Author's Accepted Manuscript

Glutamate level in anterior cingulate predictsanxiety in healthy humans: A magneticresonance spectroscopy study

Shilpi Modi, Poonam Rana, Prabhjot Kaur,Nisha Rani, Subash Khushu

PII: S0925-4927(14)00056-0DOI: http://dx.doi.org/10.1016/j.pscychresns.2014.03.001Reference: PSYN10178

To appear in: Psychiatry Research: Neuroimaging

Received date: 7 August 2013Revised date: 28 January 2014Accepted date: 7 March 2014

Cite this article as: Shilpi Modi, Poonam Rana, Prabhjot Kaur, Nisha Rani,Subash Khushu, Glutamate level in anterior cingulate predicts anxiety inhealthy humans: A magnetic resonance spectroscopy study, PsychiatryResearch: Neuroimaging, http://dx.doi.org/10.1016/j.pscychresns.2014.03.001

This is a PDF file of an unedited manuscript that has been accepted forpublication. As a service to our customers we are providing this early version ofthe manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting galley proof before it is published in its final citable form.Please note that during the production process errors may be discovered whichcould affect the content, and all legal disclaimers that apply to the journalpertain.

www.elsevier.com/locate/psychresns

Glutamate level in anterior cingulate predicts anxiety in healthy humans: A magnetic

resonance spectroscopy study

Shilpi Modi, Poonam Rana, Prabhjot Kaur, Nisha Rani, and Subash Khushu*

NMR Research Centre, Institute of Nuclear Medicine and Allied Sciences (INMAS), Lucknow Road, Timarpur, Delhi, India

*Corresponding author.

Dr. Subash Khushu,

Scientist ‘G’ and Head,

NMR Research Center, INMAS,

Lucknow Road, Timarpur,

Delhi -110054, India.

Tel.: +91-11-23905313; Fax: +91-11-23919509

E-mail address: skhushu@inmas.drdo.in

Glutamate level in anterior cingulate cortex predicts anxiety in healthy humans: A

magnetic resonance spectroscopy study

Abstract

Anxiety, a personality dimension in healthy humans, has been found to be associated with

many functional consequences such as increased distractibility and attentional bias in favour

of threat-related information, along with morphological and microstructural changes in the

brain. The associated metabolic/ neurochemical alterations are sparsely studied. In the present

magnetic resonance spectroscopy (MRS) study, we investigated the possible relationship

between regional brain chemistry within anterior cingulate cortex (4-cm3 voxel) and

hippocampus (2.5-cm3 voxel) and anxiety (measured by State-Trait Anxiety Inventory) in our

subject group. In the anterior cingulate cortex, multivariate analysis of covariance showed an

increase in myo-inositol and combined glutamate and glutamine levels in the high anxiety

subject group as compared with the low anxiety group. In the partial correlation analysis

between neurochemicals and anxiety, glutamate and combined glutamate and glutamine also

showed a predictive value for anxiety. On analysing the trait anxiety sub-score separately, we

found glutamate, inositol and combined glutamate and glutamine levels to be increased in the

high trait anxiety group as compared with the low trait anxiety group. All three resonances

also had a predictive value for trait anxiety. In the hippocampus, none of the neurochemicals

showed significant difference between high and low anxiety groups. The study provides a

first account of alterations in anterior cingulate cortex neurochemistry in relation to anxiety in

healthy subjects. The study thus contributes to the limited literature available on altered

metabolism and neural mechanisms underlying sub-clinical anxiety.

Keywords: Trait anxiety; State anxiety; Neuroimaging; Hippocampus

1. Introduction

Anxiety is an aversive emotional and motivational state that can be characterized by harm

avoidance behaviour such as worrying, irritability, and difficulty in relaxing, or by a

predisposition to interpret ambiguous situations as threatening (Grachev and Apkarian, 2000;

Eysenck et al., 2007). These effects are seen not only in clinical anxiety but also within the

normal population as a personality dimension that is generally assessed by measures of

anxiety such as Spielberger’s State-Trait Anxiety Inventory (STAI) (Spielberger, 1983;

Eysenck et al., 2007). The STAI consists of two scales – Y2, which measures the

participant’s disposition to anxiety (trait), and Y1, which measures how anxious the

participant feels at the moment (state). The model proposed by Sandi and Richter-Levin

(Sandi and Richter-Levin, 2009) supports the etiological hypotheses that the high anxiety trait

is a vulnerability factor to develop both depression and anxiety disorders. Therefore,

identifying hallmarks of anxiety to identify individuals that are at risk for the development of

clinical anxiety disorders and depression becomes important, so that timely preventive

interventions may be given to them. Both state and trait anxiety have been found to be

associated with many functional consequences such as, increased distractibility, attentional

bias in favour of threat-related information and a hyper-responsive amygdala even for

unattended threat-related stimuli (Bishop et al., 2004; Eysenck et al., 2007; Bishop, 2009).

