Corpus Callosum Abnormalities and Their

Association with Psychotic Symptoms in Patients

with Schizophrenia
Citation
Whitford, Thomas J., Marek Kubicki, Jason S. Schneiderman, Lauren J. O’Donnell, Rebecca
King, Jorge L. Alvarado, Usman Khan, et al. 2010. Corpus Callosum Abnormalities and Their
Association with Psychotic Symptoms in Patients with Schizophrenia. Biological Psychiatry 68,
no. 1: 70–77. doi:10.1016/j.biopsych.2010.03.025.

Published Version
doi:10.1016/j.biopsych.2010.03.025

Permanent link
http://nrs.harvard.edu/urn-3:HUL.InstRepos:28548995

Terms of Use
This article was downloaded from Harvard University’s DASH repository, and is made available
under the terms and conditions applicable to Other Posted Material, as set forth at http://
nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA

Share Your Story

The Harvard community has made this article openly available.
Please share how this access benefits you. Submit a story .

Accessibility
 NIH Public Access
Author Manuscript
Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.
Published in final edited form as:
NIH-PA Author Manuscript

Biol Psychiatry. 2010 July 1; 68(1): 70–77. doi:10.1016/j.biopsych.2010.03.025.

Corpus Callosum Abnormalities and their Association with

Psychotic Symptoms in Patients with Schizophrenia

Thomas J. Whitford, Ph.D.1,2,*, Marek Kubicki, M.D., Ph.D.1,3, Jason S. Schneiderman, Ph.D.
1, Lauren J. O'Donnell, Ph.D.4,5, Rebecca King, B.Sc.1, Jorge L. Alvarado, B.Sc.1, Usman
Khan, B.Sc.1, Douglas Markant, B.Sc.1, Paul G. Nestor, Ph.D.3,6, Margaret Niznikiewicz,
Ph.D.3, Robert W. McCarley, M.D., PhD.3, Carl-Fredrik Westin, Ph.D.4, and Martha E. Shenton,
Ph.D.1,3
1Psychiatry Neuroimaging Laboratory, Department of Psychiatry, Brigham and Women's Hospital,

Harvard Medical School, Boston, MA, USA

2Melbourne Neuropsychiatry Centre, Department of Psychiatry, University of Melbourne and
Melbourne Heath, Melbourne, VIC, Australia
NIH-PA Author Manuscript

3Clinical Neuroscience Division, Laboratory of Neuroscience, Department of Psychiatry, Veterans

Affairs (VA) Boston Healthcare System, Harvard Medical School Brockton, MA, USA
4Laboratoryof Mathematics in Imaging, Brigham and Women's Hospital, Harvard Medical School,
Boston, MA, USA
5Golby Laboratory, Department of Neurosurgery, Brigham and Women's Hospital, Harvard Medical
School, Boston, MA, USA
6College of Liberal Arts, University of Massachusetts - Boston, Boston, MA, USA

Abstract
Background—Whilst the neuroanatomical underpinnings of the functional brain disconnectivity
observed in patients with schizophrenia remain elusive, white matter fiber bundles of the brain are
a likely candidate given that they represent the infrastructure for long-distance neural communication.
Methods—This study investigated for diffusion abnormalities in 19 patients with chronic
schizophrenia (SZ), relative to 19 matched controls, across tractography-defined segments of the
Corpus Callosum. Diffusion-weighted images were acquired with 51 non-collinear gradients on a
NIH-PA Author Manuscript

3T scanner (1.7mm isotropic voxels). The Corpus Callosum was extracted by means of whole-brain
tractography and automated fiber-clustering, and was parcellated into six segments on the basis of
fiber trajectories. The diffusion indices of Fractional Anisotropy (FA) and Mode were calculated for
each segment.
Results—Relative to the healthy controls, the SZ patients exhibited Mode increases in the Parietal
fibers, suggesting a relative absence of crossing fibers. SZ patients also exhibited FA reductions in
the Frontal fibers, which were underpinned by increased in Radial Diffusivity, consistent with myelin
abnormalities. Significant correlations were observed between patients' degree of Reality Distortion

