Review article
Review of 1H magnetic resonance spectroscopy findings in major depressive disorder: A meta-analysis

https://doi.org/10.1016/j.pscychresns.2005.12.004Get rights and content

Abstract

In a review of the current literature, we identified 1H MRS studies of major depressive disorder (MDD) that examined the metabolites N-acetylaspartate (NAA), choline (Cho), myo-inositol (mI), glutamate/glutamine/γ-aminobutyric acid-GABA (Glx), and creatine (Cr). Separate meta-analyses comparing adult and pediatric MDD patients with healthy controls were performed. For adults, 14 studies with 227 patients/246 controls for NAA, 15 studies with 240 patients/261 controls for Cho, seven studies with 96 patients/104 controls for mI, six studies with 86 patients/109 controls for Glx, and nine studies with 146 patients/173 controls for Cr were identified. There were six studies containing a total of 79 pediatric depressed patients. We performed 15 separate meta-analyses to combine results from studies with similar characteristics. Adult MDD patients had higher Cho/Cr values than controls in the basal ganglia. In contrast, three studies on Glx levels indicated significantly lower Glx levels in the frontal lobe of MDD patients. The review indicated increased Cho/Cr in the basal ganglia in MDD and no alteration of NAA, suggesting an increased membrane turnover in MDD without a neurodegenerative outcome. Lower Glx levels in depressed patients in contrast to a likely hyperglutamatergic state in bipolar disorder may implicate a different pathophysiological ground in MDD.

Introduction

Unipolar or major depressive disorder (MDD) is a severe, prevalent, and often chronically disabling illness with a lifetime prevalence of 15–20% (Kessler et al., 1994, Nolan et al., 2002). About 25–50% of patients have a first degree relative with MDD, indicating genetically transmittable underlying brain pathology (Nolan et al., 2002). However, the nature of brain pathology or neural abnormality that precipitates MDD has yet to be identified. Through the use of magnetic resonance spectroscopy (MRS), differences in tissue concentration or resonance frequency of various chemical compounds in human brain can be detected in vivo. Among the several nuclei assessed in MRS examinations, proton, or [1H] MRS, is the most commonly used in investigations of the neurochemical basis of MDD (Chang et al., 2003).

N-Acetyl aspartate (NAA), 2.02 ppm, is the main peak in the 1H MRS spectra from the normal brain, and it constitutes approximately 3–4% of the total brain osmolarity. NAA is synthesized from acetyl coenzyme A and aspartate, primarily in neurons, by the mitochondrial enzyme l-aspartate N-acetyltransferase (Baslow, 2000). NAA functions in the brain as an acetyl donor for acetyl coenzyme A and takes part in lipid biosynthesis, including for myelin (Moore and Galloway, 2002). NAA is commonly considered to be a putative neuronal marker (Stanley, 2002). However, considering the high NAA levels in Canavan's disease (Matalon et al., 1988), which involves neurodegeneration of white matter, reversible decreases of NAA levels in neurological diseases involving white matter (De Stefano et al., 1995), and intercompartmental cycling of NAA between neurons and oligodendrocytes (Baslow, 2000), the role or interpretation of NAA may also reflect the formation and maintenance of myelin (Stanley, 2002). Although most studies have found negative results for the involvement of NAA in the pathophysiology of depression, there are some reports indicating an increase of NAA resonance with antidepressant treatment (Charles et al., 1994). In keeping with this, Van der Hart et al. (2002) reported significantly decreased concentrations of NAA in chronically stressed animals and reversal of that decrease upon treatment with antidepressants.

Choline (Cho), 3.23 ppm, is an essential precursor of the neurotransmitter acetylcholine and the membrane lipids, phosphatidylcholine and sphingomyelin (Kusumakar et al., 2001). Thus, Cho resonance is typically higher (30%) in white matter (Kusumakar et al., 2001). Cytosolic choline compounds such as glycerophosphocholine (GPC) and phosphocholine (PC) contribute as much as 50% to choline signal (Kusumakar et al., 2001). Thus, changes in Cho resonance are typically thought to be derived from changes in GPC and PC. The Cho peak is considered a potential biomarker for the status of membrane phospholipid metabolism (Glitz et al., 2002, Moore and Galloway, 2002), and an elevated Cho signal most likely reflects an increase in membrane turnover (Ende et al., 2000). Phosphatidylcholine, the major choline-containing metabolite of the normal brain, is MR invisible in myelin, cell membranes, and other brain lipids under normal circumstances. However, in cases of pathologies marked by membrane breakdown such as neurodegenerative states, bound Cho moieties may be liberated into the free Cho pool, thus contributing to an increase of this resonance (Ende et al., 2000, Glitz et al., 2002). Since GPC is a catabolite whereas PC is a precursor of phosphatidylcholine (Renshaw et al., 1997), and phosphatidylcholine is a source of secondary messenger diacylglycerol, which plays a crucial role in intracellular signal transduction (Kusumakar et al., 2001), alterations in Cho signal may have an impact on signal transduction in MDD (Steingard et al., 2000). Based on theories of cholinergic overactivity and depression (Janowsky et al., 1972) and of phosphatidylcholine/membrane phospholipids metabolism and signal transduction systems, the Cho peak in 1H MRS has received considerable attention in investigations of MDD.