Similarly, morphological and microstructural changes have been reported to be associated

with anxiety-related personality traits (Pujol et al., 2002; Westlye et al., 2011; Baur et al.,

2012; Lochner et al., 2012; Montag et al., 2012; Modi et al., 2013). Pujol et al. (2002)

showed that a large right anterior cingulate cortex is related to a temperamental disposition to

fear and anticipatory worry. Similarly, in another volumetric study, a positive correlation was

obtained between trait anxiety and the volume of the left amygdala and right hippocampus

(Baur et al., 2012). Lochner et al. (2012) found white matter abnormalities in the anterior

limb of the internal capsule (ALIC) and cingulum that were related to obsessive-compulsive

disorder (OCD). Similarly, increased harm avoidance behaviour has been shown to be

associated with decreased white matter (WM) microstructure in major WM tracts such as

corticolimbic pathways, the anterior thalamic radiations, inferior fronto-occipital cortex and

right superior longitudinal fasciculus (Westlye et al., 2011). Individual differences in negative

emotionality have also been found to be positively correlated with the fractional anisotropy

(FA) values of several WM tracts in the temporal lobe of the left hemisphere (Montag et al.,

2012). A recent diffusion tensor tractography study showed a positive correlation between

trait anxiety scores and the mean FA value in the fornix and left uncinate fasciculus (Modi et

al., 2013).

The functional, morphological and microstuctural changes associated with anxiety might also

be related to alterations in regional chemistry of the brain. In vivo proton magnetic resonance

spectroscopy (1H-MRS) has been used primarily to measure concentrations of

neurochemicals in brain tissue such as N-acetyl-l-aspartate (NAA), which is a marker of

neuronal viability (Miller, 1991); glutamate (Glu) and glutamine (Gln), which are excitatory

neurotransmitters (Whiteside et al., 2006); choline (Cho), which is an indirect marker of

myelination and cell membrane metabolism (Miller, 1991); myo-inositol (mI), which is

involved in phospholipid metabolism (Whiteside et al., 2006); and creatine (Cr), which is a

marker of cellular energetics and commonly used as a reference level (Miller, 1991;

Whiteside et al., 2006). In vivo MRS studies on the neurobiology of anxiety-related disorders

are relatively few and confined to DSM-IV anxiety disorders like panic disorder (PD) (Dager

et al., 1999), OCD (Whiteside et al., 2006), posttraumatic stress disorder (PTSD)

(Mahmutyaz�c�o�lu et al., 2005; Kimbrell et al., 2005), and social anxiety disorder (SAD)

(Phan et al., 2005). MRS studies in non-clinical populations in which trait or state anxiety is

assessed as a personality dimension are very few. The first ever MRS study examining the

relationship between regional brain chemistry within the orbital frontal cortices and

physiologic anxiety in normal subjects demonstrated an increase in chemical concentrations

of most of the metabolites studied in the group with higher levels of anxiety as compared

with the lower anxiety group. Out of all the metabolites studied, NAA was identified as the

strongest chemical marker for anxiety (Grachev and Apkarian, 2000). The authors explained

these findings on the basis of a chemical-behavioral network in the brain of healthy subjects

as a possible mechanism for the development of anxiety. They proposed that the genetic

predisposition for anxiety may be linked to the number of neurons and their receptors in the

brain. An increase in the number of axons and synaptic connections across anxiety-related

regions (i.e., sprouting) might result in an increase in the concentration of NAA (Grachev and

Apkarian, 2000). In another study on unmedicated OCD patients, Fan et al. (2010) also

reported a correlation between the NAA/Cr ratio in the medial prefrontal cortex and trait

anxiety scores on the STAI among the control group.

Besides the orbital frontal cortex, the anterior cingulate cortex (ACC) and the hippocampus

have been previously implicated in the pathophysiology of several psychiatric disorders, like

bipolar disorder, depression, and schizophrenia, as well as anxiety disorders such as PTSD,

OCD, and SAD (Heckers et al., 2001; Sala et al., 2004; Mahmutyaz�c�o�lu et al., 2005; Phan

et al., 2005; Whiteside et al., 2006; Hong et al., 2007; Hermens et al., 2012). The anterior

cingulate cortex plays a role in the modulation of conditional fear responses and is critically

involved in performance monitoring and cognitive control. It also has a key role in emotional

and social behaviour (Mahmutyaz�c�o�lu et al., 2005; Etkin et al., 2011; Hermens et al.,

2012). The hippocampus is also one of the most important structures of the limbic system,

being involved in memory functions, processing of emotional information, and stress

response via the hypothalamo-pituitary-adrenal axis (Mahmutyaz�c�o�lu et al., 2005). In the

study by Grachev and Apkarian (2000), the authors did not find any relation between

neurochemistry in the cingulate cortex and anxiety. They attributed this negative finding to

the fact that only the middle portion of cingulate area was studied. Instead, the anterior

portion is considered to be an affective area and implied in a variety of emotional tests such

as emotional conflict regulation (Egner et al., 2008), reappraisal (Etkin et al., 2011), affect

labeling of emotional faces (Liebermen et al., 2007), and self-distraction from a fear-

conditioned stimulus (Delgado et al., 2008). Therefore, in the present study, we examined the

relationship between regional brain neurochemicals in the anterior cingulate cortex and the

anxiety levels of healthy subjects using in vivo 1H-MRS. In addition, because of the

relationship between the hippocampus and anxiety, the hippocampus was chosen as another

region of interest (ROI) for MRS. Spectroscopic findings in the anterior cingulate cortex and

the hippocampus in relation to anxiety levels of healthy subjects have the potential to

improve our understanding of the neurobiology of sub-clinical anxiety. Given the limited

literature available on the relationship between anxiety in healthy populations and regional

brain chemistry, we did not restrict our analyses to a particular neurochemical; rather, we

analysed all the possible metabolites.