*
Corresponding Author: Thomas J. Whitford, Psychiatry Neuroimaging Laboratory, Brigham and Women's Hospital, Harvard Medical
School, 1249 Boylston St, Boston, MA, 02215, Phone: +1 617 525 1059, Fax: +1 617 525 6150, whitford@bwh.harvard.edu.
Financial Disclosures: None of the authors report having any biomedical financial interests or potential conflicts of interest
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.
 Whitford et al. Page 2

and their FA and Radial Diffusivity, such that the most severely psychotic patients were the least
abnormal in terms of their Frontal fiber diffusivity.
Conclusions—The SZ patients exhibited a variety of diffusion abnormalities in the Corpus
NIH-PA Author Manuscript

Callosum, which were related to the severity of their psychotic symptoms. To the extent that diffusion
abnormalities influence axonal transmission velocities, these results provide support for those
theories that emphasize neural timing abnormalities in the etiology of schizophrenia.

Keywords
Schizophrenia; Corpus Callosum; Diffusion Tensor Imaging; Tractography; Fractional Anisotropy;
Mode

Introduction
The unifying tenet of the increasingly popular ‘disconnectivity’ theories of schizophrenia (SZ)
(1-3) is that the disorder is ultimately caused by abnormal interactions between pathological
brain regions, as opposed to regional neuropathology per se. In addition to providing a prima
facie account of the cognitive disorganization characteristic of the disease, support for the
‘disconnectivity’ theories has been provided by a number of electrophysiological and
functional imaging studies that have reported abnormalities in the degree of correlated activity
NIH-PA Author Manuscript

between spatially disparate brain regions in patients with schizophrenia (4-6). The
neuroanatomical underpinnings of this aberrant functional connectivity are as yet unknown
and represent a point of divergent focus (but not, it should be emphasized, contradiction)
between the various ‘disconnectivity’ theories. Whilst some models have emphasized the role
of aberrant synaptic plasticity in the etiology of the disorder (2), others have focused on the
role of abnormalities in the physical infrastructure for long-distance neural communication,
i.e., white matter (WM) (3). With respect to the latter, white matter is primarily constituted of
the phospholipid processes (known as myelin) of a specialized class of neuroglia called
oligodendrocytes (7). Myelin ensheaths axons in the nervous system, electrically insulating
the axon membrane and increasing the conduction velocity of action potentials. Myelinated
axons with similar destinations fasciculate into fiber bundles, which constitute the primary
infrastructure for communication between spatially disparate brain regions. It has previously
been suggested that abnormalities in WM fiber bundles could represent the neuroanatomical
bases for many of the observed abnormalities in functional connectivity in patients with
schizophrenia (3).

The Corpus Callosum (CC) is the largest WM fiber bundle in the brain and connects
homologous regions of the two cerebral hemispheres. The CC has been implicated in SZ by
NIH-PA Author Manuscript

those who emphasize the role of abnormal hemispheric specialization and abnormal
interhemispheric communication in the etiology of the disease (8). Whilst conventional MRI
has provided only equivocal evidence for CC abnormalities in patients with schizophrenia (see
(9) for a review), a more consistent picture of structural abnormality has been provided by
those studies that have employed the more sensitive (at least with respect to WM pathology)
modality of Diffusion Tensor Imaging (DTI) (see (10) for a review). However, whilst these
previous studies have provided novel and valuable information regarding the location and
extent of SZ-specific structural abnormalities in the CC, they have also exhibited some
methodological limitations.

