PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Schizophr Res. Author manuscript; available in PMC 2007 March 30.
Published in final edited form as:
PMCID: PMC1839817
NIHMSID: NIHMS15978

Structural Analysis of the Basal Ganglia in Schizophrenia

Abstract

Increases in the total volume of basal ganglia structures have been reported in schizophrenia. However, patterns of basal ganglia shape change, which can reveal localized changes in substructure volumes, have not been investigated. In this study, the total volume and shape of several basal ganglia structures were compared in subjects with and without schizophrenia.

T1-weighted magnetic resonance scans were collected in 54 schizophrenia and 70 comparison subjects. High-dimensional (large-deformation) brain mapping was used to assess the shape and volume of several basal ganglia structures. The relationships of shape and volume measures with psychopathology, cognition and motor function were also assessed.

Left and right volumes of the caudate and putamen, as well as the right globus pallidus volume, were significantly increased in subjects with schizophrenia as compared to comparison subjects after total brain volume was included as a covariate. Significant differences in shape accompanied these volume changes in the caudate, putamen and globus pallidus, after their total volumes were included as covariates. There were few significant correlations between volume or shape measures and either cognitive function or clinical symptoms, other than a positive correlation between an attention/vigilance cognitive dimension and the volume of the caudate and putamen, and a negative correlation between nucleus accumbens volume and delusions.

In conclusion, basal ganglia volumes relative to total brain volume were larger in schizophrenia subjects than healthy comparison subjects. Specific patterns of shape change accompanied these volume differences.

1. INTRODUCTION

The basal ganglia are a collection of nuclei deep within the cerebrum. These nuclei include the caudate, the nucleus accumbens and the putamen - which are collectively called the striatum, and also the globus pallidus, the subthalamic nucleus and the substantia nigra. Through their extensive cortical connections, the basal ganglia can influence both motor and cognitive functions (Parent & Hazrati, 1995; DeLong, 2000). There has been increasing evidence for the involvement of the basal ganglia in cognitive and behavioral syndromes (Levy et al, 1997; Mendez et al, 1989). Also, emotional and cognitive dysfunction has been observed in basal ganglia-related movement disorders (Zgaljardic et al, 2003; Joel, 2001).

Abnormal activity of the basal ganglia has been reported in subjects with schizophrenia at rest and during various cognitive tasks by several investigators (Manoach et al, 2000; Menon et al, 2001). However, structural imaging studies of the basal ganglia in schizophrenia have yielded less consistent results. Most investigators report enlargement of the volumes of various basal ganglia (Staal et al 2000; Breier et al, 1992), although normal (Gunduz et al, 2002) or even decreased volumes (Corson et al, 1999; Keshavan et al, 1998) have also been reported. Basal ganglia enlargement, when found, has usually been interpreted to be the result of exposure to antipsychotic medications.

In contrast to volume studies, there have been few studies of basal ganglia shape or conformation in schizophrenia (however see Shihabuddin et al, 1998). Shape assessment can be used to demonstrate subtle abnormalities in the contouring of a structure that reflect localized changes in regional subvolumes (Csernansky et al, 1998; 2002). Also, comparing the shape of a structure can allow for better discrimination between normal and pathologic conditions than that observed by comparing the volume alone (Csernansky et al, 2002, 2004). We have previously used large-deformation high-dimensional brain mapping (HDBM-LD; Haller et al, 1997; Wang et al, 2001) to characterize the shape of the hippocampus (Csernansky et al, 2002) and thalamus (Csernansky et al, 2004) in patients with schizophrenia.

In the present study, we used HDBM-LD to compare the shape, as well as the symmetry and volume, of several basal ganglia structures in 54 schizophrenia subjects and 70 healthy subjects. The relationships between the neuroanatomical measures (i.e. volume and shape) and selected clinical and cognitive features of the subjects were assessed in an exploratory analysis.

2. METHODS

2.1 Subjects

The demographic and clinical characteristics of the 54 schizophrenia and 70 comparison subjects are summarized in TABLE 1. The majority of these subjects were included in prior studies of hippocampal and thalamic shape (Csernansky et al, 2002; 2004). All subjects were diagnosed using DSM-IV criteria on the basis of a consensus between a research psychiatrist who conducted a semi-structured interview and a trained research assistant who used the Structured Clinical Interview for DSM-IV Axis I Disorders (First et al, 1995). The healthy comparison subjects had no prior history of mental illness, nor any first-degree relative with a psychotic disorder. Subjects were excluded if they had neurologic disorders, unstable medical disorders, head injury with loss of consciousness, or if they met DSM-IV criteria for substance abuse or dependence during the 3 months preceding the study. A distant lifetime history of substance abuse or dependence was reported by fourteen schizophrenic subjects and seven comparison subjects. Handedness was evaluated in all subjects (Oldfield, 1971).