2. Methods

2.1. Subjects

Twenty-four right-handed, healthy, educated participants drawn from the institute (males =

11, females = 13, mean age = 23.16 years, SD = 2.20 years) were screened for current or past

medical and psychiatric illness using the Medical History Questionnaire of the Hindi version

of the Diagnostic Interview for Genetic Studies (version 2) (Deshpande et al., 1998). None of

the subjects recruited for the study had any clinical evidence of stroke, head injury,

cardiovascular diseases, history of alcohol or drug dependence, hypertension, neurological or

psychiatric disorder or sensori-cognitive impairment; nor did they have any cortical

infarctions on the T2-weighted magnetic resonance images. In the case of females, care was

taken that while undergoing spectroscopy session they were not in their premenstrual or

menstrual periods. The procedure followed in the current study was in accordance with the

guidelines of the ethical committee of the Institute. Further, all subjects gave their consent to

participate in the study, and the procedure was thoroughly explained to them. After the

spectroscopy session, participants’ state-anxiety (inside the scanner) and trait-anxiety levels

were assessed using the STAI self-report questionnaires for adults (Spielberger, 1983).

Following Grachev and Apkarian (2000), the scores of state and trait anxiety were added

together to obtain the total anxiety scores of the participants. To evaluate the current

depression levels, participants were also asked to complete the Beck Depression Inventory

(BDI) (Beck et al., 1996).

2.2. Localized invivo single voxel 1H-MRS

The study was carried out using a 3-T whole body MR system (Magnetom Skyra, Siemens,

Germany) with a 32-channel head coil. The subject’s head was immobilized using

expandable ear cushions. Anatomical imaging was performed in all three orthogonal planes

for positioning the MRS voxels. T2-weighted multislice images (repetition time (TR) = 5600

ms, echo time (TE) = 100 ms, number of excitations = 1, 312×512 matrix, field of view =

180×220 mm, 25 slices, slice thickness = 4.0 mm, distance factor = 1.2 mm) covering the

entire brain were obtained.

For the anterior cingulate cortex, a 20 × 20 × 10 mm3 voxel was positioned in the midline

(covering samples from both the right and left sides of the brain) on the axial slice in which

the caudate nucleus was well formed visually. The voxel posteriorly abuts the genu of the

corpus callosum. For the hippocampus, axial oblique sections were obtained by planning

slices parallel to the body of the hippocampus in a parasagittal section; coronal oblique

sections were obtained by positioning slices perpendicular to the body of the hippocampus. A

voxel of 25 × 10 × 10 mm3 was then positioned on these axial oblique, coronal oblique, and

sagittal sections for better delineation and coverage of the hippocampus. Representative voxel

placement are shown in Fig. 1a and Fig. 1b.

MRS was obtained using a PRESS (point-resolved spectroscopy) sequence with the

following acquisition parameters: TR/TE = 2000 ms/33 ms; 2048 spectral points; 1200-Hz

spectral bandwidth and 196 averages. Automated global shimming was used to minimize the

B0 inhomogeneties, and localized shimming was done to further minimize B0 field variations

over the voxel of interest. Unsuppressed water (with 10 averages) spectra were also acquired

immediately before the water-suppressed metabolic acquisition and was used for spectral

quantifications.

The spectra were processed using LCModel software (Provencher, 1993). Only the

metabolites/neurochemicals with Cramer-Rao Lower Bounds less than 20% were analysed.

Relative concentrations of Glu, mI, NAA, combined peak of glycerophosphocholine and

choline (tCh = total choline), combined peak of NAA and N-acetylaspartyl-glutamate

(NAA+NAAG), and combined peak of glutamate and glutamine (Glu+Gln) relative to total

creatine (tCr, creatine + phosphocreatine) were measured. The tCr values were used as

reference because the total amount of Cr is a measure of general brain metabolism and

appears to be stable in subjects who are naïve to psychotropic medications (Keshavan et al.,

1991). Representative spectra from the anterior cingulate cortex and the hippocampus are

shown in Fig. 2a and Fig. 2b, respectively.

2.3. Statistical analysis

Controlling for the effects of age, sex, sub-clinical depression (BDI scores) and spectral

quality, anxiety group (high anxiety (total STAI score = 70–95) vs. lower anxiety (total STAI

score = 40–69)) and metabolite ratios were analyzed in both the ROIs with multivariate

analysis of covariance (MANCOVA) using the general linear model (SPSS (version 15.0,

SPSS Inc., Chicago, IL, USA) statistical software). To control for the effect of age, sex, sub-

clinical depression and spectral quality on the observed spectral differences between the two

groups, age, sex, BDI scores, full width at half maximum (FWHM) and signal-to-noise ratio

(SNR) of the spectra were taken as covariates of no interest. Adjustment for multiple

comparisons was done by applying Bonferroni correction. To determine the predictive value

of anxiety on neurochemical concentration, a partial correlation analysis (one-tailed)

including age, sex, sub-clinical depression and spectral quality (FWHM and SNR) as

covariates of no interest was computed between total anxiety scores and the neurochemical

ratios in both the ROIs, with the assumption that there was no correlation between

neurochemical ratios and anxiety (Ho=0). Alternatively, if a correlation of greater than 0.001

is observed at �=0.05 and 90% power of the test, the null hypothesis was rejected. P values of

� 0.05 were considered to be significant. A separate analysis for the trait anxiety sub-score of

the STAI was also carried out.