Firstly, the majority of previous studies have employed a somewhat arbitrary protocol for
segmenting the CC, often based on geometric rather than anatomical boundaries, which may
be insufficiently sensitive to detect subtle, regionally-specific diffusion abnormalities.
Secondly, most previous DTI studies have investigated for group-wise differences on only a

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 3

single diffusivity metric - namely Fractional Anisotropy (FA), which is a measure of the
asphericity of water diffusion. However, as noted by Hasan (11), FA is only able to provide a
limited picture as to the three-dimensional shape of the observed diffusion. FA cannot, for
NIH-PA Author Manuscript

example, distinguish between prolate (i.e., cigar shaped) and oblate (i.e., pancake shaped)
diffusion, despite these presumably having markedly different microstructural underpinnings.
The present study aimed to address these methodological limitations in two ways. Firstly, this
study employed a previously validated analysis method (12) to distinguish between fiber
bundles on the basis of the cortical regions to which they projected, thus avoiding the need for
an inflexible geometric schema. Secondly, in addition the stalwart metric of FA, this study also
investigated a second, orthogonal diffusion index: Mode (13). Mode is a measure of the
prolateness / oblateness of a diffusion ellipsoid, thus providing additional insight into the three-
dimensional shape (and hence microstructural underpinnings) of observed diffusion, relative
to that provided by FA alone.

The second aim of this study was to investigate the basis for a consistently reported, if somewhat
paradoxical finding of a positive correlation between SZ patients' FA and the severity of their
hallucinations and delusions (14-17). Although the explanation behind this curious relationship
has not yet been established, we have recently suggested that while psychotic symptoms may
arise when the brain attempts to integrate mildly temporally dysmetric activity in spatially
disparate cerebral regions, more severe temporal discoordinations might not be able to be
integrated and thus might be incapable of giving rise to psychotic symptoms (18). Given the
NIH-PA Author Manuscript

role that WM (and especially myelin) is known to play in modulating the speed of neural
transmission, we predicted that severely psychotic SZ patients would show less extensive
myelin abnormalities relative to SZ patients with less severe psychotic symptoms.

In summary, the two aims of the present study were: 1) to provide a comprehensive description
of the shape and extent of water diffusion across tractography-defined subregions of the CC,
with the aim of inferring the microstructural underpinnings of any observed diffusion
abnormalities in SZ patients, and 2) to investigate the relationship between psychotic symptom
severity and WM integrity (as assessed with FA and Radial Diffusivity, the latter of which has
been proposed as a putative measure of myelin integrity (19)) in SZ patients.

Methods and Materials

Participants
Nineteen male patients with chronic schizophrenia were recruited from out-patient, in-patient,
day treatment and foster care programs at the VA Boston Healthcare System, Brockton, MA.
Diagnosis of schizophrenia was made in accordance with DSM-IV criteria on the basis of the
Structured Clinical Interview for DSM-IV (conducted by a clinically and research-trained
NIH-PA Author Manuscript

psychologist (PN)), and a review of the medical record. Nineteen male healthy control subjects
were recruited from the general community. The control subjects were group matched to the
patients on age, handedness, parental socioeconomic status (20), and estimated premorbid IQ,
as assessed by performance on the Reading scale of Wide Range Achievement Test (WRAT-3)
(21). More than 90% of the participants in this study were new (10% were previously tested
on the 1.5T magnet), and none overlapped with the samples described in previous studies by
our group (e.g., 22,23,24).

Exclusion criteria for all subjects were left-handedness, a history of electroconvulsive shock
therapy, a history of neurological illness including epilepsy, a lifetime history of substance
dependence or a history of substance abuse within the past 5 years, a history of steroid use,
and estimated premorbid IQ below 75. Furthermore, control subjects were screened for the
presence of an Axis-I disorder using the SCID-Non-Patient edition (25), and were also excluded
if they reported having first-degree relative with an Axis I disorder.

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 4

The study was approved by the VA Boston Healthcare System, the Harvard Medical School
Internal Review Board Committee, and the Brigham and Women's Hospital Human Subjects
Committee. After a detailed description of the study, each subject gave written informed
NIH-PA Author Manuscript

consent to participate. The demographic details for the patients and controls are summarized
in Table 1.