TABLE 1
Demographic and Clinical Characteristics of Schizophrenia and Healthy Comparison Subjects. Values are means (standard deviation) unless stated otherwise

All schizophrenic subjects were clinically stable; the global severity of their symptoms had remained unchanged for at least 2 weeks. 19 of the schizophrenia subjects had one or more extrapyramidal motor symptoms (dyskinesia, dystonia or parkinsonism) raging in severity from borderline to moderately severe. In the subjects who were receiving antipsychotic drugs, their most recent (last four weeks) drug treatment was categorized as either typical or atypical. Atypical antipsychotic drugs included risperidone, olanzapine, clozapine and quetiapine. Typical antipsychotic drugs included haloperidol, thiothixene, and fluphenazine. The median duration of treatment was 12 weeks (range 1 to 520 weeks) with atypical drugs and 78 weeks (range 2 to 468 weeks) with typical drugs.

2.2 Rating of Clinical Function

The severity of psychopathology was assessed in the schizophrenia subjects using the Scale for the Assessment of Positive Symptoms (SAPS) and the Scale for the Assessment of Negative Symptoms (SANS) (Andreasen et al, 1995; Andreasen & Olsen, 1982).

To investigate relationships between neuroanatomical variables and specific domains of cognitive function, a principal component analysis was applied to data from a battery of neuropsychological tests. The principal axis method was used to extract the components, followed by a Varimax (orthogonal) rotation. Significant test items and corresponding factor loadings are presented in TABLE 2. This PCA identified some factors that were similar to, although not identical to, cognitive domains previously reported in groups of schizophrenic patients (Nuechterlein et al, 2004).

TABLE 2
Rotational factor pattern from principal component analysis of neuropsychological tests administered to schizophrenic patients (N=54). Factor loadings greater than .40 are in bold.

Extrapyramidal motor symptoms of schizophrenia subjects were evaluated by a psychiatrist using the Extrapyramidal Symptom Rating Scale (ESRS) (Gharabawi GM et al, 2005). From the ESRS, performance on the clinical global impression of severity of dyskinesia, parkinsonism and dystonia were used to assess correlations between motor symptoms and neuroanatomical measures.

2.3 Image Acquisition and Preprocessing

Magnetic resonance (MR) scans were collected using a turbo-fast low-angle shot (turbo-FLASH) sequence (TR=20, TE=5.4, flip angle=30 degrees, number of acquisitions=1, matrix=256×256, scanning time=13.5 minutes) that acquired three-dimensional datasets with 1 mm × 1 mm × 1 mm isotropic voxels across the entire cranium (Venkatesan and Haacke, 1997). Raw MR data were reformatted for analysis using Analyze software (Rochester, Minn.), and signed 16-bit MR datasets were compressed to unsigned 8-bit MR datasets by linearly rescaling voxel intensities such that voxels with intensity levels at two standard deviations above the mean of white matter (corpus callosum) were mapped to 255, and voxels with intensity levels at two standard deviations below the mean of CSF (lateral ventricle) were mapped to 0. The white matter and CSF means and standard deviations were obtained by sampling voxels from these respective regions.

Landmarks were placed in all scans at the external boundaries of the brain, and at points where the anterior and posterior commissures intersect the midsagittal plane (Haller et al, 1997). Additional landmarks were also placed at selected points throughout structures of interest as previously described (Wang et al 2006). These landmarks provided starting values for HDBM-LD (see below).

2.4 Large-Deformation High Dimensional Brain Mapping (HDBM-LD)

An MR scan collected from a healthy comparison subject not otherwise included in the study was used to construct the neuroanatomical template. The basal ganglia in the right hemisphere were manually outlined in this scan by expert raters (MG, LW) according to a priori neuroanatomical guidelines. Target scans were landmarked at defined positions within the basal ganglia-thalamus complex that corresponded to landmarks placed in the template scan.

