Proton MRS in twin pairs discordant for schizophrenia

Article metrics


Proton magnetic resonance spectroscopy (1H MRS) neurometabolite abnormalities have been detected widely in subjects with and at risk for schizophrenia. We hypothesized that such abnormalities would be present both in patients with schizophrenia and in their unaffected twin siblings. We acquired magnetic resonance spectra (TR/TE=3000/30 ms) at voxels in the mesial prefrontal gray matter, left prefrontal white matter and left hippocampus in 14 twin pairs discordant for schizophrenia (2 monozygotic, 12 dizygotic), 13 healthy twin pairs (4 monozygotic, 9 dizygotic) and 1 additional unaffected co-twin of a schizophrenia proband. In the mesial prefrontal gray matter voxel, N-acetylaspartate (NAA), creatine+phosphocreatine (Cr), glycerophosphocholine+phosphocholine (Cho) and myo-inositol (mI) did not differ significantly between patients with schizophrenia, their unaffected co-twins or healthy controls. However, glutamate (Glu) was significantly lower in patients with schizophrenia (31%, percent difference) and unaffected co-twins (21%) than in healthy controls (collapsed across twin pairs). In the left hippocampus voxel, levels of NAA (23%), Cr (22%) and Cho (36%) were higher in schizophrenia patients compared with controls. Hippocampal NAA (25%), Cr (22%) and Cho (37%) were also significantly higher in patients than in their unaffected co-twins. Region-to-region differences in metabolite levels were also notable within all three diagnosis groups. These findings suggest that 1H MRS neurometabolite abnormalities are present not only in patients with schizophrenia, but also in their unaffected co-twins. Thus, reduced mesial prefrontal cortical Glu and elevated hippocampal NAA, Cr and Cho may represent trait markers of schizophrenia risk and, when exacerbated, state markers of schizophrenia itself.


Twin studies have established genetic contributions to anatomical, neuropsychological and physiological abnormalities associated with schizophrenia.1, 2, 3 Twin cohort designs involving monozygotic (MZ) and dizygotic (DZ) twin pairs discordant for schizophrenia and matched control twin pairs allow for the dissociation of genetic from shared and unique environmental contributions. As such, the design also allows for the dissociation of state from trait effects by examining the unaffected relatives of patients in relation to patients and unrelated controls.

Magnetic resonance spectroscopy (MRS) is an imaging method that allows for noninvasive in vivo quantification of metabolite concentrations in the brain. Although MRS can be performed using a variety of nuclei, only hydrogen (1H) and phosphorus (31P) exist in vivo in concentrations high enough for routine clinical evaluation. In this study, proton magnetic resonance spectroscopy (1H MRS) was used to examine concentrations of N-acetylaspartate (NAA), glutamate (Glu), creatine+phosphocreatine (Cr), glycerophosphocholine+phosphocholine (Cho) and myo-inositol (mI).

Each of the 1H MRS-detectable metabolites is thought to reflect different, though related, aspects of cell energetic or membrane physiology. In the case of NAA, although controversy remains,4 Baslow et al.5 has recently argued that its principal function is as an osmolyte. NAA transports the large quantities of water generated by glucose (Glc) catabolism particularly during oxidative phosphorylation in the mitochondrion out of the neuron. This maintains proper osmotic pressure across the phospholipid membrane. Consequently, NAA is indirectly linked to energy metabolism. Rather than permanent losses in neuron numbers, NAA deficits observed in many diseases may therefore represent reversible mitochondrial dysfunction.4, 6, 7, 8 N-acetylaspartylglutamate (NAAG), closely related to NAA, contributes a small percentage to the measured NAA and Glu signals, so the presumed role of NAAG cannot be evaluated with this technique. In the visual cortex of rats, NAAG is maintained at 2% of the level of NAA,9 and produces a much weaker singlet resonance signal10 that is difficult to separate from the NAA signal at clinical nuclear magnetic resonance (NMR) magnetic strengths. Low extracellular Glu levels are maintained by astroglial uptake. In turn, the astrocytes convert Glu to glutamine (Gln) which is then transported back to the presynaptic neuron and reconverted to Glu.11, 12 As many amino acids, including Glu, are also involved in intermediary metabolism and protein synthesis, it is difficult to separate their biochemical role from their transmitter role using 1H MRS. The present results cannot separate the metabolic versus transmitter roles of Glu and NAAG, so further work is necessary to clarify.

Creatine (Cr) is also related to the energetic state of brain tissue, because the interconversion of creatine and phosphocreatine maintains an ATP ‘buffer’ to help meet short-term cellular energy demands.13, 14 Finally, the cyclic sugar alcohol mI is a precursor to the phospholipid membrane component phosphatidylinositol and a substrate for the phosphoinositide second-messenger system; changes in mI levels may reflect abnormalities in membrane metabolism or in intracellular signaling mechanisms (reviewed in Irvine and Schell15). Thus, levels of 1H MRS metabolites reflect the status of multiple important functions of neurons and glial cells that may be disturbed in subjects with and at risk for schizophrenia.

Reviews of the MRS schizophrenia literature to date16, 17, 18, 19, 20 show ample evidence of abnormalities in NAA; suggestive evidence for abnormalities in glutamate+glutamine (Glx), Cr and choline (Cho); and scant evidence for abnormalities in mI. In particular, bilateral deficits in NAA, NAA/Cr and/or NAA/Cho in the ‘dorsolateral prefrontal’ region (usually middle frontal gyrus) and mesial temporal lobes (predominantly hippocampus) were detected in early work in adult schizophrenia21 and have since become fairly accepted. Wobrock et al.17 count a majority of studies, including the more rigorous ones, supporting these NAA deficits. Low levels of NAA in schizophrenia are thought to reflect low density or low metabolic activity of neurons, perhaps due to abnormal development or to excitotoxic damage. However, elevated neuronal density in superior frontal cortex (Brodmann area 9) and prefrontal cortex (Brodmann area 46) has been shown in post-mortem schizophrenia brain specimens.22, 23 In addition to chronic (usually medicated) schizophrenia patients, NAA deficits have been observed in first episode, treatment-naive patients,24, 25, 26 in schizophrenia spectrum disorder patients,27 in schizophrenia patients’ psychotic relatives,28 in unaffected siblings of schizophrenia patients,29 in children and adolescents diagnosed with schizophrenia30 and in subjects at risk for schizophrenia,31 indicating that these deficits may occur in the absence of overt symptoms. Glutamatergic models of schizophrenia (Farber et al.,32 Olney et al.,33 Jentsch et al.,51 Laruelle et al.,52 Olney et al.53 and Goff et al.78) have recently sharpened interest in in vivo MRS assays of Glu, Gln and their sum Glx. Various abnormalities have been reported involving these compounds in schizophrenia (reviewed in Abbott and Bustillo,16 Wobrock et al.17 and Stanley34). Such abnormalities may signal disturbances in neurotransmission (including excitotoxicity), cell energetics or the balance between neurons and glial cells that maintains the Glu–Gln cycle. Cho findings in schizophrenia (reviewed in Wobrock et al.,17) are less consistent than those for NAA. Elevated Cho levels have been interpreted as supportive of the ‘membrane hypothesis’ of schizophrenia35, 36 with more severe phospholipid disturbances resulting in an earlier disease onset.37 Theberge et al.38 found that Cho correlated positively with duration of untreated psychosis and suggested that untreated psychosis may increase membrane turnover, perhaps on account of excitotoxicity. However, Cho results may be confounded because the Cho resonance includes Cho, phosphocholine (PCh) and other trimethylamines. Published MRS findings have failed to show significant mI abnormalities in schizophrenia.39, 40