Image Acquisition
Diffusion data were collected on 3 Tesla GE Echospeed system (General Electric Medical
Systems, Milwaukee, WI). Diffusion-weighted images (DWIs) were acquired using an echo
planar imaging sequence, and a double echo option to reduce eddy-current related distortions.
To reduce impact of EPI spatial distortion, an 8 Channel coil and ASSET (Array Spatial
Sensitivity Encoding techniques, GE) with a SENSE-factor (speed-up) of 2 was used. Eighty-
five axial slices parallel to the AC-PC line covering whole brain were acquired, in 51 diffusion
directions with b=900. In addition, 8 baseline scans with b=0 were also acquired. The scan
parameters were as follows: TR 17000 ms, TE 78 ms, FOV 24 cm, 144×144 matrix, 1.7 mm
slice thickness, producing isotropic 1.7×1.7×1.7mm voxels. The total scanning time was 17
minutes.

Whole-Brain Tractography
The methods used in this study have been described in detail elsewhere (12). For each subject
NIH-PA Author Manuscript

and each image element (voxel), diffusion tensors were estimated from the 59 DWIs. The
resultant DTIs were first edited to remove extra-cerebral voxels via a semi-automated
generation of a brain ‘mask’. Deterministic (streamline) tractography was then performed via
a Runge-Kutta second order protocol with a fixed step size of 0.5mm. Seed-points were placed
at every point for which Westin's Linear Anisotropy measure (CL) was greater than 0.3 (the
seeding criterion). Tractography proceeded from each seed point in 0.5mm steps, and followed
the principal direction of the diffusion ellipsoid. Tracking was stopped when CL fell below
0.15. The seeding and stopping thresholds were based on the CL rather than on FA in order to
avoid the circularity inherent in using a dependent variable to define the trajectories of the fiber
tracts being measured. A length threshold was also employed whereby fibers were excluded
if they were shorter than 20mm.

Fiber Clustering
The fiber-clustering procedure has been described in detail previously (26). The whole-brain
tractography procedure (see Fig.1a) generated somewhere in the vicinity of twenty thousand
fibers per subject. The purpose of fiber clustering was to group fibers with similar shapes and
spatial positions into clusters. Firstly, Fractional Anisotropy (FA) maps calculated for each
subject were mapped into a common coordinate system using a congealing registration (27),
NIH-PA Author Manuscript

and the parameters for this registration were applied to each subjects' 3-dimensional set of fiber
trajectories. Secondly, each fiber was compared with every other fiber in the trajectory set, and
the average distance between pairs of nearest points on the fibers was calculated. This mean
fiber-distance (Dij) was then converted into a ‘similarity value’ (Wij) via the formula
, where θ = 60mm. The role of θ was to set the distance over which fibers could be
considered similar (26). The ‘similarity value’ of each fiber to every other fiber was entered
into an ‘affinity matrix’ (W), and the top 15 eigenvectors of W were used to calculate the most
important shape similarity information for each fiber. The clustering algorithm used k-way
normalized cuts, which produces clusters with high within-cluster similarity and low between-
cluster similarity (28). Fiber clustering was performed on the amalgamated fiber tracts of all
subjects in the study (as opposed to clustering each subject separately), as this enabled the
identification of homologous fiber-clusters (FCs) between subjects. Tractography and fiber

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 5

clustering were performed using Matlab 7.0 (www.mathworks.com) and 3D-Slicer

(www.slicer.org), which is freely available to the research community.
NIH-PA Author Manuscript

Extraction and Segmentation of the Corpus Callosum

The output of the clustering procedure was 400 FCs, each consisting of a spatially and
morphologically similar subset of the fibers generated from the whole-brain tractography (see
Fig.1b). Of these 400 FCs, 75 were judged to constitute the Corpus Callosum, based on the
consensus of two independent raters (TW and JS) on three randomly-selected participants
(inter-rater reliability = 0.901). These 75 Corpus FCs were automatically extracted from all
participants (on the basis of their unique cluster labels (29)), and the validity of these extracted
FCs confirmed by a third rater (MK). These CC FCs were then subdivided into six segments
on the basis of the cortical regions to which they projected (see Fig.1c). Using these cluster
labels, the FCs constituting the CC were automatically extracted for each subject, and
subsequently subdivided into six segments on the basis of the cortical regions to which they
projected. The six segments, which are described and illustrated in Fig.1d, were: Frontal fibers
(CC1), Premotor fibers (CC2), Sensorimotor fibers (CC3), Parietal fibers (CC4), Occipital
fibers (CC5) and Temporal fibers (CC6).