Transformation of the template onto the target MR scans occurred in a two-step process. First, it was coarsely aligned to the left and right sides of each target scan by using the landmarks, and then a probabilistic, large-deformation transformation was applied to it (Miller et al, 1997). During this transformation, the movement and deformation of template voxels were constrained by assigning them the physical properties of a fluid. The reliability and validity of high-dimensional brain mapping forsegmenting subcortical structures with respect to expert manual outlining has been previously demonstrated (Haller et al, 1997). To check the validity of high-dimensional brain mapping as used in this study, we compared the segmentations generated by this process to those manually outlined by experts (M.G., L.W.) in the MR scans of a randomly selected subgroup of ten subjects. Values estimating the overlap of caudate, putamen and globus pallidus contours produced by HDBM versus manual outlining averages exceeded 80% in all subjects, which is comparable to the accuracy of repeated attempts at manual outlining of these structures by the same expert. The values estimating overlap of the nucleus accumbens contours produced by HDBM versus manual outlining approached but did not exceed 80% in all subjects. Total cerebral volumes were derived using elastic-based transformations of the template scan (Miller et al 1997) so that comparisons of structural volume could be performed using total cerebral volume as a covariate.

To quantify the shape and volume of individual structures, a triangulated surface was first superimposed onto each structure outlined in the template scan. These surfaces were then carried along as the template scan was transformed to match left and right sides of each of the target scans. Volumes of the selected basal ganglia were calculated by computing the volumes enclosed by the transformed surfaces. To compare structural shapes between subject groups, vector fields were derived from the displacements of the triangulated surface points during the transformations. The pooled covariance of the vector fields yielded eigenvalues and a complete orthonormal set of eigenvectors representing shape variation for the population under study via singular value decomposition (Joshi et al 1997). Coefficients (eigenscores) associated with these eigenvalues and eigenvectors were then calculated for each structures in each hemisphere jn every subject. Eigenscores based on the fewest number of eigenvectors needed to explain ~75% of the total variance (i.e. 10) were then used for statistical analysis.

2.5 Statistical Analysis

All statistical analyses were performed using SAS 9.0 software. Structural volumes were analyzed using one-way analysis of variance with diagnostic group as an independent variable. Group comparisons of individual structural volumes were repeated using cerebral volume as a covariate, since cerebral volume (F=4.59, df=1,122, p<0.05; see TABLE 3) but not intracranial volume (F= 3.7, df=1,122, p=0.06) showed significant group differences in volume.

TABLE 3
Cerebral and basal ganglia volumes in schizophrenic subjects (n=54) and healthy comparison subjects (N=70). Values are displayed as volumes (standard deviation).

To compare the shape of the basal ganglia between groups, the first 10 eigenvectors representing variation in the shape of each structure were selected a priori. These eigenvectors were then used in a one-way multiple analysis of variance, with diagnostic group as an independent variable, to test the hypothesis there was a significant group effect on shape. Then, to identify the eigenvectors that contributed most to group discrimination, a logistic backward regression was performed using a significant likelihood ratio statistic for discrimination. The eigenvectors selected by the logistic regression model were later used in a “leave-one-out” discrimination function analysis to determine the percentage of correctly classified subjects in each group.

To visualize the physical pattern of basal ganglia shape difference between groups, we reconstructed maps of the composite surfaces of individual structures in the schizophrenia and comparison subjects at every point on the graphical surfaces for each of the structures. The displacements were calculated at each surface point as the difference between the means of the group vectors in magnitude.

Bivariate correlations between volume measures and clinical measures of psychopathology, motor function, and performance on the various cognitive domains were examined in the schizophrenia subjects on an exploratory basis, using non-parametric statistics. To examine the relationship between shape and clinical measures, the canonical variate discriminating groups was determined separately for each of the basal ganglia structures using its shape eigenvectors. Correlations between these canonical variables and clinical measures were determined using bivariate statistics.

3. RESULTS

3.1 Combined Basal Ganglia Volumes

Mean combined volume of all basal ganglia structures that were assessed was 19972 mm3 (SD= 2284) in the schizophrenic subjects and 19930 mm3 (SD=2061) in the comparison subjects. The volume difference between groups was not statistically significant. After covarying the combined basal ganglia volumes for total cerebral volume, the effect of diagnosis became significant (F=7.58, df=1, 122, p<0.007).

3.2 Caudate Volume and Shape

The effect of diagnosis on left and right caudate volumes was not significant (TABLE 3). After covarying caudate volumes for total cerebral volumes, the effect of diagnosis became significant on the left (F=10.76, df=1, 122, p<0.002; see FIGURE 1A) and right (F=7.51, df=1, 122, p<0.01; see FIGURE 1B). In both cases the schizophrenia subjects had larger caudate volumes relative to total cerebral volumes than the comparison subjects. This result was not affected by covarying caudate volumes for age, gender, race, handedness, type of antipsychotic drug treatment (i.e., atypical versus typical), or a lifetime history of substance abuse/dependence.