The signal of myo-inositol (mI), 3.56 ppm, represents predominantly mI with minor contributions (< 5%) from glycine and inositol-1-phosphate (Moore and Galloway, 2002). It is a sugar involved in the regulation of neuronal osmolarity. Traditionally, mI is seen as a marker of glial cells (Frey et al., 1998). However, it has recently been demonstrated that high levels of mI are present in some types of neurons (Moore and Galloway, 2002), which challenges this notion. Inositol has a function in osmoregulation in brain glial cells (Frey et al., 1998). Glial cells may have a storage function for mI, which can then be gradually passed on to the neurons, where it becomes a precursor in the phosphoinositide cycle (Frey et al., 1998). In cerebrospinal fluid (CSF), markedly reduced levels of mI have been reported in depressed patients with unipolar or bipolar affective disorder (Barkai et al., 1978). Under double-blind conditions, the intake of inositol has been reported to lead to an improvement in depression (Levine et al., 1995).

At low field strength, the broad resonance centered at approximately 2.3 ppm contains overlapping resonances arising from glutamate (Glu)/glutamine (Gln)/gamma-aminobutyric acid (GABA), which are often indistinguishable (Glitz et al., 2002, Malhi et al., 2002). Glutamate and its N-methyl-d-aspartate (NMDA) receptors have been implicated in the pathophysiology of depression (Petrie et al., 2000, Pfleiderer et al., 2003).

GABA is the major inhibitory neurotransmitter in central nervous system and is integral to managing brain excitability (Chang et al., 2003). The 1H spectrum of GABA contains complex multiple peaks that overlap with Glu and Gln, with estimated proportions of human brain GABA relative to Glu and Gln around 1/9 and 1/6, respectively (Stanley, 2002). Consequently, the in vivo quantification of GABA has poor reliability, unless a spectral-difference or editing-type pulse sequence is used to isolate the GABA peak from other overlapping peaks (Stanley, 2002). GABA is synthesized in neurons from Gln with Glu as an intermediate step. Released GABA is taken up by glial cells and enters the tricarboxylic acid cycle, which eventually yields Glu, which is then converted to Gln. The glial Gln can then be transferred back into neurons for synthesis of new GABA (Chang et al., 2003).

Creatine (Cr)/phosphocreatine (PCr), 3.02 ppm, plays a central role in maintaining energy stores. The gray matter concentration of Cr is greater than that in white matter. Since the Cr resonance includes both Cr and PCr, it reflects total cellular creatine stores (Malhi et al., 2002). As its level is considered to be relatively constant, it has often been used as an internal standard for comparison (Malhi et al., 2002).

Although a significant amount of work using 1H MRS to investigate the neurobiology of MDD has been done over the last decade, studies have reported mixed results, thus preventing a definitive conclusion on the direction of alterations on individual neurochemicals. A meta-analysis is a systematic and quantitative approach for combining data and overcomes power limitations associated with small sample sizes. However, given the different characteristics of most of the published studies, with some on medicated subjects, in different brain regions, and with different acquisition parameters, significant limitations exist in combining studies in a meta-analysis. Yet, due to the exploratory nature of present MRS investigations, qualitative classification of studies according to the findings on individual metabolites would help to develop standards for future experiments in the field. The first generation data obtained so far with 1H MRS served to probe what type of alterations should be expected and in which brain regions. The next generation of 1H MRS studies should be driven by specific hypotheses in affected brain regions. Thus, the aim of this article is to review the current 1H MRS literature in MDD through the use of meta-analyses of the data on individual neurochemicals.