Diffusion Indices
Fractional Anisotropy (FA) and Mode were calculated at every voxel for each subject.
NIH-PA Author Manuscript

Mode is an index of the 3-dimensional shape of diffusion - specifically whether it is better

described by an oblate (shaped like a pancake; Mode ≈ -1) or prolate (shaped like a cigar; Mode
≈ 1) ellipsoid.

Fractional Anisotropy (FA) is an index of the asphericity of diffusion, and is calculated as:

Mean values of the diffusivity indices were calculated for each CC segment, for each subject.
NIH-PA Author Manuscript

This was done by averaging the diffusivity values (i.e., FA and Mode) of all voxels through
which passed any of the FCs that constituted a given CC segment.

In order to help further infer the microstructural underpinnings of any abnormalities in FA,
two additional diffusion indices were calculated for only those segments for which group-wise
differences in FA were observed. These two indices – namely Radial Diffusivity and Axial
Diffusivity – have been proposed as being sensitive to abnormalities in myelination (19,30)
and axon integrity (31), respectively. Radial Diffusivity is a measure of the extent of diffusion

perpendicular to the principal axis of the diffusion ellipsoid, and is calculated as: . Axial
Diffusivity is a measure of the absolute extent of diffusion along the principal axis of the
diffusion ellipsoid, and is calculated as: λ1

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 6

Statistical Analysis
The statistical analysis was performed using SPSS v11 (www.spss.com). Given the confounds
associated with using a Repeated-Measures design in the context of severe violations of
NIH-PA Author Manuscript

sphericity (32), a Multivariate Analysis of Variance (MANOVA) was used to identify between-
group differences in FA and Mode. To control for multiple comparisons, Fisher's Protected
Test was employed (33). If (and only if) the MANOVA identified an overall between-group
difference in a diffusion index were post hoc contrasts (Fisher's LSD) used to identify the CC
segments responsible for the omnibus effect.

Pearson's partial correlations (controlling for CPZ-equivalent medication dosage) were used
to investigate the relationship between patients' FA and Radial Diffusivity and their Reality
Distortion symptom score. As per Liddle (34), Reality Distortion was calculated for each
patient as the sum of their scores on the Hallucination and Delusion subscales of the Positive
and Negative Symptom Scale (PANSS). Correlations between FA / Radial Diffusivity and
patients' scores on the PANSS-Negative and PANSS-General subscales were also investigated.
In order to limit the number of statistical comparisons, these correlations were only investigated
for those CC segments which were identified as being abnormal in the patient group.

Results
Figure 2 shows the scatterplots for FA (Fig.2a) and Mode (Fig.2b) across the 6 CC segments,
NIH-PA Author Manuscript

for the SZ (red) and Control (blue) groups.

Multivariate ANOVA revealed significant between-group differences in FA (F(6,31)=3.537,

p=0.009). Post hoc analysis revealed that the FA of CC1 (i.e., the Frontal fibers) was
significantly decreased in the SZ patients relative to controls (t(1,36)=2.282 p=0.029).
Additional analyses were performed in order to determine whether the FA reductions in CC1
were underpinned by changes in Axial or Radial Diffusivity. The SZ patients exhibited
significant increases in Radial Diffusivity in CC1 (t(1,36)=2.218, p=0.033), but no differences
in Axial Diffusivity (t(1,36)=0.113, p=0.911), relative to controls (Fig.3).

Significant between-group differences in Mode were also observed (F(6,31)=2.417, p=0.049)

(Fig.2b). Post hoc analysis revealed that Mode was significantly increased (indicating a more
cigar-shaped diffusion ellipsoid) in CC4 (i.e., the Parietal fibers) in the SZ patients relative to
controls (t(1,36)=2.291, p=0.028).