FIGURE 1FIGURE 1FIGURE 1FIGURE 1
Volumes of basal ganglia structures in relation to cerebral volume in schizophrenia subjects (N=54) and healthy comparison subjects (N=70). Solid line (and filled circle) represent comparison subjects; and dashed line (and open circle) represent schizophrenia ...

Using the first 10 eigenvectors representing the shape of the left and right caudate nucleus, a MANOVA revealed a trend towards a significant group effect (Wilks' Lambda = 0.86, p=0.07). The group effect became highly significant after including caudate volume as a covariate in the analysis (Wilks' Lambda = 0.79, p=0.002). Eigenvectors 3 and 8 maximally discriminated the two subject groups. In a “leave one out” discriminant function analysis using these eigenvectors, 32 (59.3%) schizophrenic and 42 (60.0%) comparison subjects were correctly classified. When total caudate volume was included in the analysis, 33 (61.1%) schizophrenic and 47 (67.1%) comparison subjects were correctly classified. Visual representations of the group difference in caudate shape is shown in FIGURE 2A.

FIGURE 2FIGURE 2FIGURE 2FIGURE 2
Surface maps depicting basal ganglia shape difference in schizophrenia subjects (N=54) and healthy comparison subjects (N=70). Blue-to-purple shading denotes regions of inward deformity of the basal ganglia surfaces in schizophrenia versus comparison ...

3.3 Nucleus Accumbens Volume and Shape

The group effect on nucleus accumbens volumes was nonsignificant before and after adding total cerebral volume as a covariate. (TABLE 3; FIGURE FIGURE1C1C and and1D).1D). This finding was not altered by adding age, gender, race, handedness, type of antipsychotic drug or a lifetime history of substance abuse/dependence as covariates.

Using the first 10 eigenvectors representing the shape of the left and right nucleus accumbens, a MANOVA suggested no significant difference in the shape of the nucleus accumbens between the groups and this result was not altered by adding total cerebral volume as a covariate.

3.4 Putamen Volume and Shape

The effect of diagnosis on left and right putamen volumes was not significant (TABLE 3). After covarying putamen volumes for total cerebral volumes, the effect of diagnosis became significant on the left (F=3.30, df=1, 122, p<0.01; FIGURE 1E) and the right (F=7.01, df=1, 122, p<0.01; FIGURE 1F). This result was not altered by adding age, gender, race, handedness, type of antipsychotic drug or a lifetime history of substance abuse/dependence as covariates.

Using the first 10 eigenvectors representing the shape of the left and right putamen, there was significant group effect (F=2.31, p=0.02). The group effect became more significant after including caudate volume as a covariate in the analysis (Wilks' Lambda = 0.77, p=0.0012). Eigenvectors 2, 4 and 8 maximally discriminated the subject groups. In a “leave one out” discriminant function analysis using these eigenvectors, 29 (53.7%) of the schizophrenia subjects and 40 (57.1%) of the comparison subjects were correctly classified. When total putamen volume was include in the analysis, 34 (63.0%) of the schizophrenic subjects and 44 (63.0%) of the comparison subjects were correctly classified. Visual representations of the group difference in putamen shape is shown in FIGURE 2B.

3.5 Globus Pallidus Volume and Shape

The group effect on left and right globus pallidus volumes was not significant (TABLE 3), and after covarying globus pallidus volumes for total cerebral volumes, the group effect remained non-significant on the left (F=0.15, df=1, 122, p<0.7; FIGURE 1G), but became significant on the right (F=4.14, df=1, 122, p<0.05; FIGURE 1H). This result was not altered by adding age, gender, race, handedness, type of antipsychotic drug or a lifetime history of substance abuse/dependence as covariates.

Using the first 10 eigenvectors representing the shape of the left and right globus pallidus, the group effect almost reached significance (Wilks' Lambda = 0.86, p=0.06) between the two subject groups. The group effect reached statistical significance after globus pallidus volume was included as a covariate in the analysis (Wilks' Lambda=0.82, p=0.013). Eigenvector 3 maximally discriminated the subject groups, and in a “leave one out” discriminant function analysis using these eigenvectors, 26 (48.2%) of the schizophrenia subjects and 40 (57.1%) of the comparison subjects were correctly classified. When total globus pallidus volume was included in the analysis, 31 (57.4%) of the schizophrenia subjects and 42 (60.0%) of the comparison subjects were correctly classified. Visual representations of the group difference in globus pallidus shape is shown in FIGURE 2C.