3. Results

3.1. Behavioral data

Descriptive statistics for the self-report measures are as follows: state anxiety score (STAI-

Y1) = 35.21 ± 9.16 (range 21-48), trait-anxiety score (STAI-Y2) = 37.21 ± 10.07 (range 20-

52), STAI total anxiety = 72.42 ± 16.42, BDI = 6.17 ± 4.51 (range 0-13). Subjects with total

STAI scores between 40 and 69 were considered as the low anxiety group (in ACC: males =

6, females = 5, mean age = 23.91 years, SD = 2.34 years; in hippocampus: males = 6, females

= 4, mean age = 23.50 years, SD = 2.01 years), and those with total STAI scores above 70

were considered as the high anxiety group (in ACC: males = 5, females = 8, mean age =

22.54 years, SD = 1.94 years; in hippocampus: males = 3, females = 8, mean age = 22.27

years, SD = 1.95 years) (Grachev and Apkarian, 2000). There was no significant difference

between the BDI scores of the two groups (high anxiety group: BDI score = 6.62 + 5.04; low

anxiety group: BDI score = 5.64 + 3.96; p = 0.607). For the analysis taking trait anxiety sub-

score separately, subjects with trait anxiety scores between 20 and 35 were considered as the

low trait anxiety group (in ACC: males = 7, females = 4; in hippocampus: males = 6, females

= 3), and those with trait anxiety scores above 35 were considered as the high trait anxiety

group (median split on the basis of trait anxiety scores) (in ACC: males = 4, females = 9; in

hippocampus: males = 3, females = 9).

3.2. Between-group differences in neurochemical variables

There was no significant group difference in tCr concentration (institutional units) in any of

the ROIs as obtained using the LCModel (ACC, p = 0.89 (high anxiety group, tCr = 5.36 +

0.38, n = 13; low anxiety group, tCr = 5.33 + 0.44, n = 10); hippocampus, p = 0.36 (high

anxiety group, tCr = 5.82 + 0.50, n = 9; low anxiety group, tCr = 6.02 + 0.44, n = 10)). In the

ACC, the levels of mI/tCr (p = 0.036) and (Glu+Gln)/tCr (p = 0.026) were significantly

higher in high anxiety group than in the low anxiety group (Table 1, Fig. 3a, 3b, 3c, 3d, 3e,

3f, 3g, 3h, 3i, 3j, 3k, 3l). The level of Glu/tCr also showed an increase in the high anxiety

group as compared with the low anxiety group, but with marginal significance (p = 0.051). In

the hippocampus, none of the metabolites/neurochemicals showed significant differences

between the two groups (Table 1).

When the subjects were divided into two groups based on their trait anxiety sub-score, the

levels of Glu/tCr (p = 0.008), mI/tCr (p = 0.046) and (Glu+Gln)/tCr (p = 0.003) were found

to be significantly elevated in the high trait anxiety group as compared with the low trait

anxiety group in the anterior cingulate cortex (Table 2). In the hippocampus, none of the

metabolites/neurochemicals showed a significant difference between the two groups (based

on trait anxiety scores) (Table 2).

3.3. Correlation between neurochemical variables and anxiety

In the anterior cingulate cortex, partial correlation analysis revealed that the total anxiety

scores of the subjects were positively correlated with the Glu/tCr ratio and the Glu+Gln/tCr

ratio (r =0.462, p = 0.023 for Glu and r = 0.573, p = 0.005 for Glu+Gln, Fig. 4). mI/tCr and

tCh/tCr also showed a positive correlation with total anxiety scores but with marginal

significance (r =0.377, p = 0.056 for mI and r = 0.358, p = 0.066 for tCh). In the

hippocampus, a marginally significant positive correlation was obtained between total

anxiety and the Glu/tCr ratio (r = 0.415, P = 0.055).

Similarly, when the trait anxiety sub-score was analysed separately, the Glu/tCr (r =0.402, p

= 0.044) and Glu+Gln/tCr (r =0.461, p = 0.024) ratios showed a predictive value for trait

anxiety in the anterior cingulate cortex (Fig. 5). mI/tCr also showed a positive correlation

with trait anxiety scores but with marginal significance (r =0.380, p = 0.054) On the other

hand, in the hippocampus, no correlation was obtained between trait anxiety scores and any

of the metabolites/neurochemicals analysed.

4. Discussion

Ours is one of the few reports that demonstrate the relationship between regional

neurochemical changes in the brain and anxiety in healthy individuals. Considering the

importance of the anterior cingulate cortex (ACC) and the hippocampus in regulating

emotional processing and their implication in anxiety, they were chosen as ROIs. In the ACC,

we obtained an increase in mI, Glu and the combination Glu+Gln in high anxious subjects as

compared with low anxious ones. Preliminary findings further suggest that total anxiety

predicts the concentration of Glu and the combination Glu+Gln in the ACC. The Glu levels

in the hippocampus were also found to be positively correlated with total anxiety, although at

a marginally significant level. It has been consistently suggested in the literature that changes

in the unresolved glutamate and glutamine (Glu+Gln) resonance are predominantly due to

changes in Glu (Phan et al., 2004; Cortese and Phan, 2005). Glutamate is the major excitatory

neurotransmitter in the mammalian brain. Limbic and associated paralimbic structures (that

include the hippocampus and the ACC, respectively) have been identified as being richly

innervated by glutamatergic pyramidal cells (Cortese and Phan, 2005). In a recent study that

carried out metabolomic analyses on a trait anxiety mouse model, the major excitatory

neurotransmitter glutamate, which binds to the N-methyl-D-aspartate (NMDA) receptor, was

found at higher levels in the plasma of a High Anxiety Behaviour mouse as compared with a