Significant positive correlations (controlling for patients' CPZ-equivalent medication dosage)

were observed between FA in CC1 and SZ patients' Reality Distortion scores (r(16)=.601,
p=0.008) (Fig.4a). Conversely, significant negative correlations were observed between Radial
NIH-PA Author Manuscript

Diffusivity in CC1 and SZ patients' Reality Distortion scores (r(16)=-.538, p=0.021) (Fig.4b).
The results of these correlations were similar (and remained statistically significant) when
CPZ-equivalent medication dosage was removed as a covariate.

There was a (non-significant) trend for a negative correlation between patients' FA in CC1 and
their total score on the PANSS-Negative subscale, controlling for medication dosage (r(16)=-.
341, p=0.166). Each of the individual PANSS-Negative items was found to be negatively
correlated with FA in CC1, although this only reached statistical significance for item N5
(difficulty in abstract thinking) (r(16)=-.477, p=0.046). No statistically significant correlations
were observed between patients' FA in CC1 and their scores on the PANSS-General scale or
subscales.

No significant correlations were observed between patients' CPZ-equivalent medication

dosage and their FA or Mode in any of the six CC segments (see Supplement 1). In contrast,

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 7

significant negative correlations were observed between age and FA across groups, consistent
with several previous studies (35-37) (see Supplement 1).
NIH-PA Author Manuscript

Discussion
The primary findings of this study were of FA and Radial Diffusivity abnormalities in the
Frontal CC fibers, and Mode abnormalities in the Parietal CC fibers in 19 SZ patients relative
to 19 matched healthy controls. Furthermore, the severity of patients' symptoms of Reality
Distortion (i.e., hallucinations and delusions) were found to be positively correlated with FA
and negatively correlated with Radial Diffusivity in the Frontal CC fibers. In light of the FA
reductions and Radial Diffusivity increases which were observed in the SZ patients in this CC
segment, these results suggest that the most severely psychotic patients were the least abnormal
in terms of their FA and Radial Diffusivity in these fibers.

Whilst FA reductions in the genu of the CC in SZ patients have been reported previously
(38,39), including in patients with first-episode SZ (40) (see (10,41) for reviews), their
microstructural underpinnings have not been well established. FA reductions have been found
to occur in response to axon death, myelin damage, damage to the axon membrane, and reduced
‘fiber coherence’ (i.e., more variable fiber trajectories within a WM bundle) (10,42). While
there is little evidence to suggest the presence of any substantial degree of abnormal neuron
death in SZ patients (e.g., see (43)), there is, in contrast, a growing body of evidence suggesting
NIH-PA Author Manuscript

that SZ patients exhibit abnormalities in both their myelin and in the integrity of their axon
membranes (44,45). Distinguishing between these two distinct neuropathologies was one
motivation behind the development of the diffusion indices Axial and Radial Diffusivity. In a
series of animal studies, Song (19) demonstrated that while damage to the myelin of the optic
nerve resulted in increased Radial (but unchanged Axial) Diffusivity, damage to the axonal
membrane of the optic nerve (but preservation of the myelin) resulted in reduced Axial but
unchanged Radial Diffusivity (31). The results of these studies support the claim that Radial
Diffusivity is a putative measure of myelin integrity while Axial Diffusivity is a putative
measure of axonal integrity. Hence the results of the present study in which the SZ patients
were observed to exhibit abnormally increased Radial Diffusivity but unchanged Axial
Diffusivity, relative to controls, in the Frontal CC fibers, suggests that the FA abnormalities
in these fibers were more likely underpinned by myelin abnormalities as opposed to damage
to the axon membrane.

If the observed diffusion abnormalities in the Frontal CC fibers were indeed the result of myelin
damage, then this might be expected to result in slowed impulse conduction (46). Furthermore,
it might also seem likely that those SZ patients with the most profound diffusion abnormalities
(and hence presumably the most severe conduction delays) would exhibit the most severe
NIH-PA Author Manuscript

psychopathology. Such a relationship, however, was not observed in the present study. On the
contrary, those SZ patients with the most subnormal levels of FA (and most supranormal levels
of Radial Diffusivity) exhibited the least severe symptoms of Reality Distortion. Far from
being an unprecedented finding, this seemingly paradoxical result of a positive correlation
between FA and psychotic symptom severity has been reported many times previously,
including in the CC (17,47), cingulum bundle (17), arcuate fasciculus (17), superior
longitudinal fasciculus (14,15) and inferior fronto-occipital fasciculus (16). It was also notable
that, in the present study, the most severely psychotic patients did not show abnormally high
levels of FA, but instead exhibited FA values lower than the healthy controls but higher than
the less-psychotic patients (see Fig.4). What is the explanation for this seemingly paradoxical
finding that the more floridly psychotic a SZ patient, the less severe their FA abnormalities?