3.6 Typical and Atypical Antipsychotic Drug Effects

TABLE 3 shows the mean volumes of the basal ganglia structures in the schizophrenia subjects exposed to typical and atypical antipsychotic drugs. A one-way ANOVA comparing the two treatment groups did not show any significant volume differences for any basal ganglia structure on either side. Comparing schizophrenia subjects exposed to atypical antipsychotic drugs (N=35) to comparison subjects (N=70) also did not reveal significant volume differences for any basal ganglia structure. Similar comparisons were not carried out with subjects on typical antipsychotics, due to the low number of subjects in this group (N=9).

3.7 Group Discrimination Using Combined Basal Ganglia Shape Information

Visual representation of the group difference in the shape of the entire basal ganglia complex is displayed in FIGURE 2D. However, use of the combination of eigenvectors that maximally discriminated the groups for each of the basal ganglia in which a significant group effect was found did not improve the proportion of subjects correctly classified, over and above the proportion of subjects correctly classified when each of the structures were analyzed separately.

3.8 Cognitive and Clinical Relationships

Correlation analysis of cognitive factors showed a significant relationship between scores on cognitive factor 3 and both total caudate (r=0.41, p=0.02) and total putamen volumes (r=0.35, p<0.05) in schizophrenic subjects (p values uncorrected for multiple comparisons). When left and right volumes were investigated separately to determine relationship with this cognitive factor, significant correlations were found between cognitive factor 3 and caudate volumes bilaterally (right, r=0.47, p=0.005; left, r=0.39, p=0.05) and the right putamen volume (r=0.38, p=0.03). There were no significant relationships observed between any cognitive factor and the shape discriminant canonical variate for any structure.

When the relationships between basal ganglia volume and shape variables and SAPS and SANS scores were investigated, significant correlations were found between smaller total nucleus accumbens volume and increased total SAPS scores (r= −0.29, p=0.037). Separate analysis of the left and right volumes showed a correlation between the right nucleus accumbens volume and the total SAPS scores (r = −0.31, p=0.02), and with its delusion component (r=0.29, p=0.03).

There were no significant correlations between volume or shape variables and extrapyramidal motor symptoms. There were no significant differences in basal ganglia volume or shape between schizophrenic subjects treated recently with typical antipsychotic drugs and those treated recently with atypical antipsychotic drugs. The duration of illness in the schizophrenic subjects also did not correlate significantly with and volume or shape measure.

4. DISCUSSION

Our results suggest that the volumes of some basal ganglia structures (i.e., caudate, putamen and perhaps also the globus pallidus) are abnormally large relative to total cerebral volume in schizophrenia subjects. Further, these differences in relative volumes were associated with significant differences in the shape of these structures. This finding suggests that differences in basal ganglia volumes are not due to uniform changes throughout the structure. Rather, the shape changes we observed suggest that specific subregions within these complex nuclei are altered in individuals with schizophrenia.

Most investigators report that treatment with antipsychotic drugs (Gur et al, 1998; Andersson et al, 2002) is the cause of basal ganglia volume enlargement in schizophrenia. It has also been suggested that basal ganglia structures may have been spared from a pathological process that affected other structures of the cerebrum, especially the cerebral cortex (Swayze et al, 1992). Regarding the effects of antipsychotic drugs on the basal ganglia, typical antipsychotic drugs have been most often associated with volume increases. In contrast, second generation (atypical) neuroleptics have been described as devoid of these effects (Lang et al, 2004; Andersson et al, 2002). Also, treatment-naïve schizophrenia patients have been reported to have normal (Gur et al, 1998) or even decreased basal ganglia volumes (Corson et al, 1999; Keshavan et al, 1998). In our study, there were no differences in basal ganglia volumes between schizophrenic subjects who were receiving typical versus atypical antipsychotic drugs. However, we were only able to make this comparison based only on the type of drugs used at the time of assessment. The influence of other antipsychotic drugs used in the past cannot be excluded. Furthermore, there were relatively few subjects who were receiving typical drugs (n=9) as compared to atypical drugs (n=35), which limits statistical power to examine the differential effects of different types of antipsychotic medication. Therefore, with our sample of subjects, we could not adequately address the question of differential antipsychotic drug effects on basal ganglia volume.

Another limitation of our study was that we were not able to completely exclude the possible confounding effects of prior use of other medications or recreational substances. Varying degrees of exposure to substances over a lifetime may have affected the structure of the basal ganglia in both groups. Nonetheless, the structural differences that we observed between schizophrenic and comparison subjects were not altered after controlling for the presence of a lifetime history of substance abuse or dependence.