The use of twin pairs discordant for schizophrenia permits examination of 1H MRS metabolites in subjects identical in age and who often have undergone much of the same experiences in life. These subjects either share 100% (MZ) or on average 50% (DZ) of their genomic sequence. When one twin has developed schizophrenia and the other has not, this represents an opportunity to identify MRS metabolite measures that may signal the presence of (disease state markers) or liability for (trait markers) schizophrenia. To our knowledge, no previous study has acquired 1H MRS in twin pairs discordant for schizophrenia. We sampled mesial prefrontal cortex (mesPFC), left prefrontal white matter (L-PFWM) and left hippocampus (L-hip), three regions readily accessible to localized MRS. As discussed above, NAA deficits have been found repeatedly in hippocampus in schizophrenia. There are scant positive reports of metabolite abnormalities in white matter41, 42, 43 and there are multiple reports of metabolite abnormalities in mesPFC, in particular the anterior cingulate cortex.41, 44, 45, 46, 47, 48

On the basis of the meta-analysis of MRS studies in schizophrenia by Steen et al.,20 we anticipated reduced NAA levels in mesPFC, L-PFWM and L-hip in schizophrenic subjects, and, to a lesser degree, in their unaffected co-twins. Also based on Steen et al.,20 we hypothesized increased Cho and Cr levels in the frontal lobes of schizophrenic patients compared with controls. Although not yet consistently supported by MRS studies, which have shown increased levels of Glu in the mesPFC,49 and reduced levels in anterior cingulate gyrus46 and thalamus50 of schizophrenia subjects compared with controls, based on the hypoglutamatergic hypothesis of schizophrenia,51, 52, 53 we predicted a reduction in Glu in patients compared with controls.

Materials and methods


Participants were drawn from a twin cohort consisting of all same-sex twins born in Finland between 1940 and 1957 (N=9562 pairs) identified through the Finnish national population registry. Questionnaire-based classification identified 2495 MZ twins, 5378 DZ twins and 1689 twins of unknown zygosity.54 This cohort was screened for history of hospitalization, prescription of psychotropic medicines and/or work disability due to psychiatric indication occurring in the years 1969–1991 in three national computerized databases: the Hospital Discharge Register, the Free Medicine Register and the Pension Register.55 These searches identified 348 index twin pairs with at least one co-twin with a diagnosis of schizophrenia or schizoaffective disorder and 9214 healthy pairs with no schizophrenia diagnosis in either co-twin according to any of the three databases. After exclusions due to death or emigration, a total of 260 twins consisting of 60 (27 MZ and 33 DZ) index pairs were chosen randomly from the available index pairs (n=229: 50 MZ, 121 DZ and 58 unknown zygosity), along with 70 (34 MZ and 36 DZ) demographically balanced healthy pairs. Index pairs in which, on direct interview, either the proband had a diagnosis of schizoaffective disorder-bipolar type, or the co-twin had a psychotic disorder diagnosis were excluded (one concordant MZ pair). Healthy pairs were excluded if there was a history of psychosis-related treatment or work disability in any of their first-degree relatives or if either co-twin was found, on direct interview, to meet diagnostic criteria for a psychotic disorder or schizotypal, paranoid or schizoid personality disorder (15 pairs: 6 MZ and 9 DZ).

Each co-twin was interviewed using the Structured Clinical Interview for DSM-III-R disorders, patient or nonpatient edition,56 by an examiner who was blind to the zygosity and diagnostic status of the co-twin, and the twins were assigned diagnoses according to DSM-IV.57 Co-twins and healthy individuals were also interviewed and rated on the cluster A items from the Personality Disorder Examination.58

MRS scans were acquired on a total of 55 subjects, including 14 pairs discordant for schizophrenia (2 MZ and 12 DZ) and 13 healthy pairs (4 MZ and 9 DZ) and 1 additional unaffected co-twin of a schizophrenia proband. Data were excluded from further analysis when the voxel coordinates were not or incorrectly recorded, when a high-resolution T2-weighted magnetic resonance imaging (MRI) scan was unavailable to determine voxel tissue composition, or when a quality control exclusion criterion (spectral width (full width at half maximum, FWHM) larger than 0.1 p.p.m., signal-to-noise ratio (SNR) of less than 3 or a Cramer–Rao lower bound (CRLB) larger than 25% for Glu and larger than 20% for all other metabolites) was exceeded.

The estimated singlet line width of the LCModel fit (FWHM) was measured for each spectrum in all regions. The mean±s.d. FWHM was 0.060±0.021 p.p.m. for mesPFC, 0.063±0.032 p.p.m. for L-PFWM and 0.080±0.037 p.p.m. for L-hip. No subjects were excluded based on estimated FWHM because all were less than 0.1 p.p.m.

Estimated SNR of the LCModel fit were measured for each spectrum in all regions. Subjects were excluded from analysis if SNR were less than 3. With this cutoff, one subject's mesPFC voxel and another's L-hip voxel were excluded from analysis. The mean±s.d. SNR was 8.4±2.5 for mesPFC, 8.4±2.9 for L-PFWM and 5.2±2.1 for L-hip.

We used the recommended 20% CRLB cutoff59 for all of the metabolites, except for Glu, for which we used 25%. This slightly more liberal cutoff for Glu was used because this allowed four additional subjects whose CRLB values fell above 20% and below 25% to be used in the analysis and group differences with the liberal and conservative cutoff remained largely the same. Further, although a CRLB cutoff of 25% is greater than LCModel's recommended 20%, several studies have used 25,60, 61 30,62, 63 4064 and 50%.65, 66 The mean±s.d. CRLBs in the mesPFC were 6.6±1.7 for Cr, 10.7±2.6 for mI, 5.7±1.4 for NAA, 14.9±4.0 for Glu and 7.7±1.9 for Cho. The CRLBs (mean±s.d.) in the L-PFWM were 7.0±1.4 for Cr, 10.6±2.7 for mI, 6.4±2.8 for NAA, 18.3±3.5 for Glu and 6.8±1.8 for Cho. The CRLBs (mean±s.d.) in the L-hip were 8.4±2.6 for Cr, 11.0±2.9 for mI, 7.2±2.1 for NAA, 16.8±4.8 for Glu and 9.3±3.5 for Cho. On the basis of the CRLBs, three subjects were excluded from mesPFC mI, two subjects from L-PFWM mI and six subjects from L-hip NAA analyses because a metabolite-specific CRLB exceeded 20%.

On the basis of the combined exclusion criteria, for the L-PFWM voxel, a total of 13 subjects (5 patients, 3 co-twins, 5 controls) were excluded, either because the voxel coordinates were not recorded correctly (N=4) or because a high-resolution T2-weighted MRI scan was unavailable to determine voxel tissue composition (N=9). For the mesPFC voxel, a total of 14 subjects were excluded (the same as for the L-PFWM but also including 1 subject with SNR less than three). For the L-hip voxel, a total of 17 subjects were excluded; the same as for the mesPFC plus 3 additional subjects who were missing voxel coordinates.

In sum, after exclusions, data from 21 controls, 12 co-twins and 9 probands were analyzed (Table 1). Statistical comparison of demographic data from included and excluded subjects showed differences in gender distribution (χ2 (1, N=55)=4.61; P=0.0318). The groups did not differ in zygosity (χ2 (1, N=55)=2.76, P=0.0964); age (t(53)=1.02, P=0.3112); duration of illness (t(12)=0.54, P=0.600) and illness onset age (t(12)=1.13, P=0.2809).