To the extent that axonal conduction delays can be inferred from diffusivity abnormalities in
DTI, these results suggest that while some amount of neural desynchronization may be

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 8

necessary for the development of psychotic symptoms, too much desynchronization may in
fact preclude the development of highly systematized hallucinations and delusions such as
would score highly on a clinical rating scale such as the PANSS. This idea has some support
NIH-PA Author Manuscript

in the WM literature. For example, while psychotic symptoms have been commonly reported
in patients with the degenerative demyelinating disease metachromatic leukodystrophy (48),
they are more likely to occur in the early stages of the disease when myelin pathologies are
relatively minor compared to the late stages of the disease when severe demyelination is
apparent (49). If it is true that psychotic symptoms represent the brain's attempt to incorporate
disjointed neural activity into a coherent (albeit pathological) framework (50), then perhaps
this “pathological integration” is only possible up to a certain level of temporal
desynchronization. In other words, while mild asynchronies between the activities of spatially
discrete brain regions might give rise to psychotic symptoms (such as, for example, between
the different brain regions stimulated by a primary discharge and its corollaries (51)), severe
asynchronies (such as might be caused by severe WM damage) might not be incorporable into
a coherent phenomenological framework and thus not give rise to psychotic symptoms. Such
severe WM damage could instead result in something of a ‘cognitive shutdown’ which could
underlie the negative symptoms of SZ. This would explain the negative correlation between
FA and negative symptom severity that has been observed previously in SZ patients (16,47,
52), and for which there was a non-significant trend in the present study. Testing these
hypotheses could potentially provide a fruitful avenue for future research.
NIH-PA Author Manuscript

Our finding of abnormally increased Mode (i.e., abnormally prolate diffusion ellipsoids) in the
Parietal CC fibers in the SZ patients has not (to our knowledge) been reported previously. Mode
is a relatively recently-developed measure which provides independent and complementary
information to the stalwart index of FA (13). In terms of its physiological determinants, Mode
has been found to decrease in the presence of fiber crossings (13). Thus one explanation as to
why the SZ patients exhibited abnormally increased Mode but normal levels of FA in the
parietal CC fibers might be that the patients evince abnormalities in a fiber bundle adjacent to
these CC fibers (e.g., the cingulum bundle, as recently suggested by Mandah et al., (53)),
resulting in a reduction in the density of fiber crossings, and hence elevated Mode, in this region
of the CC. Our finding of diffusion abnormalities in the Parietal CC fibers is also consistent
with the results of several early-onset studies which have reported parietal lobe white matter
to be among the first cerebral areas to evince structural abnormalities in patients with
schizophrenia – abnormalities that have been argued to contribute to the emergence of
psychotic symptoms through their effects on frontal-parietal connectivity (54,55).

A limitation of the present study relates to the fact that all of the SZ patients were suffering
from chronic schizophrenia and had been at least intermittently (and in many cases chronically)
exposed to neuroleptic medications. This is relevant in light of evidence from primate studies
NIH-PA Author Manuscript

suggesting that exposure to both typical and atypical neuroleptics can influence brain structure
in and of itself (56). Notwithstanding the fact that a) patients' chlorpromazine-equivalent
medication dosages were statistically controlled for in the present study, and b) there were no
significant correlations between patients' CPZ-equivalent medication dosages and their FA,
Mode, or Radial Diffusivity in any of the 6 CC segments (see Supplement 1), replicating these
results in a population of first-episode, neuroleptic-naive SZ patients would strengthen
confidence into their validity, and would be a worthwhile aim for future research. A second
limitation relates to the fact that all the patients were male. Whilst this was advantageous in
the sense that it increased the homogeneity of the sample and thus the power, it obviously
limited the extent to which the results can be generalized to females.

