive response DNA binding protein 43 kDa (TDP-43), and α-synuclein pathologies that are common and may be overlapping in older adults. While this study demonstrates that reduced Glx measured by MRS is a cellular marker of neurodegeneration even in cognitively unimpaired older individuals, it does not relate these changes to a specific neurodegenerative pathologic process. Based on the fact that AD-related pathology is the most common pathology observed in older adults, we can assume that a majority of these alterations would be explained with AD biomarkers. For example, reduced glutamate in cognitively impaired elderly with MCI was associated with β-amyloid deposition on Pittsburgh compound-B positron emission tomography (Zeydan et al., Reference Zeydan2017) and MD on DTI was associated with the neurofibrillary tangle tau pathology at autopsy within the AD spectrum (Kantarci et al., Reference Kantarci2017). Thus, future studies with AD biomarkers may explain the basis of the relationship between these biochemical and structural alterations more clearly.

The results of the current study are intriguing, since the basis of Glx alterations in neurodegenerative diseases is relatively unknown. The current study demonstrates that the disruption of gray matter microstructure is associated with changes in glutamate+glutamine metabolism, making Glx concentration a potential biomarker for early detection of neurodegeneration.