Low Anxiety Behaviour mouse (Zhang et al., 2011). Our findings are also consistent with

previous MRS studies providing evidence that anxiety disorders are associated with

alterations in the glutamate system. A study on generalized SAD (Phan et al., 2005) showed

an increase in Glu/Cr in the ACC of patients as compared with controls along with a

correlation of Glu/Cr with the intensity of social anxiety symptoms. The authors explained

the findings on the basis of proposed models for glutamate’s role in anxiety according to

which an increased glutamatergic transmission and/or excessive glutamate release within the

limbic system is associated with fear-related learning and reactivity (Walker and Davis,

2002). Further, the hyperresponsive limbic system in response to social threat/scrutiny and

anxiety-provoking situations in patients with social anxiety, as well as its attenuation on

successful treatment, suggests the functional significance of glutamate in SAD (Phan et al.,

2005). Similarly, Grachev and Apkarian (2000) reported an increase in Glu in the orbito-

frontal cortex in healthy individuals with high state-trait anxiety. A recent study on pediatric

generalized anxiety disorder (GAD) patients showed that Glu/Cr is related to the severity of

anxiety symptoms and suggested the possibility that dysregulation of Glu within the ACC

may be linked to the pathophysiology of pediatric GAD (Strawn et al., 2013). Dynamic

changes in Glu concentrations have also been observed in the ACC during experimentally

induced panic (Zwanzger et al., 2013), suggesting a disturbance of the inhibitory-excitatory

equilibrium. Human genetic studies and clinical drug trials also present evidence for the

involvement of glutamate in anxiety disorders (Cortese and Phan, 2005). In a study by Arnold

et al. (2004), a significant association between a glutamate system gene, GRIN2B (glutamate

NMDA receptor subtype 2B), and OCD diagnosis and lifetime symptom severity was

obtained, thereby supporting glutamate system changes in the pathophysiology of anxiety.

Clinical drug trials using compounds such as phenytoin and topiramate that have direct

actions on glutamate neurotransmission have been reported for the treatment of anxiety

disorders (Cortese and Phan, 2005). Earlier, neuroimaging-based morphological and functional

studies have also shown that glutamate-rich brain regions such as the hippocampus,

amygdala, and ACC are either structurally altered or functionally hyperactive in patients

diagnosed with anxiety disorders (Cortese and Phan, 2005). Our findings of increased Glu/tCr

in the ACC and hippocampus document a direct relationship between the glutamate system

and the neurobiology of anxiety in healthy individuals, thereby suggesting a role of glutamate

in mediating the physiological and behavioral consequences associated with anxiety even at a

sub-clinical level.

Other metabolites that showed a relation with anxiety were mI and tCh. mI is an intermediate

in the cerebral inositol poly-phosphate (IPP) cascade and plays an important role in

intracellular signaling pathways. It acts as a second messenger that liberates Ca2+ from the

endoplasmic reticulum and is involved in the recognition of chemical signals (Grachev and

Apkarian, 2000). In a recent study on monozygotic twin pairs, it was found that GAD genetic

risk predicted increases in myo-inositol in the amygdala (Hettema et al., 2012). The

proteomic and metabolomic study by Zhang et al. (2011) has also shown the inositol pathway

to be critically involved in the anxiety phenotype with several proteins and metabolites that

are part of phosphatidylinositol signaling having altered expression levels. However,

according to the authors, there is not enough evidence available to know whether these

pathways are causative or the result of distinct anxiety endophenotypes. Grachev and

Apkarian (2000) also reported an increase in mI in the orbitofrontal cortex of high anxiety

individuals. Our findings further suggest that inositol pathways are implicated in the anxiety

phenotype even at a sub-clinical level. An increased tCh in the ACC might suggest an

increase in overall cell density and/or the rate of membrane turnover as a function of anxiety.

Similar findings in Glu, Glu+Gln and mI were obtained when the trait anxiety sub-score of

the STAI was used as the parameter to classify the subjects into two groups – high trait

anxious and low trait anxious. The results further provide neuroimaging-based evidence that

both Glu and mI levels are altered as a function of sub-clinical anxiety.

In contrast to the findings in healthy subjects (Grachev and Apkarian, 2000; Fan et al., 2010)

and subjects with various anxiety disorders like SAD and OCD (Phan et al., 2005; Whiteside

et al., 2006), i.e., that increased NAA concentration was associated with anxiety symptoms,

we did not find any significant difference between NAA/tCr ratios between the two anxiety

groups. NAA is a pre-product of the excitatory neurotransmitter aspartate and is considered to

be a marker of neuronal density and integrity. Therefore, earlier MRS researchers had

hypothesized that a perpetual level of high anxiety leads to an increase in excitatory

neurotransmitter release and subsequent neuronal reorganization. It may further lead to an

increase in the number of axons and synaptic connections, which is suggested by an increase

in the levels of NAA (Grachev and Apkarian, 2000; Phan et al., 2005). One of the reasons for

this discrepancy in findings might be the region of the brain studied. The earlier findings

were mainly obtained in the orbito-frontal cortex or the medial prefrontal cortex, as against

the ACC or hippocampus in our study. Moreover, as suggested by the authors of earlier

studies (Grachev and Apkarian, 2000; Phan et al., 2005), the findings in NAA should be

considered preliminary and should be reexamined using morphometric techniques. Moreover,

a recent morphometric study in major depressive disorder and anxiety (SAD and PD),

showed lower grey matter volumes in the rostral anterior cingulate gyrus extending into the

dorsal anterior cingulate gyrus in both depression and anxiety patients (van Tol et al., 2010).

Therefore, the putative association between NAA and sub-clinical anxiety in healthy

populations needs further investigation.

