INTRODUCTION

Functional magnetic resonance spectroscopy (fMRS) is a potential powerful tool to investigate the neurochemicals dynamic in response to various type of interventions and stimulations. However, there is limited consensus on the paradigms, stimulus domains, and response functions associated with neurochemical dynamics; studies vary in effect size, and there continues to be a discussion on suitable acquisition duration and analytical approach. We aim to inform these unknowns by performing a meta-analysis of existing ¹H-fMRS studies of GABA and Glutamate (either as Glu or as Glx; Glutamate+Glutamine) in the healthy brain. In this meta-analysis, we systematically summarise and analyse reported metabolite level changes based on fMRS paradigms and stimulus domains from existing studies.

METHODS

Data extraction
A systematic search was performed according to PRISMA guidelines¹ focused on functional MRS studies of GABA and Glu/Glx. Forty-nine eligible studies were included (Fig.1). Standardized mean differences in metabolite levels between paradigm conditions and 95% confidence intervals (Hedge’s G) was calculated from extracted data using the Meta-Essentials². MRS reporting quality assurance was performed following MRS-Q checklists^3,4.

Data analysis
Extracted effect sizes were pooled using the robust variance estimation method in R-package robumeta⁶ to account for non-independence of effect size as multiple experiments and time-points were reported in each individual study. Effect sizes were first pooled based on type of data extracted (mean or percentage change from the baseline (change scores))⁷, then divided into sub-groups based on paradigm (block or event-related), and stimulus domain. If only one study with multiple outcomes was included for subgroup analysis, the R-package meta⁸ was used. Results were considered statistically significant at p < 0.05, and at p < 0.01 when the degrees of freedom < 4 for RVE t-tests to account for Type-I error⁹.

RESULTS

Quality assessment
Most studies included were assessed to be of high quality through the MRS-Q (Fig.2A). 36.7% (n = 18/49) of the studies were identified as ‘low-quality’, mostly due to an insufficient number of transients or small voxel size (see Discussion). Fig. 2B and 2C summarize studies by brain region investigated for Glx/Glu and GABA and additionally reports on studies by stimulus domains.

Glu/Glx and GABA level across studies
Across 49 studies, mean Glu and Glx increased significantly [compared to the respective control condition] in fMRS studies (Hedge’s G = 0.371 and 0.292 respectively). The percentage change in Glu was positive on average (Hedge’s G = 0.473) (Fig.3A). We found no significant change in GABA across 20 included studies.


Glu/Glx and GABA level by type of paradigm
When effect sizes were differentiated between the type of paradigm regardless of brain region and stimulus domain, block designs showed lower variation in effect size and a significant overall positive change in Glu/Glx (Hedge’s G = 0.286 to 0.365). A reduction in GABA was observed for event-related designs (but not block designs); Hedge’s G = -0.756) (Fig.3B).


Glu/Glx and GABA level by stimulus domain
Figure 4 shows metabolite level changes compared between stimulation domains. Mean Glu/Glx changes were significantly positive during cognitive stimulation (Hedge’s G = 0.398). Percentage change in Glu/Glx were positive during learning (Hedge’s G = 0.287). Mean GABA level was positively increased during motor (Hedge’s G = 0.322) and exercise (Hedge’s G = 0.465), while percentage change of GABA was negative during transcranial direct-current stimulation (Hedge’s G = -0.122). Stress stimulation was associated with a significant negative change of both Glu/Glx (Hedge’s G = -0.413) and GABA (Hedge’s G = -0.869). There was no average significant change related to visual stimulation for both Glu/Glx and GABA.


Glu/Glx and GABA level by time course
Figure 5 illustrates effect sizes in relation to time of data acquisition during fMRS. The results show different fluctuation patterns in GABA/Glu/Glx for different stimulus domains relative to the time of stimulation.

DISCUSSION

Results show a wide range effect size variability and directionality in fMRS of GABA and Glu/Glx. The observed changes strongly depend on the type of paradigm and the stimulus domain. Time-resolved fluctuation patterns suggest a potential response delay for different brain regions or stimulus domains, and between GABA and Glu/Glx. This delay may explain the variation of effect sizes observed in event-related designs and associate with different parts of the neurometabolite response¹⁰, which may also explain differences in directionality of findings.

Currently available quality assurance and reporting metrics (MRS-Q and MRSinMRS) were designed for static MRS^3,11, and do not take into account functional approaches where a reduced number of transients for averaging are often used and modelled. Notably, many studies used small voxel sizes compared to consensus recommendation³, in particular for spectral editing of GABA¹², making the current findings less reliable. The variation in effect sizes observed is likely due to different approaches used, typical for an emerging technique such as fMRS.

CONCLUSION

Establishing fMRS paradigms and parameters that evoke metabolic responses with high reliability and reproducibility is of great interest to investigate healthy and atypical brain function. Here, we establish effect sizes and directionality of the GABA and Glu response in all currently available fMRS studies, with the aim to improve consensus and reproducibility of fMRS research and show the importance of standardised reporting and minimal best practice for fMRS.

Acknowledgements

The work was supported in part by Chiang Mai University, Chiang Mai, Thailand.

References

1. Tricco AC, Lillie E, Zarin W, et al. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med. 2018;169(7):467-473.
2. Suurmond R, van Rhee H, Hak T. Introduction, comparison, and validation of Meta-Essentials: A free and simple tool for meta-analysis. Res Synth Methods. 2017;8(4):537-553.
3. Peek AL, Rebbeck T, Puts NA, Watson J, Aguila MER, Leaver AM. Brain GABA and glutamate levels across pain conditions: A systematic literature review and meta-analysis of 1H-MRS studies using the MRS-Q quality assessment tool. NeuroImage. 2020;210:116532.
4. Puts N, Peek A. MRS-Q. Published online January 22, 2020.
5. Fisher Z, Tipton E. robumeta: An R-package for robust variance estimation in meta-analysis. ArXiv150302220 Stat. Published online March 7, 2015. http://arxiv.org/abs/1503.02220 Accessed October 31, 2021.
6. 9.4.5.2 Meta-analysis of change scores. https://handbook-5-1.cochrane.org/chapter_9/9_4_5_2_meta_analysis_of_change_scores.htm Accessed November 5, 2021.
7. Schwarzer G. meta: An R package for meta-analysis. R News. 2007;7(3):40-45.
8. Tanner-Smith EE, Tipton E, Polanin JR. Handling Complex Meta-analytic Data Structures Using Robust Variance Estimates: a Tutorial in R. J Dev Life-Course Criminol. 2016;2(1):85-112.
9. Moher D, Shamseer L, Clarke M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4(1):1.
10. Mullins PG. Towards a theory of functional magnetic resonance spectroscopy (fMRS): A meta-analysis and discussion of using MRS to measure changes in neurotransmitters in real time. Scand J Psychol. 2018;59(1):91-103.
11. Lin A, Andronesi O, Bogner W, et al. Minimum Reporting Standards for in vivo Magnetic Resonance Spectroscopy (MRSinMRS): Experts’ consensus recommendations. NMR Biomed. 2021;34(5):e4484.
12. Puts NAJ, Edden RAE. In vivo magnetic resonance spectroscopy of GABA: A methodological review. Prog Nucl Magn Reson Spectrosc. 2012;60:29-41.