Section snippets

Selection of studies

A computer-assisted literature search of the National Library of Medicine's Medline–PubMed literature system between 1978 and November 2005 was conducted with ‘1H magnetic resonance spectroscopy’ and ‘unipolar/major depressive/affective disorder/illness’ as the key words, and complemented by a manual search of bibliographic cross-referencing. In vivo 1H MRS studies were included in the data documentation/analysis if the following criteria were met: the report has been published in English and

NAA

The inclusion process identified 14 studies (one study provided data for both euthymic and depressed states; two studies investigated two different brain regions) on the NAA resonance in adult patients with MDD (N = 227) as compared with healthy controls (N = 246) (Table 1a). In 2 of the 14 studies, both unipolar and bipolar depressed patients (Auer et al., 2000, Michael et al., 2003b), and in one study, unipolar patients in both the depressed and euthymic states (Brambilla et al., 2005b) were

Discussion

In this review MDD patients were found to have similar NAA/Cr and NAA values to those of healthy controls in the basal ganglia and frontal lobes. This finding parallels the observation of normal neuronal cell numbers in MDD (Pfleiderer et al., 2003). An unaltered NAA value in MDD has different implications than for bipolar disorder (BD) since accumulated data suggest lower NAA levels in euthymic bipolar patients in the frontal lobe structures and hippocampus (Brambilla et al., 2005a,

References (58)

  • V. Gallo et al.

    Glutamate receptors in glia: new cells, new inputs and new functions

    Trends in Pharmacological Sciences

    (2000)
  • D.S. Janowsky et al.

    A cholinergic–adrenergic hypothesis of mania and depression

    Lancet

    (1972)
  • R.C. Kessler et al.

    Sex and depression in the national comorbidity survey: II. Cohort effects

    Journal of Affective Disorders

    (1994)
  • J. Levine et al.

    Increased cerebrospinal fluid glutamine levels in depressed patients

    Biological Psychiatry

    (2000)
  • Y. Mirza et al.

    Reduced anterior cingulate cortex glutamatergic concentrations in childhood major depression

    Journal of the American Academy of Child and Adolescent Psychiatry

    (2004)
  • G. Nowak et al.

    Alterations in the N-methyl-d-aspartate (NMDA) receptor complex in the frontal cortex of suicide victims

    Brain Research

    (1995)
  • R.X. Petrie et al.

    The N-methyl-d-aspartate receptor, synaptic plasticity, and depressive disorder: a critical review

    Pharmacology and Therapeutics

    (2000)
  • F. Petty

    GABA and mood disorders: a brief review and hypothesis

    Journal of Affective Disorders

    (1995)
  • B. Pfleiderer et al.

    Effective electroconvulsive therapy reverses glutamate/glutamine deficit in the left anterior cingulum of unipolar depressed patients

    Psychiatry Research: Neuroimaging

    (2003)
  • G. Rajkowska

    Postmortem studies in mood disorders indicate altered numbers of neurons and glial cells

    Biological Psychiatry

    (2000)
  • G. Rajkowska et al.

    Reductions in neuronal and glial density characterize the dorsolateral prefrontal cortex in bipolar disorder

    Biological Psychiatry

    (2001)
  • P.F. Renshaw et al.

    Proton MRS of basal ganglia in major depression

    Biological Psychiatry

    (1994)
  • P.F. Renshaw et al.

    Basal ganglia choline levels in depression and response to fluoxetine treatment: an in vivo proton magnetic resonance spectroscopy study

    Biological Psychiatry

    (1997)
  • J.C. Soares et al.

    The anatomy of mood disorders – review of structural neuroimaging studies

    Biological Psychiatry

    (1997)
  • R.J. Steingard et al.

    Increased orbitofrontol cortex levels of choline in depressed adolescents as detected by in vivo proton magnetic resonance spectroscopy

    Biological Psychiatry

    (2000)
  • M. Vythilingam et al.

    Focal and lateralized subcortical abnormalities in unipolar major depressive disorder: an automated multivoxel proton magnetic resonance spectroscopy study

    Biological Psychiatry

    (2003)
  • A.I. Barkai et al.

    Reduced myo-inositol levels in cerebrospinal fluid from patients with affective disorder

    Biological Psychiatry

    (1978)
  • M.H. Baslow

    Functions of N-acetyl-l-aspartate and N-acetyl-l-aspartylglutamate in the vertebrate brain: role in glial cell specific signaling

    Journal of Neurochemistry

    (2000)
  • N. Binesh et al.

    Neurochemistry of late-life depression: a pilot two dimensional MR spectroscopic study

    Journal of Magnetic Resonance Imaging

    (2004)
  • Cited by (207)

    • Brain temperature as an indicator of neuroinflammation induced by typhoid vaccine: Assessment using whole-brain magnetic resonance spectroscopy in a randomised crossover study

      2022, NeuroImage: Clinical
      Citation Excerpt :

      The lack of NAA/CR alteration post-vaccine suggests that, like LPS, typhoid vaccine does not affect neuronal morphology. The absence of a treatment effect on NAA/CR is consistent with studies of MDD, where normal neuronal cell numbers were observed and there was no significant alteration in NAA (Yildiz-Yesiloglu and Ankerst, 2006). The metabolite alterations in CHO/CR and MI/CR, and absence of a change in NAA/CR, suggests that typhoid vaccine may elicit a similar inflammatory response to that observed in MDD.

    View all citing articles on Scopus
    View full text