The basis of the observed shape differences between groups appeared to be uneven changes in the surface of individual basal ganglia, perhaps reflecting localized changes in substructure volumes. In the case of the caudate nucleus, there was a suggestion of localized volume loss in the anterior pole and anterior deflection of the tail. The anatomical distinctions observed are notable, in that the anterior pole of the caudate has reciprocal connections to prefrontal and limbic cortices (Parent & Hazrati, 1995; Lehericy et al, 2004; DeLong, 2000). In the putamen, more irregular shape changes were observed; however, it is notable that the anterior-lateral regions of the putamen with prominent connections to non-motor cortical areas (Lehericy et al, 2004) were affected. As discussed above, the structures of the basal ganglia may be affected differently by the presence of the disease state and by treatment (typical versus atypical drugs); thus, it may not be surprising that we observed a complex pattern of basal ganglia shape changes. Neuronal loss within the basal ganglia in schizophrenia has not been reported, and so is unlikely to be the basis for the observed changes in shape. Rather, it seems more plausible to suggest that basal ganglia shape changes could result from changes in the position of underlying neurons and their processes. Also, the ventral deflection of the tail of the caudate could be secondary to structural changes in the underlying thalamus (Csernansky et al, 2004).

In considering the patterns of shape change observed, it is also interesting to consider the patterns of neurotransmitter receptors that are differentially distributed in the basal ganglia. Opposite directional gradients have been reported for dopamine receptors in the basal ganglia, with a greater density of D1 receptors ventrally, and a greater density of D2/3 receptors dorso-posteriorly (Rosa-Neto et al, 2004; Hall et al, 1994). Therefore, antipsychotic drugs that block D2/3 receptors, might be expected to have disproportionate effects on dorso-posterior regions of the basal ganglia.

In our study, larger volumes of the caudate and putamen were correlated with greater performance on the third cognitive factor in schizophrenic subjects. Inspection of the cognitive measures loading on this factor, particularly performance on the continuous performance test, suggests that this dimension may be related to attention/vigilance (Nuechterlein et al, 2004). It is notable, however, that the WAIS Picture Completion task which also loaded highly on this factor, has not been linked to the attention/vigilance dimension in previous reports, but to reasoning and problem solving (Nuechterlein et al, 2004). Attention tasks are well known to be severely impaired in schizophrenia (Suwa et al, 2004). In particular, right hemispheric involvement has been reported in attention and vigilance (Pardo et al, 1991), and in our study, correlations between structural volumes and the attention cognitive factor were more substantial for the right-sided structures. We would caution that these findings are not conclusive in establishing a relationship between basal ganglia structural measures and cognitive function, especially considering that changes in such structures could have resulted from confounding from antipsychotic treatment. Furthermore, the cognitive battery used in this study was not specifically designed to measure basal ganglia functions, and overall, we found little evidence for a relationship between the structural measures and cognitive function. Future research using more focused cognitive batteries is needed to help elucidate whether there is a systematic relationship between either volume or shape changes in the basal ganglia in schizophrenia and specific aspects of cognitive function that may be supported by these structures (e.g., set switching).

In our exploratory analysis of correlations between structural and clinical measures, the right nucleus accumbens volume was inversely correlated with the severity of positive symptoms (and delusions alone). While the nucleus accumbens has traditionally been associated with reward, pleasure and addiction, an important role for this structure in the pathophysiology of schizophrenia has been suggested (Gray, 1998; Grace, 2000). Of course, our finding was the result of an exploratory analysis (uncorrected for multiple comparisons), and the precision of mapping the nucleus accumbens was not as strong as the other basal ganglia structures in this study. Thus, this observation must be considered highly preliminary.

As mentioned above, a limitation of our study was our inability to assess the impact of antipsychotic drug treatment other than the most recent on the neuroanatomical measures collected in our study. To determine the relationship between alterations of the basal ganglia in subjects with schizophrenia, the disease process and treatment, siblings of schizophrenia patients that later become ill or treatment naïve patients would need to be studied. The elucidation of abnormalities in neuroanatomical volume and shape in schizophrenia using MR imaging and computational anatomy may one day be useful for improving clinical diagnosis and selecting treatment. However, before such measures can be used for clinical purposes, the causes and consequences of such structural abnormalities must be better understood.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