In summary, this is the first study suggesting alterations in ACC neurochemistry in relation

with anxiety in healthy subjects, with glutamate, combined glutamate and glutamine and

myo-inositol showing the maximum changes. The study thus contributes to the limited

literature available on altered metabolism and neural mechanisms underlying sub-clinical

anxiety. The limitations of our study include a small sample size and the absence of measures

of absolute concentrations of the neurochemicals. Moreover, even though the healthy subjects

were recruited by screening them for current or past medical and psychiatric illness using the

medical history questionnaire of the Hindi version of the Diagnostic Interview for Genetic

Studies (version 2) (Deshpande et al., 1998), the absence of full psychiatric evaluation, for

example, the structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) (First et

al., 2002), is another limitation of the study. Along with earlier etiological models that

suggest high anxiety trait to be a precursor to anxiety disorders and depression, the observed

neurochemical changes in our study, which are similar to those obtained in various anxiety

disorders, further support the need for preventive interventions in high anxiety individuals.

Acknowledgement This work was supported by DRDO R&D Project No. INM 311 (4.1).

References

Arnold, P.D., Rosenberg, D.R., Mundo, E., Tharmalingam, S., Kennedy, J.L., Richter, M.A.,

2004. Association of a glutamate (NMDA) subunit receptor gene (GRIN2B) with obsessive-

compulsive disorder: a preliminary study. Psychopharmacology (Berl) 174, 530-538.

Baur, V., Hänggi, J., Jäncke, L., 2012. Volumetric associations between uncinate fasciculus,

amygdala, and trait anxiety. BMC Neuroscience 13, 4.

Beck, A.T., Steer, R.A., Brown, G.K., 1996. Beck Depression Inventory: Manual, 2nd ed. The

Psychological Corporation, San Antonio, TX.

Bishop, S.J., Duncan, J., Brett, M., Lawlence, A.D., 2004. Prefrontal cortical function and

anxiety: controlling attention to threat related stimuli. Nature Neuroscience 7(2), 184-188.

Bishop, S.J., 2009. Trait anxiety and impoverished prefrontal control of attention. Nature

Neuroscience 12 (1), 92-98.

Cortese, B.M., Phan, K.L., 2005. The role of glutamate in anxiety and related disorders. CNS

Spectrums 10 (10), 820-830.

Dager, S.R., Friedman, S.D., Heid, A., Layton, M.E., Richards, T., Artru, A. Strauss, W.,

Hayes, C., Posse, S., 1999. Two-dimensional proton echo-planar spectroscopic imaging of

brain metabolic changes during lactate-induced panic. Archives of General Psychiatry 56, 70–

77.

Delgado, M.R., Nearing, K.I., Ledoux, J.E., Phelps, E.A., 2008. Neural circuitry underlying the

regulation of conditioned fear and its relation to extinction. Neuron 59, 829–838.

Deshpande, S.N., Mathur, M.N.L., Das, S.K., Bhatia, T., Sharma, S.D., Nimgaonkar, V.L.,

1998. A Hindi version of the diagnostic interview for genetic studies. Schizophrenia Bulletin

24 (3), 489–493.

Egner, T., Etkin, A., Gale, S., Hirsch, J., 2008. Dissociable neural systems resolve conflict

from emotional versus nonemotional distracters. Cerebral Cortex 18, 1475–1484.

Etkin, A., Egner, T., Kalisch, R., 2011. Emotional processing in anterior cingulate and medial

prefrontal cortex. Trends in Cognitive Sciences 15 (2), 85-93.

Eysenck, M.W., Derakshan, N., Santos, R., Calvo, M.G., 2007. Anxiety and cognitive

performance: attentional control theory. Emotion 7 (2), 336–353.

Fan, Q., Tan, L., You, C., Wang, J., Ross, C.A., Wang, X., Zhang, T., Li, J., Chen, K., Xiao,

Z., 2010. Increased N-Acetylaspartate/creatine ratio in the medial prefrontal cortex among

unmedicated obsessive-compulsive disorder patients. Psychiatry and Clinical Neurosciences 64

(5), 483-490.

First, M. B., Spitzer, R. L., Miriam, G., Williams, J. B.W., 2002. Structured Clinical Interview

for DSM-IV-TR Axis I Disorders, Research Version, Patient edition (SCID-I/P). Biometrics

Research, New York State Psychiatric Institute, New York, NY.

Grachev, I.D., Apkarian, A.V., 2000. Anxiety in healthy humans is associated with orbital

frontal chemistry. Molecular Psychiatry 5, 482–488.

Heckers, S., 2001. Neuroimaging studies of the hippocampus in schizophrenia. Hippocampus

11 (5), 520-528.

Hermens, D.F., Lagopoulos, J., Naismith, S.L., Tobias-Webb, J., Hickie, I.B., 2012. Distinct

neurometabolic profiles are evident in the anterior cingulate of young people with major

psychiatric disorders. Translational Psychiatry 2, e110.

Hettema, J.M., Kettenmann, B., Ahluwalia, V., McCarthy, C., Kates, W.R., Schmitt, J.E.,

Silberg, J.L., Neale, M.C., Kendler, K,S., Fatouros, P., 2012. A pilot multimodal twin imaging

study of generalized anxiety disorder. Depression and Anxiety 29 (3), 202–209.

Hong, S.B., Shin, Y.W., Kim, S.H., Yoo, S.Y., Lee, J.M., Kim, I.Y. Kim, S.I., Kwon, J.S.,

2007. Hippocampal shape deformity analysis in obsessive-compulsive disorder. European

Archives of Psychiatry and Clinical Neurosciences 257 (4), 185-190.