Table 1 Subject demographics

1H MR spectroscopy

Proton spectra were acquired using a 3.0 Tesla scanner (GE, Milwaukee, WI, USA) at the Advanced Magnetic Imaging Centre (AMI Centre) of the Helsinki University of Technology. Three volumes of interest: mesPFC (2 × 2 × 2 cm, 8 cc); L-PFWM (2 × 2 × 2 cm) and L-hip (1.5 × 1.5 × 1.5 cm) were obtained for each subject. These were localized on axial T1-weighted MRI in a way that maximized gray matter (mesPFC and L-hip voxels) or white matter (L-PFWM voxel) content (Figure 1). More specifically, the mesPFC voxel included parts of the cingulate sulcus, cingulate gyrus, frontal pole and superior frontal gyrus; the L-PFWM voxel contained the middle frontal gyrus and superior frontal gyrus and the L-hip voxel included the hippocampus, parahippocampal gyrus, fusiform gyrus and collateral sulcus.

Figure 1

Axial (left, center) and coronal (right) T1-weighted magnetic resonance imaging (MRI) of the brain showing placements (white boxes) of mesial prefrontal gray matter (mesPFC) (left; 2 × 2 × 2 cm), left prefrontal white matter (L-PFWM) (center; 2 × 2 × 2 cm) and left hippocampal (L-hip) (right; 1.5 × 1.5 × 1.5 cm) proton magnetic resonance spectroscopy (1H MRS) single voxels.

Automated global shimming was performed and water-suppressed spectra were acquired using a point-resolved spatially localized spectroscopy sequence (PRESS, TR=3000 ms, TE=30 ms, NEX=64 for mesPFC and L-PFWM, NEX=128 for L-hip). There were 2048 complex points and the spectral bandwidth was 2500 Hz. The GE PROBE-P sequence, which incorporates eight nonwater-suppressed excitations before the 64 water-suppressed excitations, was used so the absolute metabolite levels are normed to unsuppressed water. Metabolite concentrations were reported in arbitrary units due to estimation uncertainty of the NMR-visible water signal.

Spectra (Figure 2) were analyzed offline in the frequency domain using the LCModel commercial spectral-fitting package and a basis set provided by Dr Provencher.67 The metabolites included in the LCModel basis set are: L-alanine (Ala), aspartate (Asp), creatine (Cr), γ-aminobutyric acid (GABA), Glc, Gln, Glu, glycerophosphocholine (GPC), PCh, L-lactate, mI, NAA, NAAG, scyllo-inositol, taurine, creatine methylene group (–CrCH2), guanidoacetate, GPC+PCh, NAA+NAAG, Glu+Gln, lipids and macromolecules: Lip13a, Lip13b, Lip09, MM09, Lip20, MM20, MM12, MM14, MM17, Lip13a+Lip13b, MM14+Lip13a+Lip13b+MM12, MM09+Lip09 and MM20+Lip20.

Figure 2

Sample proton magnetic resonance (1H MR) spectrum acquired from left prefrontal white matter (L-PFWM) of a 56-year-old female twin with schizophrenia as fit by LCModel (Provencher, 199367).

Modeling Glu separately or in combination with Gln produced similar findings across regions and diagnoses. The same was true for modeling NAA separately and in combination with NAAG. Data shown reflect quantification of Glu separately and NAA separately. Choline was quantified by LCModel as GPC, PCh, and the combination of GPC and PCh (GPCPCh). On the basis of CRLBs, modeling the combination, GPCPCh, produced the most reliable data; thus, Cho quantification reflects the combination of these peaks.

Tissue segmentation was performed using FAST (FMRIB's Automated Segmentation Tool 3.51) part of FSL (FMRIB's Software Library 3.2,, 69 The gray matter (GM), white matter (WM) and cerebrospinal fluid (CSF) content of each MRS voxel was calculated. To correct for the tissue composition in each voxel, the percent GM in each voxel was included as a co-variate in the statistical analyses. In addition, to obtain atrophy-corrected metabolite intensities and to verify that metabolic changes were not due to partial volume effects, the following CSF correction70 was applied:

where met(GM+WM) represents the amount of metabolite in the GM and WM of the voxel, respectively, and Vol( ) represents the volume of each particular tissue type in the voxel. According to McLean et al.,72 and Lynch et al.,71 it is safe to assume the metabolite concentrations in CSF are close to 0 to correct for partial volume effects due to CSF and provide metabolite levels per volume of brain tissue. The metabolite levels generated by LCModel represent met(GM+WM), whereas the denominator of the above equation provides the volume fraction CSF correction. Statistical analyses were subsequently performed on the corrected metabolite levels.

Statistical methods

The data were analyzed using the general linear mixed model with repeated measures (Proc Mixed, SAS 8.2; SAS Institute, Cary, NC, USA) treating twin pair as a random variable and the model error terms were adjusted accordingly to correct for dependency between co-twins. Degrees of freedom were estimated from the data using the Satterthwaite option. The hypothesis that diagnosis would be associated with differences in NAA, Glu, Cr, Cho and mI was tested by modeling diagnosis (schizophrenia, unaffected co-twin, healthy control) as a fixed-effect predictor, whereas co-varying for age, sex, region of interest, percent GM and the interaction of diagnosis with region. Region of interest entered the model as a within-subject repeated-measure factor. Significant main effects were followed up with post hoc two-tailed t-tests.



The group-wise analysis of the Glu peaks revealed significant effects for diagnosis (F(2,89)=4.09, P=0.0201); region (F(2,89)=16.34, P<0.0001) and age (F(1,89)=7.49, P=0.0075) (Figure 3; Table 2). The effect of age reflects greater Glu levels in older subjects. The effect of diagnosis revealed significantly greater Glu in controls than probands (t(89)=2.55, P=0.0328) and controls compared with co-twins (t(89)=2.02, P=0.0468), whereas the effect of region showed significantly greater Glu in mesPFC (t(89)=5.47, P<0.0001) and L-hip (t(89)=3.98, P=0.0004) compared with L-PFWM. Although the group by region interaction was not significant, further specification of the group differences within specific regions was undertaken for exploratory analysis, with the most interesting Glu differences present in the mesPFC, in which controls had significantly greater Glu than probands (t(89)=2.75, P=0.0073) and their unaffected co-twins (t(89)=2.08, P=0.0401). Moreover, when the analysis was repeated using the more conservative 20% CRLB cutoff, the group differences in Glu remained substantively the same. The only difference between the two sets of results was that in the analysis using the 20% cutoff, the group by region interaction term was not significant (but the overall group effect remained significant), indicating that patients (and to a lesser extent, co-twins) showed lower Glu levels throughout the ROIs.

Figure 3

Least square mean metabolite quantity (arbitrary units)±standard error of creatine+phosphocreatine (Cr, upper left); glutamate (Glu, upper right); glycerophosphocholine+phosphocholine (Cho, middle left); myo-inositol (mI, middle right) and N-acetylaspartate (NAA, lower left), acquired with proton magnetic resonance spectroscopy (1H MRS) from dorsal mesial prefrontal cortex (mesPFC), left prefrontal white matter (L-PFWM) and left hippocampus (L-hip) in twins with schizophrenia (proband), their unaffected co-twins (co-twin), and healthy control twin pairs (control). Overall (collapsing across regions) metabolite levels are also presented. Note significantly lower Glu in mesPFC of schizophrenia patients and their unaffected co-twins compared with controls. Also, note above normal hippocampal levels of NAA, Cr and Cho in twins with schizophrenia compared with both unaffected co-twins and controls. Significant between-diagnosis metabolite differences in post hoc head-to-head comparisons using a two-way protected t-test are denoted by an * for P<0.05. Where the group by region interaction was not significant, further specification of the group differences within specific regions was undertaken for exploratory analysis. These metabolite differences were compared using a post hoc head-to-head two-way protected t-test and are denoted by a # for P<0.05.