5. References

1. Andersson C, Hamer RM, Lawler CP, Mailman RB, Lieberman JA. Striatal volume changes in the rat following long-term administration of typical and atypical antipsychotic drugs. Neuropsychopharmacology. 2002 Aug;27(2):143–51. [PubMed]
2. Andreasen NC, Arndt S, Alliger R, Miller D, Flaum M. Symptoms of schizophrenia: methods, meanings and mechanisms. Arch Gen Psychiatry. 1995;52:341–351. [PubMed]
3. Andreasen NC, Olsen S. Negative v positive schizophrenia. Arch Gen Psychiatry. 1982;39:789–94. [PubMed]
4. Corson PW, Nopoulos P, Andreasen NC, Heckel D, Arndt S. Caudate size in first-episode neuroleptic-naive schizophrenic patients measured using an artificial neural network. Biol Psychiatry. 1999 Sep 1;46(5):712–20. [PubMed]
5. Csernansky JG, Schindler MK, Splinter NR, Wang L, Gado M, Selemon LD, Rastogi-Cruz D, Posener JA, Thompson PA, Miller MI. Abnormalities of thalamic volume and shape in schizophrenia. Am J Psychiatry. 2004;161:896–902. [PubMed]
6. Csernansky JG, Wang L, Jones D, Rastogi-Cruz D, Posener JA, Hedebrand G, Miller JP, Miller MI. Hippocampal deformities in schizophrenia characterized by high dimensional brain mapping. Am J Psychiatry. 2002;159:2000–2006. [PubMed]
7. DeLong MR. The Basal Ganglia. In: Kandel ER, Schwartz JH, Jessell TM, editors. Principles of Neural Science. McGraw Hill; 2000. pp. 853–867.
8. First MB, Spitzer RL, Gibbon M, Williams JBW. Biometrics Research. New York State Psychiatric Institute; New York: 1995. Structured Clinical Interview for DSM-IV Axis I Disorders, Patient Edition (SCID-P), version 2.
9. Gharabawi GM, Bossie CA, Lasser RA, Turkoz I, Rodriguez S, Chouinard G. Abnormal Involuntary Movement Scale (AIMS) and Extrapyramidal Symptom Rating Scale (ESRS): cross-scale comparison in assessing tardive dyskinesia. Schizophr Res. 2005 Sep 15;77(23):119–28. [PubMed]
10. Grace AA. Gating of information flow within the limbic system and the pathophysiology of schizophrenia. Brain Res Brain Res Rev. 2000 Mar;31(23):330–41. [PubMed]
11. Gray JA. Integrating schizophrenia. Schizophr Bull. 1998;24(2):249–66. [PubMed]
12. Gunduz H, Wu H, Ashtari M, Bogerts B, Crandall D, Robinson DG, Alvir J, Lieberman J, Kane J, Bilder R. Basal ganglia volumes in first-episode schizophrenia and healthy comparison subjects. Biol Psychiatry. 2002 May 15;51(10):801–8. [PubMed]
13. Gur RE, Maany V, Mozley PD, Swanson C, Bilker W, Gur RC. Subcortical MRI volumes in neuroleptic-naive and treated patients with schizophrenia. Am J Psychiatry. 1998 Dec;155(12):1711–7. [PubMed]
14. Hall H, Sedvall G, Magnusson O, Kopp J, Halldin C, Fardel L. Distribution of D1- and D2-dopamine receptors, and dopamine and its metabolites in the human brain. Neuropsychopharmacology. 1994;11:245–256. [PubMed]
15. Haller JW, Banerjee A, Christensen GE, Gado M, Joshi S, Miller MI, Sheline YI, Vannier MW, Vannier MW, Csernansky JG. Three-dimensional hippocampal MR morphometry by high-dimensional transformation of a neuroanatomic atlas. Radiology. 1997;202:504–510. [PubMed]
16. Joel D. Open interconnected model of basal ganglia-thalamocortical circuitry and its relevance to the clinical syndrome of Huntington's disease. Mov Disord. 2001 May;16(3):407–23. [PubMed]
17. Joshi SC, Miller MI, Grenander U. On the geometry and shape of brain sub-manifolds. Int J Pattern Recognit Artif Intell. 1997;11:1317–1343.
18. Keshavan MS, Rosenberg D, Sweeney JA, Pettegrew JW. Decreased caudate volume in neuroleptic-naive psychotic patients. Am J Psychiatry. 1998 Jun;155(6):774–8. [PubMed]
19. Lang DJ, Kopala LC, Vandorpe RA, Rui Q, Smith GN, Goghari VM, Lapointe JS, Honer WG. Reduced basal ganglia volumes after switching to olanzapine in chronically treated patients with schizophrenia. Am J Psychiatry. 2004 Oct;161(10):1829–36. [PubMed]