Keshavan, M.S., Kapur, S., Pettegrew, J.W., 1991. Magnetic resonance spectroscopy in

psychiatry: potential, pitfalls, and promise. American Journal of Psychiatry 148 (8), 976-85.

Kimbrell, T., Leulf, C., Cardwell, D., Komoroski, R.A., Freeman, T.W., 2005. Relationship of

in vivo medial temporal lobe magnetic resonance spectroscopy to documented combat

exposure in veterans with chronic posttraumatic stress disorder. Psychiatry Research:

Neuroimaging 140 (1), 91-94.

Lieberman, M.D. Eisenberger, N.I., Crockett, M.J., Tom, S.M., Pfeifer, J.H., Way, B.M., 2007.

Putting feelings into words: affect labeling disrupts amygdala activity in response to affective

stimuli. Psychological Science 18, 421–428.

Lochner, C., Fouché, J.-P., Plessis, S., Spottiswoode, B., Seedat, S., Fineberg, N.,

Chamberlain, S.R., Stein, D.J., 2012. Evidence for fractional anisotropy and mean diffusivity

white matter abnormalities in the internal capsule and cingulum in patients with obsessive–

compulsive disorder. Journal of Psychiatry and Neuroscience 37 (3), 193–199.

Mahmutyaz�c�o�lu, K., Konuk, N., Özdemir, H., Atasoy, N., Atik, L., Gündo�du, S., 2005.

Evaluation of the hippocampus and the anterior cingulate gyrus by proton MR spectroscopy in

patients with post-traumatic stress disorder. Diagnostic and Interventional Radiology 11, 125-

129.

Miller, B.L., 1991. A review of chemical issues in 1H NMR spectroscopy: N-acetyl-L-

aspartate, creatine and choline. NMR in Biomedicine 4, 47-52.

Modi, S., Trivedi, R., Singh, K., Kumar, P., Rathore, R.K.S., Tripathi, R.P., Khushu, S., 2013.

Individual differences in trait anxiety are associated with white matter tract integrity in fornix

and uncinate fasciculus: preliminary evidence from a DTI based tractography study.

Behavioural Brain Research 238, 188–192.

Montag, C., Reuter, M., Weber, B., Marketta, S., Schoene-bake, J-C., 2012. Individual

differences in trait anxiety are associated with white matter tract integrity in the left temporal

lobe in healthy males but not females. Neuroscience 217, 77–83.

Phan, K.L., Fitzgerald, D.A., Cortese, B.M., Seraji-Bozorgzad, N., Tancer, M.E., Moore, G.J.,

2004. Anterior cingulate neurochemistry in social anxiety disorder: 1H-MRS at 4Tesla.

Neuroreport 16 (2), 183-186.

Provencher, S.W., 1993. Estimation of metabolite concentrations from localized in vivo NMR

spectra. Magnetic Resonanace in Medicine 30, 672–679.

Pujol, J., Lo´pez, A., Deus, J., Cardoner, N., Vallejo, J., Capdevila, A., Paus, T., 2002.

Anatomical variability of the anterior cingulate gyrus and basic dimensions of human

personality. NeuroImage 15, 847–855.

Sala, M., Perez, J., Soloff, P., Ucelli di Nemi, S., Caverzasi, E., Soares, J.C., Brambilla, P.,

2004. Stress and hippocampal abnormalities in psychiatric disorders. European

Neuropsychopharmacology 14 (5), 393-405.

Sandi, C., Richter-Levin, G., 2009. From high anxiety trait to depression: a neurocognitive

hypothesis. Trends in Neuroscience 32 (6), 312-320.

Spielberger, C.D., 1983. Manual for the State-Trait Anxiety Inventory. Consulting

Psychologists Press, Palo Alto, CA.

Strawn, J.R., Chu, W.J., Whitsel, R.M., Weber, W.A., Norris, M.M., Adler, C.M., Eliassen,

J.C., Phan, K.L., Strakowski, S.M., DelBello, M.P., 2013. A pilot study of anterior cingulate

cortex neurochemistry in adolescents with generalized anxiety disorder. Neuropsychobiology

67 (4), 224-229.

van Tol, M.J., van der Wee, N.J., van den Heuvel, O.A., Nielen, M.M., Demenescu, L.R.,

Aleman, A., Renken, R., van Buchem, M.A., Zitman, F.G., Veltman, D.J., 2010. Regional

brain volume in depression and anxiety disorders. Archives of General Psychiatry 67 (10),

1002-1011.

Walker, D.L., Davis, M., 2002. The role of amygdala glutamate receptors in fear learning, fear-

potentiated startle, and extinction. Pharmacology, Biochemistry and Behavior 71, 379–392.

Westlye, L.T., Bjørnebekk, A., Grydeland, H., Fjell, A.M., Walhovd, K.B., 2011. Linking an

anxiety-related personality trait to brain white matter microstructure diffusion tensor imaging

and harm avoidance. Archives of General Psychiatry 68 (4), 369–377.

Whiteside, S.P., Port, J.D., Deacon, B.J., Abramowitz, J.S., 2006. A magnetic resonance

spectroscopy investigation of obsessive–compulsive disorder and anxiety. Psychiatry

Research: Neuroimaging 146, 137–147.

Zhang, Y., Filiou, M.D., Reckow, S., Gormanns, P., Maccarrone, G., Kessler, M.S., Frank, E.,

Hambsch, B., Holsboer, F., Landgraf, R., Turck, C.W., 2011. Proteomic and metabolomic

profiling of a trait anxiety mouse model implicate affected pathways. Molecular & Cellular

Proteomics 10, 12.