Table 2 Raw group mean metabolite levels by group in arbitrary units (mean±s.d.)


The group-wise analysis showed significant effects for region (F(2, 83)=20.94, P<0.0001) and the diagnosis-by-region interaction (F(4, 84)=3.32, P=0.0142) in predicting NAA levels. The effect of region was driven by higher NAA levels in mesPFC compared with L-PFWM (t(83)=5.18, P<0.0001) and by higher levels in L-hip compared with L-PFWM (t(83)=5.68, P<0.0001). Decomposition of the diagnosis-by-region interaction showed that probands had higher L-hip NAA levels than controls (t(90)=2.84, P=0.0056) and their unaffected co-twins (t(86)=3.88, P=0.0061). With mesPFC and L-PFWM, there were no significant between-group differences in NAA. Furthermore, the control subjects had higher NAA levels in mesPFC compared with L-PFWM (t(82)=4.49, P=0.0007) and L-hip compared with L-PFWM (t(89)=2.99, P=0.0036). Probands had significantly greater NAA in L-hip compared with both mesPFC (t(83)=2.57, P=0.0120) and L-PFWM voxels, (t(83)=4.79, P=0.0002).


Group-wise analysis showed significant effects for region (F(2, 90)=24.81, P<0.0001) in predicting Cr levels. The region effect reflects greater Cr in L-hip compared with L-PFWM (t(90)=6.18, P<0.0001). Although neither the main effect of group nor the group by region interaction were significant, exploratory analysis of group differences in specific regions showed that within L-hip, Cr was significantly higher in probands compared with their unaffected co-twins (t(93)=2.67, P=0.0088) and controls (t(100)=2.45, P=0.0159), but unaffected co-twins did not differ significantly from controls.


Group-wise analysis showed significant effects for region (F(2, 88)=19.49, P<0.0001) in predicting mI levels. Region effects reflect greater mI in the L-hip compared with both mesPFC (t(91)=3.14, P=0.0065) and L-PFWM (t(86)=6.23, P<0.0001), and more mI in mesPFC compared with L-PFWM (t(89)=2.88, P=0.0135).


Group-wise analysis showed significant effects for region (F(2, 106)=7.43, P=0.0010) and age (F(1, 106)=4.54, P=0.0354) in predicting Cho levels. L-hip Cho levels were larger than those in the mesPFC (t(106)=2.54, P=0.0332) and L-PFWM (t(106)=3.80, P=0.0007). The effect of age reflects greater Cho levels in older subjects. Although neither the main effect of group nor the group by region interaction were significant, exploratory analysis of group differences in specific regions showed that in the L-hip, probands had significantly greater Cho compared with both their unaffected co-twins (t(106)=3.37, P=0.0281) and controls (t(106)=2.63, P=0.0098).


The principal findings of this study are (1) that mesPFC Glu levels are lower in both twins with schizophrenia and in their unaffected co-twins compared with in healthy controls and (2) that levels of L-hip NAA, Cr and Cho were higher in twins with schizophrenia compared with in their unaffected co-twins and controls. This pattern of findings suggests that some 1H MRS neurometabolite abnormalities are present both in patients with schizophrenia and in their unaffected co-twins and may represent markers of genetic or shared environmental liability to the disorder, whereas others are present only in patients and are therefore likely disease related. As such, mesial prefrontal cortical Glu reductions may represent markers of schizophrenia risk, whereas elevated hippocampal NAA, Cr and Cho may represent markers of schizophrenia itself.

Glutamate was reduced in twins with schizophrenia and their unaffected co-twins in mesPFC compared with controls. Our findings fortify the notion that Glu abnormalities may occur in the absence of overt schizophrenia symptoms. We can conclude that the decreased Glu levels are not due to drug effects because decreased Glu is found both in twins with schizophrenia and their nonmedicated co-twins. As many amino acids, including Glu, are also involved in intermediary metabolism and protein synthesis, it is difficult to separate their biochemical role from their transmitter role. Even though we see changes in Glu levels, one does not know how these changes relate to vesicular Glu and neurotransmission,73 so further work is necessary to clarify. Previous studies have reported conflicting data on Glu levels in schizophrenia patients, including: increased Glu levels in left dorsolateral prefrontal cortex in first episode schizophrenia patients,74 decreased levels of Glu and Gln in left anterior cingulate cortex of medicated patients with schizophrenia,46 increased levels of Glu in prefrontal cortex and hippocampus of schizophrenia patients,75 and increased Glx levels in multiple WM regions in elderly schizophrenia patients.61

The observed reductions in Glu levels among probands are consistent with the N-methyl D-aspartate (NMDA) receptor hypofunction model of schizophrenia, which assumes decreased glutamatergic neurotransmission. However, direct measurement of Glu neurotransmission is not possible because 1H MRS measures both metabolic and vesicular Glu. The one study that induced NMDA blockade in humans found increased Gln in the anterior cingulate.76 Rowland et al. used 1H MRS to measure Gln levels, thus, indirectly assessing Glu neurotransmitter release. In contrast, our results indicated less Glu across the regions of interest in both twins with schizophrenia and their unaffected co-twins compared with controls. Although it is not known if this Glu is vesicular, the net Glu levels are reduced in schizophrenia patients and their unaffected co-twins. Although we cannot measure glutamatergic neurotransmission directly, the reduction in Glu is consistent with NMDA hypofunction model.

The activity of dopamine neurons is modulated by projections involving Glu transmission from the prefrontal cortex and other areas, such as the amygdala.52 Carlsson et al.77 proposed a model depicting bimodal modulation of dopamine activity in the ventral tegmentum area by glutamatergic projections originating in the frontal cortex. This model provides an anatomical framework relating putative neurochemical dysregulation involved in the pathophysiology of schizophrenia, namely a deficit in Glu transmission.52 Both probands and their co-twins show reduced Glu levels in the mesPFC, consistent with the aforementioned model. Glu functions not only as an excitatory agent in the brain, but also as a major regulator of inhibitory tone. This is achieved by Glu tonically activating NMDA receptors on GABAergic, serotonergic and noradrenergic neurons, thereby driving these neurons to inhibit the activity of major excitatory pathways (both glutamatergic and cholinergic) that convergently innervate primary neurons in neocortical and limbic brain regions. These primary neurons use Glu as a transmitter and regulate their own firing by sending a recurrent inhibitory collateral to an NMDA receptor on a GABAergic neuron that feeds back onto the primary neuron.33 Several lines of evidence support the hypothesis that schizophrenia might be associated with a persistent dysfunction of Glu transmission involving NMDA receptors.51, 53, 78, 79, 80

A meta-analysis of 64 studies (88% of them performed at 1.5 T) showed consistent evidence that NAA is reduced in a broad range of tissues in the schizophrenic brain, though some studies failed to show a reduction.20 In particular, most (8 of 15) published studies of the hippocampus report a significant reduction of NAA in patients. The present findings therefore add to evidence of NAA abnormalities in schizophrenia,16, 17, 18, 19, 20 in subjects at risk for schizophrenia,27, 28, 29, 31 and extend them to the particular case of twin siblings. But the direction of our NAA findings (excess vs deficit) is at odds with the above-cited literature. On the basis of post hoc power calculations, Steen et al.20 concluded that few of these prior studies were sufficiently powered to detect deficits in NAA reliably;20 further they suggest publication bias in favor of studies that show NAA deficits as opposed to no difference with controls or an excess. If present, a local excess of NAA could reflect local energetic hypermetabolism or a local deficit in glial cell populations, leading to densification of remaining neurons. Although elevated neuronal density has not been shown in the hippocampus, it has been observed in the superior frontal cortex (Brodmann area 9) and the prefrontal cortex (Brodmann area 46) of post-mortem schizophrenia brain specimens.22, 23 Accordingly, our excess NAA findings may reflect elevated neuronal density in the hippocampus of schizophrenia patients.