20. Lehericy S, Ducros M, Van de Moortele PF, Francois C, Thivard L, Poupon C, Swindale N, Ugurbil K, Kim DS. Diffusion tensor fiber tracking shows distinct corticostriatal circuits in humans. Ann Neurol. 2004 Apr;55(4):522–9. [PubMed]
21. Levy R, Friedman HR, Davachi L, Goldman Rakic PS. Differential activation of the caudate nucleus in primates performing spatial and nonspatial working Memory tasks. J Neurosci. 1997;17:3870–3882. [PubMed]
22. Manoach DS, Gollub RL, Benson ES, Searly MM, Goff DC, Halpern E, Saper CB, Rauch SL. Schizophrenic subjects show aberrant fMRI activation of dorsolateral prefrontal cortex and basal ganglia during working memory performance. Biol Psychiatry. 2000 Jul 15;48(2):99–109. [PubMed]
23. McCarley RW, Wible CG, Frumin M, Hirayasu Y, Levitt JJ, Fischer IA, Shenton ME. MRI anatomy of schizophrenia. Biol Psychiatry. 1999 May 1;45(9):1099–119. [PMC free article] [PubMed]
24. Mendez MF, Adams NL, Lewandowski KS. Neurobehavioral changes associated with caudate lesions. Neurology. 1989;39(3):349–354. [PubMed]
25. Menon V, Anagnoson RT, Glover GH, Pfefferbaum A. Functional magnetic resonance imaging evidence for disrupted basal ganglia function in schizophrenia. Am J Psychiatry. 2001 Apr;158(4):646–9. [PubMed]
26. Miller M, Banerjee A, Christensen G, Joshi S, Khaneja N, Grenander U, Matejic L. Statistical methods in computational anatomy. Stat Methods Med Res. 1997 Sep;6(3):267–99. [PubMed]
27. Nuechterlein KH, Barch DM, Gold JM, Goldberg TE, Green MF, Heaton RK. Identification of separable cognitive factors in schizophrenia. Schizophrenia Research. 2004;72:29–39. [PubMed]
28. Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neurospcyhologia. 1971;9:97–113. [PubMed]
29. Pardo JV, Fox PT, Raichle ME. Localization of a human system for sustained attention by positron emission tomography. Nature. 1991 Jan 3;349(6304):61–4. [PubMed]
30. Parent A, Hazrati LN. Functional anatomy of the basal ganglia: I. The cortico-basal gangliatalamo-cortical loop. Brain Res. Brain Res. Rev. 1995;20:91–127. [PubMed]
31. Rosa-Neto P, Doudet DJ, Cumming P. Gradients of dopamine D1- and D2/3-binding sites in the basal ganglia of pig and monkey measured by PET. Neuroimage. 2004;22:1076–1083. [PubMed]
32. Shihabuddin L, Buchsbaum MS, Hazlett EA, Haznedar MM, Harvey PD, Newman A, Schnur DB, Spiegel-Cohen J, Wei T, Machac J, Knesaurek K, Vallabhajosula S, Biren MA, Ciaravolo TM, Luu-Hsia C. Dorsal striatal size, shape, and metabolic rate in never-medicated and previously medicated schizophrenics performing a verbal learning task. Arch Gen Psychiatry. 1998 Mar;5593:235–43. [PubMed]
33. Staal WG, Hulshoff Pol HE, Schnack HG, Hoogendoorn ML, Jellema K, Kahn RS. Structural brain abnormalities in patients with schizophrenia and their healthy siblings. Am J Psychiatry. 2000 Mar;157(3):416–21. [PubMed]
34. Suwa H, Matsushima E, Ohta K, Mori K. Attention disorders in schizophrenia. Psychiatry Clin Neurosci. 2004 Jun;58(3):249–56. [PubMed]
35. Swayze VW, Andreasen NC, Alliger RJ, Yuh WT, Ehrhardt JC. Subcortical and temporal structures in affective disorder and schizphrenia: a magnetic resonance imaging study. Biol Psychiatry. 1992;31:221–240. [PubMed]
36. Vankatesan R, Haacke EM. Role of high resolution in magnetic resonance (MR) imaging: applications of MR angiography, intracranial T1-weighted imaging, and image interpolation. Int J Imaging Systems and Technology. 1997;8:529–543.
37. Wang L, Joshi SC, Miller MI, Grenander U, Csernansky JG. Statistical analysis of hippocampal asymmetry. Neuroimage. 2001;14:531–545. [PubMed]
38. Zgaljardic DJ, Borod JC, Foldi NS, Mattis P. A review of the cognitive and behavioral sequelae of Parkinson's disease: relationship to frontostriatal circuitry. Cogn Behav Neurol. 2003 Dec;16(4):193–210. [PubMed]