Zwanzger, P., Zavorotnyy, M., Gencheva, E., Diemer, J., Kugel, H., Heindel, W., Ruland, T.,

Ohrmann, P., Arolt, V., Domschke, K., Pfleiderer, B., 2013. Acute shift in glutamate

concentrations following experimentally induced panic with cholecystokinin tetrapeptide--a

3T-MRS study in healthy subjects. Neuropsychopharmacology 38 (9), 1648-1654.

Figure captions:

Fig. 1: Location of (a) 4-cm3 voxel in anterior cingulate cortex and (b) 2.5-cm3 voxel in right

hippocampus of a normal subject.

Fig. 2: Typical spectrum from (a) anterior cingulate cortex and (b) right hippocampus as

analysed by the LCModel. The concentrations of the metabolites and their associated

Cramer-Rao bounds are listed in the box on the right.

Fig. 3: Scatter plots for (a) Glu/tCr (b) Ins/tCr (c) NAA/tCr (d) tCh/tCr (e)

(NAA+NAAG)/tCr (f) (Glu+Gln)/tCr in the anterior cingulate cortex and (g) Glu/tCr (h)

Ins/tCr (i) NAA/tCr (j) tCh/tCr (k) (NAA+NAAG)/tCr (l) (Glu+Gln)/tCr in the hippocampus

in low and high anxiety groups. The horizontal lines in the scatter plot represent the mean

values.

Fig. 4: Partial correlation plots (controlling for age, sex, BDI, FWHM and SNR of the

spectra) between levels of (a) Glu/tCr and total anxiety scores and (b) (Glu+Gln)/tCr and total

anxiety scores in anterior cingulate cortex of the healthy subjects.

Fig. 5: Partial correlation plots (controlling for age, sex, BDI, FWHM and SNR of the

spectra) between levels of (a) Glu/tCr and trait anxiety scores and (b) (Glu+Gln)/tCr and trait

anxiety scores in anterior cingulate cortex of the healthy subjects.

Table 1

Relative values of neurochemicals for the high and low anxiety groups

Experimental group

High anxious Low anxious P valuea Effect Size

Mean (SD) Mean (SD)

Anterior Cingulate

Cortex

(N = 13) (N=11)

Glu/tCr 1.52 (0.166) 1.36 (0.210) 0.051# 0.85

Ins/tCr 0.79 (0.109) 0.71 (0.087) 0.036* 0.82

NAA/tCr 1.22 (0.131) 1.16 (0.127) 0.435 0.46

tCh/tCr 0.25 (0.016) 0.24 (0.025) 0.107 0.49

(NAA+NAAG)/tCr 1.23 (0.123) 1.17 (0.128) 0.418 0.48

(Glu+Gln)/tCr 2.03 (0.227) 1.76 (0.252) 0.026* 1.13

Hippocampus (N = 11) (N=10)

Glu/tCr 1.268 (0.146) 1.274 (0.186) 0.605 -0.04

Ins/tCr 0.820 (0.171) 0.992 (0.271) 0.356 -0.77

NAA/tCr 0.998 (0.104) 0.955 (0.132) 0.561 0.36

tCh/tCr 0.311 (0.032) 0.300 (0.024) 0.626 0.39

(NAA+NAAG/tCr) 1.110 (0.102) 1.085 (0.114) 0.511 0.23

(Glu+Gln)/tCr 2.057 (0.352) 2.101 (0.353) 0.868 -0.12

a With age, sex, current depression, Signal to Noise Ratio and Full Width at Half Maximum

of the spectra as covariates of no interest.

* Boldface indicates p < 0.05 (Bonferroni correction for multiple testing); # marginally

significant differences at p < 0.10 are indicated in italics.

Table 2

Relative values of neurochemicals for the high and low trait anxiety groups

Experimental group

High trait

anxious

Low trait

anxious

P valuea Effect Size

Mean (SD) Mean (SD)

Anterior Cingulate

cortex

(N = 13) (N=11)

Glu/tCr 1.54 + 0.186 1.36 + 0.179 0.008 0.99

Ins/tCr 0.79 + 0.107 0.71 + 0.092 0.046 0.80

NAA/tCr 1.18 + 0.122 1.21+ 0.143 0.349 -0.23

tCh/tCr 0.25 + 0.018 0.25 + 0.023 0.366 0.00

(NAA+NAAG)/tCr 1.20 + 0.119 1.21 + 0.138 0.508 -0.08

(Glu+Gln)/tCr 2.01 + 0.238 1.78 + 0.262 0.003 0.92

Hippocampus (N = 12) (N=9)

Glu/tCr 1.27 + 0.154 1.27 + 0.154 0.569 0.00

Ins/tCr 0.86 + 0.160 0.96 + 0.310 0.825 -0.43

NAA/tCr 0.97 + 0.088 0.96 + 0.154 0.436 0.08

tCh/tCr 0.30 + 0.030 0.31 + 0.027 0.352 -0.35

(NAA+NAAG/tCr) 1.13 + 0.105 1.05 + 0.091 0.232 0.82

(Glu+Gln)/tCr 2.10 + 0.382 2.04 + 0.305 0.337 0.17

a With age, sex, current depression, Signal to Noise Ratio and Full Width at Half Maximum

of the spectra as covariates of no interest.

* Boldface indicates p < 0.05 (Bonferroni correction for multiple testing)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)

Figure(s)