Reports of effects of schizophrenia on absolute Cr are sparse, perhaps due in part to the long-standing, tenuous assumption in spectroscopy that brain Cr remains constant across a wide range of subjects, regions and conditions. This assumption, however, has recently been challenged81 as Cr levels, as reported in the current study, have been shown to change substantially across regions82, 83 and pathologies.84 For example, Buckley et al.85 found that temporal-lobe Cr levels correlated positively with multiple measures of cognitive function, Bustillo et al.86 found that prefrontal cortex Cr levels correlated with a measure of abnormal movement, and in a cross-sectional study Ohrmann et al.87 observed that dorsolateral prefrontal region Cr levels decline with illness phase (chronic schizophrenia<first episode schizophrenia<control). In addition, Deicken et al.88 found bilateral elevation of Cr in schizophrenia in the prefrontal lobes and Auer et al.37 found elevated Cr in WM in chronic medicated schizophrenics that correlated positively with score on the Brief Psychiatric Rating Scale. Wood et al.89 showed a significant elevation of NAA/Cr and Cho/Cr in the dorsolateral prefrontal regions of a group at ultra-high risk of developing psychosis that was interpreted as a decline in Cr indicative of hypometabolism. However, in this study the ultra-high risk subjects were not age matched to their controls and it remains possible that the reported group differences reflect increased NAA rather than decreased Cr.

Myo-inositol levels differed significantly across regions, but not across diagnoses. Sharma et al.39 observed higher levels of mI/Cr in adults with schizophrenia than in controls. A report of above-normal mI in left parietal WM was contradicted by another report of normal values in parietal lobes.40 There is evidence suggesting that alterations in brain mI may be associated with psychiatric disorders such as bipolar disorder and schizophrenia,90 but most spectroscopy studies have reported normal mI concentrations in several brain regions in schizophrenic patients.18, 28, 91, 92 Hence, the verdict on mI abnormalities in schizophrenia awaits further study.

Common issues for all 1H MRS studies include that most metabolite peaks reflect multiple metabolites (for example, Cho is quantified as the combination of the GPC and PCh peaks). Similarly, creatine levels reflect both phosphocreatine and unphosphorylated creatine, and some metabolites are harder to measure separately than others (for example, Glu and Gln). At 3-T field strength, however, Glu and Gln, have been reliably separated by numerous investigators, including Purdon et al.,60 Harris et al.,93 Shibuya-Tayoshi et al.,94 Shulman et al.,95 Gallinat et al.,96 and Schubert et al.97 Given the high quality of our mesial prefrontal cortical Glu spectra (S/N=7–9, FHWM=0.05–0.068 p.p.m., CRLB=14–16.4%), we believe the results accurately reflect Glu levels. As many amino acids, including Glu, are also involved in intermediary metabolism and protein synthesis, it is difficult to separate their biochemical role from their transmitter role. Even though we see changes in Glu levels, it is currently unknown whether such changes relate to vesicular or free (neurotransmission) Glu.

This study corrects metabolite levels for voxel tissue composition to address partial-volume effects that could reflect atrophy or hypertrophy of gray or white matter. In addition, we correct for CSF content to correct for partial volume effects due to CSF. Although subject age was included in the statistical analysis, the control group was not matched with the other two groups in age. Given that only two MZ discordant twin pairs were analyzed, environmental versus genetic effects on metabolite levels cannot be teased apart. Finally, a higher number of MZ twins would have greatly strengthened our ability to examine metabolite markers of schizophrenia against a constant genetic background. Bearing these limitations in mind, this study suggests that certain MRS neurometabolite abnormalities are present in the unaffected co-twins of patients with schizophrenia. In addition, the neurometabolite pattern suggests that MRS is sensitive to inherited differences in neurochemical metabolites among schizophrenia patients and their co-twins.


  1. 1

    Karlsgodt KH, Glahn DC, van Erp TG, Therman S, Huttunen M, Manninen M et al. The relationship between performance and fMRI signal during working memory in patients with schizophrenia, unaffected co-twins, and control subjects. Schizophr Res 2007; 89: 191–197.

  2. 2

    van Erp TG, Saleh PA, Huttunen M, Lonnqvist J, Kaprio J, Salonen O et al. Hippocampal volumes in schizophrenic twins. Arch Gen Psychiatry 2004; 61: 346–353.

  3. 3

    Cannon TD, Huttunen MO, Lonnqvist J, Tuulio-Henriksson A, Pirkola T, Glahn D et al. The inheritance of neuropsychological dysfunction in twins discordant for schizophrenia. Am J Hum Genet 2000; 67: 369–382.

  4. 4

    Moffett JR, Namboodiri AM . Preface: a brief review of N-acetylaspartate. Adv Exp Med Biol 2006; 576: vii–xiii.

  5. 5

    Baslow MH, Suckow RF, Gaynor K, Bhakoo KK, Marks N, Saito M et al. Brain damage results in down-regulation of N-acetylaspartate as a neuronal osmolyte. Neuromolecular Med 2003; 3: 95–104.

  6. 6

    Bates TE, Strangward M, Keelan J, Davey GP, Munro PM, Clark JB . Inhibition of N-acetylaspartate production: implications for 1H MRS studies in vivo. Neuroreport 1996; 7: 1397–1400.

  7. 7

    Dautry C, Vaufrey F, Brouillet E, Bizat N, Henry PG, Conde F et al. Early N-acetylaspartate depletion is a marker of neuronal dysfunction in rats and primates chronically treated with the mitochondrial toxin 3-nitropropionic acid. J Cereb Blood Flow Metab 2000; 20: 789–799.

  8. 8

    Gasparovic C, Arfai N, Smid N, Feeney DM . Decrease and recovery of N-acetylaspartate/creatine in rat brain remote from focal injury. J Neurotrauma 2001; 18: 241–246.

  9. 9

    Battistuta J, Bjartmar C, Trapp BD . Postmortem degradation of N-acetylaspartate and N-acetyl aspartylglutamate: an HPLC analysis of different rat CNS regions. Neurochem Res 2001; 26: 695–702.

  10. 10

    Vrenken H, Barkhof F, Uitdehaag BM, Castelijns JA, Polman CH, Pouwels PJ . MR spectroscopic evidence for glial increase but not for neuro-axonal damage in MS normal-appearing white matter. Magn Reson Med 2005; 53: 256–266.

  11. 11

    Novotny Jr EJ, Fulbright RK, Pearl PL, Gibson KM, Rothman DL . Magnetic resonance spectroscopy of neurotransmitters in human brain. Ann Neurol 2003; 54 (Suppl 6): S25–S31.

  12. 12

    Rothman DL, Sibson NR, Hyder F, Shen J, Behar KL, Shulman RG . In vivo nuclear magnetic resonance spectroscopy studies of the relationship between the glutamate–glutamine neurotransmitter cycle and functional neuroenergetics. Philos Trans R Soc Lond B Biol Sci 1999; 354: 1165–1177.

  13. 13

    Erecinska M, Silver IA . ATP and brain function. J Cereb Blood Flow Metab 1989; 9: 2–19.

  14. 14

    Miller BL . A review of chemical issues in 1H NMR spectroscopy: N-acetyl-L-aspartate, creatine and choline. NMR Biomed 1991; 4: 47–52.

  15. 15

    Irvine RF, Schell MJ . Back in the water: the return of the inositol phosphates. Nat Rev Mol Cell Biol 2001; 2: 327–338.

  16. 16

    Abbott C, Bustillo J . What have we learned from proton magnetic resonance spectroscopy about schizophrenia? A critical update. Curr Opin Psychiatry 2006; 19: 135–139.

  17. 17

    Wobrock T, Scherk H, Falkai P . [Magnetic resonance spectroscopy in schizophrenia. Possibilities and limitations]. Radiologe 2005; 45: 124–130, 132–126.

  18. 18

    Keshavan MS, Stanley JA, Pettegrew JW . Magnetic resonance spectroscopy in schizophrenia: methodological issues and findings—part II. Biol Psychiatry 2000; 48: 369–380.

  19. 19

    Sanches RF, Crippa JA, Hallak JE, Araujo D, Zuardi AW . Proton magnetic resonance spectroscopy of the frontal lobe in schizophrenics: a critical review of the methodology. Rev Hosp Clin Fac Med Sao Paulo 2004; 59: 145–152.

  20. 20

    Steen RG, Hamer RM, Lieberman JA . Measurement of brain metabolites by 1H magnetic resonance spectroscopy in patients with schizophrenia: a systematic review and meta-analysis. Neuropsychopharmacology 2005; 30: 1949–1962.

  21. 21

    Bertolino A, Nawroz S, Mattay VS, Barnett AS, Duyn JH, Moonen CT et al. Regionally specific pattern of neurochemical pathology in schizophrenia as assessed by multislice proton magnetic resonance spectroscopic imaging. Am J Psychiatry 1996; 153: 1554–1563.

  22. 22

    Selemon LD, Rajkowska G, Goldman-Rakic PS . Elevated neuronal density in prefrontal area 46 in brains from schizophrenic patients: application of a three-dimensional, stereologic counting method. J Comp Neurol 1998; 392: 402–412.

  23. 23

    Selemon LD, Rajkowska G, Goldman-Rakic PS . Abnormally high neuronal density in the schizophrenic cortex. A morphometric analysis of prefrontal area 9 and occipital area 17. Arch Gen Psychiatry 1995; 52: 805–818; discussion 819–820.

  24. 24

    Bertolino A, Callicott JH, Elman I, Mattay VS, Tedeschi G, Frank JA et al. Regionally specific neuronal pathology in untreated patients with schizophrenia: a proton magnetic resonance spectroscopic imaging study. Biol Psychiatry 1998; 43: 641–648.

  25. 25

    Cecil KM, Lenkinski RE, Gur RE, Gur RC . Proton magnetic resonance spectroscopy in the frontal and temporal lobes of neuroleptic naive patients with schizophrenia. Neuropsychopharmacology 1999; 20: 131–140.

  26. 26

    Renshaw PF, Yurgelun-Todd DA, Tohen M, Gruber S, Cohen BM . Temporal lobe proton magnetic resonance spectroscopy of patients with first-episode psychosis. Am J Psychiatry 1995; 152: 444–446.

  27. 27

    Bertolino A, Sciota D, Brudaglio F, Altamura M, Blasi G, Bellomo A et al. Working memory deficits and levels of N-acetylaspartate in patients with schizophreniform disorder. Am J Psychiatry 2003; 160: 483–489.

  28. 28

    Block W, Bayer TA, Tepest R, Traber F, Rietschel M, Muller DJ et al. Decreased frontal lobe ratio of N-acetylaspartate to choline in familial schizophrenia: a proton magnetic resonance spectroscopy study. Neurosci Lett 2000; 289: 147–151.

  29. 29

    Callicott JH, Egan MF, Bertolino A, Mattay VS, Langheim FJ, Frank JA et al. Hippocampal N-acetylaspartate in unaffected siblings of patients with schizophrenia: a possible intermediate neurobiological phenotype. Biol Psychiatry 1998; 44: 941–950.

  30. 30

    Thomas MA, Ke Y, Levitt J, Caplan R, Curran J, Asarnow R et al. Preliminary study of frontal lobe 1H MR spectroscopy in childhood-onset schizophrenia. J Magn Reson Imaging 1998; 8: 841–846.

  31. 31

    Jessen F, Scherk H, Traber F, Theyson S, Berning J, Tepest R et al. Proton magnetic resonance spectroscopy in subjects at risk for schizophrenia. Schizophr Res 2006; 87: 81–88.

  32. 32

    Farber NB, Kim SH, Dikranian K, Jiang XP, Heinkel C . Receptor mechanisms and circuitry underlying NMDA antagonist neurotoxicity. Mol Psychiatry 2002; 7: 32–43.

  33. 33

    Olney JW, Newcomer JW, Farber NB . NMDA receptor hypofunction model of schizophrenia. J Psychiatr Res 1999; 33: 523–533.

  34. 34

    Stanley JA . In vivo magnetic resonance spectroscopy and its application to neuropsychiatric disorders. Can J Psychiatry 2002; 47: 315–326.

  35. 35

    Fenton WS, Hibbeln J, Knable M . Essential fatty acids, lipid membrane abnormalities, and the diagnosis and treatment of schizophrenia. Biol Psychiatry 2000; 47: 8–21.

  36. 36

    Horrobin DF, Glen AI, Vaddadi K . The membrane hypothesis of schizophrenia. Schizophr Res 1994; 13: 195–207.

  37. 37

    Auer DP, Wilke M, Grabner A, Heidenreich JO, Bronisch T, Wetter TC . Reduced NAA in the thalamus and altered membrane and glial metabolism in schizophrenic patients detected by 1H-MRS and tissue segmentation. Schizophr Res 2001; 52: 87–99.

  38. 38

    Theberge J, Al-Semaan Y, Drost DJ, Malla AK, Neufeld RW, Bartha R et al. Duration of untreated psychosis vs N-acetylaspartate and choline in first episode schizophrenia: a 1H magnetic resonance spectroscopy study at 4.0 Tesla. Psychiatry Res 2004; 131: 107–114.

  39. 39

    Sharma R, Venkatasubramanian PN, Barany M, Davis JM . Proton magnetic resonance spectroscopy of the brain in schizophrenic and affective patients. Schizophr Res 1992; 8: 43–49.

  40. 40

    Bluml S, Tan J, Harris K, Adatia N, Karme A, Sproull T et al. Quantitative proton-decoupled 31P MRS of the schizophrenic brain in vivo. J Comput Assist Tomogr 1999; 23: 272–275.

  41. 41

    Auer DP, Putz B, Kraft E, Lipinski B, Schill J, Holsboer F . Reduced glutamate in the anterior cingulate cortex in depression: an in vivo proton magnetic resonance spectroscopy study. Biol Psychiatry 2000; 47: 305–313.

  42. 42

    Lim KO, Adalsteinsson E, Spielman D, Sullivan EV, Rosenbloom MJ, Pfefferbaum A . Proton magnetic resonance spectroscopic imaging of cortical gray and white matter in schizophrenia. Arch Gen Psychiatry 1998; 55: 346–352.

  43. 43

    Steel RM, Bastin ME, McConnell S, Marshall I, Cunningham-Owens DG, Lawrie SM et al. Diffusion tensor imaging (DTI) and proton magnetic resonance spectroscopy (1H MRS) in schizophrenic subjects and normal controls. Psychiatry Res 2001; 106: 161–170.

  44. 44

    Ende G, Braus DF, Walter S, Weber-Fahr W, Soher B, Maudsley AA et al. Effects of age, medication, and illness duration on the N-acetylaspartate signal of the anterior cingulate region in schizophrenia. Schizophr Res 2000; 41: 389–395.

  45. 45

    Yamasue H, Fukui T, Fukuda R, Yamada H, Yamasaki S, Kuroki N et al. 1H-MR spectroscopy and gray matter volume of the anterior cingulate cortex in schizophrenia. Neuroreport 2002; 13: 2133–2137.

  46. 46

    Theberge J, Al-Semaan Y, Williamson PC, Menon RS, Neufeld RW, Rajakumar N et al. Glutamate and glutamine in the anterior cingulate and thalamus of medicated patients with chronic schizophrenia and healthy comparison subjects measured with 4.0-T proton MRS. Am J Psychiatry 2003; 160: 2231–2233.

  47. 47

    Theberge J, Bartha R, Drost DJ, Menon RS, Malla A, Takhar J et al. Glutamate and glutamine measured with 4.0 T proton MRS in never-treated patients with schizophrenia and healthy volunteers. Am J Psychiatry 2002; 159: 1944–1946.

  48. 48

    Keshavan MS, Montrose DM, Pierri JN, Dick EL, Rosenberg D, Talagala L et al. Magnetic resonance imaging and spectroscopy in offspring at risk for schizophrenia: preliminary studies. Prog Neuropsychopharmacol Biol Psychiatry 1997; 21: 1285–1295.

  49. 49

    Tibbo P, Hanstock C, Valiakalayil A, Allen P . 3-T proton MRS investigation of glutamate and glutamine in adolescents at high genetic risk for schizophrenia. Am J Psychiatry 2004; 161: 1116–1118.

  50. 50

    Omori M, Pearce J, Komoroski RA, Griffin WS, Mrak RE, Husain MM et al. In vitro 1H-magnetic resonance spectroscopy of postmortem brains with schizophrenia. Biol Psychiatry 1997; 42: 359–366.

  51. 51

    Jentsch JD, Roth RH . The neuropsychopharmacology of phencyclidine: from NMDA receptor hypofunction to the dopamine hypothesis of schizophrenia. Neuropsychopharmacology 1999; 20: 201–225.

  52. 52

    Laruelle M, Kegeles LS, Abi-Dargham A . Glutamate, dopamine, and schizophrenia: from pathophysiology to treatment. Ann NY Acad Sci 2003; 1003: 138–158.

  53. 53

    Olney JW, Farber NB . Glutamate receptor dysfunction and schizophrenia. Arch Gen Psychiatry 1995; 52: 998–1007.

  54. 54

    Kaprio J, Koskenvuo M . Genetic and environmental factors in complex diseases: the older Finnish Twin Cohort. Twin Res 2002; 5: 358–365.

  55. 55

    Cannon TD, Kaprio J, Lonnqvist J, Huttunen M, Koskenvuo M . The genetic epidemiology of schizophrenia in a Finnish twin cohort. A population-based modeling study. Arch Gen Psychiatry 1998; 55: 67–74.

  56. 56

    Spitzer RL, Williams JBW, Gibbon M, First MB . Instruction Manual for the Structured Clinical Interview for DSM-III-R (SCID). Biometrics Research: New York, NY, 1989.

  57. 57

    American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn, American Psychiatric Association: Washington, DC, 1994.

  58. 58

    Loranger AW, Susman VL, Oldham JM, Russakoff LM . Personality Disorder Examination: A Structured Interview for Making Diagnosis of DSM-III-R Personality Disorders. Cornell Medical College: White Plains, NY, 1985.

  59. 59

    Cavassila S, Deval S, Huegen C, van Ormondt D, Graveron-Demilly D . Cramer-Rao bounds: an evaluation tool for quantitation. NMR Biomed 2001; 14: 278–283.

  60. 60

    Purdon SE, Valiakalayil A, Hanstock CC, Seres P, Tibbo P . Elevated 3 T proton MRS glutamate levels associated with poor Continuous Performance Test (CPT-0X) scores and genetic risk for schizophrenia. Schizophrenia Res 2008; 99: 218–224.

  61. 61

    Chang L, Friedman J, Ernst T, Zhong K, Tsopelas ND, Davis K . Brain metabolite abnormalities in the white matter of elderly schizophrenic subjects: implication for glial dysfunction. Biol Psychiatry 2007; 62: 1396–1404.

  62. 62

    Wood SJ, Yücel M, Wellard RM, Harrison BJ, Clarke K, Fornito A et al. Evidence for neuronal dysfunction in the anterior cingulate of patients with schizophrenia: a proton magnetic resonance spectroscopy study at 3 T. Schizophrenia Res 2007; 94: 328–331.

  63. 63

    Mangia S, Tkác I, Gruetter R, Van De Moortele P-F, Giove F, Maraviglia B et al. Sensitivity of single-voxel 1H-MRS in investigating the metabolism of the activated human visual cortex at 7T. Magn Reson Imaging 2006; 24: 343–348.

  64. 64

    Venkatraman TN, Hamer RM, Perkins DO, Song AW, Lieberman JA, Steen RG . Single-voxel 1H PRESS at 4.0 T: precision and variability of measurements in anterior cingulate and hippocampus. NMR Biomed 2006; 19: 484–491.

  65. 65

    Oz G, Terpstra M, Tkác I, Aia P, Lowary J, Tuite PJ et al. Proton MRS of the unilateral substantia nigra in the human brain at 4 tesla: detection of high GABA concentrations. Magn Reson Med 2006; 55: 296–301.

  66. 66

    Posse S, Otazo R, Caprihan A, Bustillo J, Chen H, Henry PG et al. Proton echo-planar spectroscopic imaging of J-coupled resonances in human brain at 3 and 4 Tesla. Magn Reson Med 2007; 58: 236–244.

  67. 67

    Provencher SW . Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magn Reson Med 1993; 30: 672–679.

  68. 68

    Zhang Y, Brady M, Smith S . Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Trans Med Imaging 2001; 20: 45–57.

  69. 69

    Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H et al. Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 2004; 23 (Suppl 1): S208–S219.

  70. 70

    Schuff N, Amend D, Ezekiel F, Steinman SK, Tanabe J, Norman D et al. Changes of hippocampal N-acetylaspartate and volume in Alzheimer's disease. A proton MR spectroscopic imaging and MRI study. Neurology 1997; 49: 1513–1521.

  71. 71

    Lynch J, Peeling J, Auty A, Sutherland GR . Nuclear magnetic resonance study of cerebrospinal fluid from patients with multiple sclerosis. Can J Neurol Sci 1993; 20: 194–198.

  72. 72

    McLean MA, Woermann FG, Barker GJ, Duncan JS . Quantitative analysis of short echo time 1H-MRSI of cerebral gray and white matter. Magn Reson Med 2000; 44: 401–411.

  73. 73

    Kalra S, Arnold DL . Magnetic resonance spectroscopy for monitoring neuronal integrity in amyotrophic lateral sclerosis. Adv Exp Med Biol 2006; 576: 275–282; discussion 361–273.

  74. 74

    Olbrich HM, Valerius G, Rüsch N, Buchert M, Thiel T, Hennig J et al. Frontolimbic glutamate alterations in first episode schizophrenia: evidence from a magnetic resonance spectroscopy study. World J Biol Psychiatry 2008; 9: 59–63.

  75. 75

    van Elst LT, Valerius G, Büchert M, Thiel T, Rüsch N, Bubl E et al. Increased prefrontal and hippocampal glutamate concentration in schizophrenia: evidence from a magnetic resonance spectroscopy study. Biol Psychiatry 2005; 58: 724–730.

  76. 76

    Rowland LM, Bustillo JR, Mullins PG, Jung RE, Lenroot R, Landgraf E et al. Effects of ketamine on anterior cingulate glutamate metabolism in healthy humans: a 4-T proton MRS study. Am J Psychiatry 2005; 162: 394–396.

  77. 77

    Carlsson A, Waters N, Carlsson ML . Neurotransmitter interactions in schizophrenia—therapeutic implications. Biol Psychiatry 1999; 46: 1388–1395.

  78. 78

    Goff DC, Coyle JT . The emerging role of glutamate in the pathophysiology and treatment of schizophrenia. Am J Psychiatry 2001; 158: 1367–1377.

  79. 79

    Javitt DC, Zukin SR . Recent advances in the phencyclidine model of schizophrenia. Am J Psychiatry 1991; 148: 1301–1308.

  80. 80

    Tamminga CA, Holcomb HH, Gao XM, Lahti AC . Glutamate pharmacology and the treatment of schizophrenia: current status and future directions. Int Clin Psychopharmacol 1995; 10 (Suppl 3): 29–37.

  81. 81

    Barker PB, Bonekamp D, Riedy G, Smith M . Quantitation of NAA in the brain by magnetic resonance spectroscopy. Adv Exp Med Biol 2006; 576: 183–197; discussion 361–183.

  82. 82

    Soher BJ, van Zijl PC, Duyn JH, Barker PB . Quantitative proton MR spectroscopic imaging of the human brain. Magn Reson Med 1996; 35: 356–363.

  83. 83

    Hetherington HP, Mason GF, Pan JW, Ponder SL, Vaughan JT, Twieg DB et al. Evaluation of cerebral gray and white matter metabolite differences by spectroscopic imaging at 4.1 T. Magn Reson Med 1994; 32: 565–571.

  84. 84

    Stockler S, Holzbach U, Hanefeld F, Marquardt I, Helms G, Requart M et al. Creatine deficiency in the brain: a new, treatable inborn error of metabolism. Pediatr Res 1994; 36: 409–413.

  85. 85

    Buckley PF, Moore C, Long H, Larkin C, Thompson P, Mulvany F et al. 1H-magnetic resonance spectroscopy of the left temporal and frontal lobes in schizophrenia: clinical, neurodevelopmental, and cognitive correlates. Biol Psychiatry 1994; 36: 792–800.

  86. 86

    Bustillo JR, Lauriello J, Rowland LM, Jung RE, Petropoulos H, Hart BL et al. Effects of chronic haloperidol and clozapine treatments on frontal and caudate neurochemistry in schizophrenia. Psychiatry Res 2001; 107: 135–149.

  87. 87

    Ohrmann P, Siegmund A, Suslow T, Pedersen A, Spitzberg K, Kersting A et al. Cognitive impairment and in vivo metabolites in first-episode neuroleptic-naive and chronic medicated schizophrenic patients: a proton magnetic resonance spectroscopy study. J Psychiatr Res 2006; Aug: 30.

  88. 88

    Deicken RF, Zhou L, Corwin F, Vinogradov S, Weiner MW . Decreased left frontal lobe N-acetylaspartate in schizophrenia. Am J Psychiatry 1997; 154: 688–690.

  89. 89

    Wood SJ, Berger G, Velakoulis D, Phillips LJ, McGorry PD, Yung AR et al. Proton magnetic resonance spectroscopy in first episode psychosis and ultra high-risk individuals. Schizophr Bull 2003; 29: 831–843.

  90. 90

    Kim H, McGrath BM, Silverstone PH . A review of the possible relevance of inositol and the phosphatidylinositol second messenger system (PI-cycle) to psychiatric disorders—focus on magnetic resonance spectroscopy (MRS) studies. Hum Psychopharmacol 2005; 20: 309–326.

  91. 91

    Delamillieure P, Constans JM, Fernandez J, Brazo P, Benali K, Courtheoux P et al. Proton magnetic resonance spectroscopy (1H MRS) in schizophrenia: investigation of the right and left hippocampus, thalamus, and prefrontal cortex. Schizophr Bull 2002; 28: 329–339.

  92. 92

    Delamillieure P, Constans J, Fernandez J, Brazo P, Dollfus S . Proton magnetic resonance spectroscopy (1H-MRS) of the thalamus in schizophrenia. Eur Psychiatry 2000; 15: 489–491.

  93. 93

    Harris RE, Sundgren PC, Pang Y, Hsu M, Petrou M, Kim SH et al. Dynamic levels of glutamate within the insula are associated with improvements in multiple pain domains in fibromyalgia. Arthritis Rheum 2008; 58: 903–907.

  94. 94

    Shibuya-Tayoshi S, Tayoshi S, Sumitani S, Ueno S, Harada M, Ohmori T . Lithium effects on brain glutamatergic and GABAergic systems of healthy volunteers as measured by proton magnetic resonance spectroscopy. Prog Neuropsychopharmacol Biol Psychiatry 2008; 32: 249–256.

  95. 95

    Shulman Y, Grant S, Seres P, Hanstock C, Baker G, Tibbo P . The relation between peripheral and central glutamate and glutamine in healthy male volunteers. J Psychiatry Neuroscience 2006; 31: 406–410.

  96. 96

    Gallinat J, Kunz D, Lang UE, Neu P, Kassim N, Kienast T et al. Association between cerebral glutamate and human behaviour: the sensation seeking personality trait. Neuroimage 2007; 34: 671–678.

  97. 97

    Schubert F, Gallinat J, Seifert F, Rinneberg H . Glutamate concentrations in human brain using single voxel proton magnetic resonance spectroscopy at 3 Tesla. Neuroimage 2004; 21: 1762–1771.

Download references


This work was supported by a research grant from the National Institute of Mental Health (to TDC). From the Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki, Finland, we thank Ulla Mustonen, Tiia Pirkola, Pirjo Käki, Eila Voipio and Annamari Tuulio-Henriksson for their contributions to subject recruitment and assessment; Antti Tanksanen for his contributions to the register searches; Kauko Heikkilä for his contributions to data management of the Finnish Twin Cohort Study and Sami Heikkinen for his work in setting up the protocol. We also thank Dr Provencher for providing the LCModel basis set. Finally, we thank the subjects, whose participation made this project possible.

Author information

Correspondence to E S Lutkenhoff.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lutkenhoff, E., van Erp, T., Thomas, M. et al. Proton MRS in twin pairs discordant for schizophrenia. Mol Psychiatry 15, 308–318 (2010) doi:10.1038/mp.2008.87

Download citation


  • schizophrenia
  • magnetic resonance spectroscopy
  • glutamate
  • NAA
  • creatine
  • twins

Further reading