In summary, the main findings of this study were of FA reductions and Radial Diffusivity
increases in Frontal CC fibers (consistent with dysmyelination) and Mode increases in Parietal
CC fibers (consistent with a relative absence of crossing fibers) in 19 patients with

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 9

schizophrenia relative to 19 matched healthy controls. Significant correlations were also

observed between patients' FA and Radial Diffusivity in the Frontal fibers and the severity of
their psychotic symptoms, such that patients with the most abnormal levels of FA and Radial
NIH-PA Author Manuscript

Diffusivity exhibited the least severe symptoms of Reality Distortion. This result, we suggest,
provides support for those theories which emphasize neural timing abnormalities in the etiology
of psychotic symptoms.

Supplementary Material
Refer to Web version on PubMed Central for supplementary material.

Acknowledgments
Thomas Whitford is supported by an Overseas-Based Biomedical Training Fellowship from the National Health and
Medical Research Council of Australia (NHMRC 520627), administered through the University of Melbourne. Marek
Kubicki is supported by grants from the National Institutes of Health (R03 MH068464-0; R01 MH 50747 to MES),
the Harvard Medical School (Milton Award), and the National Alliance for Research on Schizophrenia and Depression.
Jason Schneiderman is supported by a fellowship from the National Institutes of Health (T32 MH 016259). Martha
Shenton is supported by grants from the National Institutes of Health (K05 MH 070047 and R01 MH 50747), the
Department of Veterans Affairs (VA Merit Award, VA Research Enhancement Award Program and VA Schizophrenia
Research Center Grant), and the Boston Center for Intervention Development and Applied Research (CIDAR) funded
through a center grant mechanism (P50 MH 080272). Robert McCarley is supported by grants from the Department
of Veterans Affairs (VA Merit Award, VA Schizophrenia Research Center Grant), the National Institute of Mental
NIH-PA Author Manuscript

Health (MH 040799), and the Boston Center for Intervention Development and Applied Research (CIDAR, P50 MH
080272). Lauren O'Donnell is supported by an R25 grant from the National Institute of Health (CA089017-06A2).
This work is also supported, in part, from the National Alliance for Medical Image Computing (NAMIC), funded by
the National Institutes of Health through the NIH Roadmap Initiative for Medical Research (Grant U54 EB005149 to
Kubicki and Shenton). All authors listed contributed to the study by either being involved in the design and
implementation (TW, MK, MES, JSS), the methodology used (TW, MK, MES, LO, CFW), the data acquisition and
analysis (TW, MK, RK, JLA, UK, DM, MN, PN), the interpretation of neuropsychological measures (PN), or were
involved in the writing (TW, MK, MES, JSS, LO, MN, RWM).

References
1. Andreasen NC. A unitary model of schizophrenia: Bleuler's “fragmented phrene” as schizencephaly.
Archives of General Psychiatry 1999;56:781–787. [PubMed: 12884883]
2. Friston K. Schizophrenia and the disconnection hypothesis. Acta Psychiatr Scand Suppl 1999;395:68–
79. [PubMed: 10225335]
3. Bartzokis G. Schizophrenia: breakdown in the well-regulated lifelong process of brain development
and maturation. Neuropsychopharmacology 2002;27:672–683. [PubMed: 12377404]
4. Norman RM, Malla AK, Williamson PC, Morrison-Stewart SL, Helmes E, Cortese L. EEG coherence
and syndromes in schizophrenia. Br J Psychiatry 1997;170:411–415. [PubMed: 9307688]
5. Fletcher P, McKenna PJ, Friston KJ, Frith CD, Dolan RJ. Abnormal cingulate modulation of fronto-
NIH-PA Author Manuscript

temporal connectivity in schizophrenia. Neuroimage 1999;9:337–342. [PubMed: 10075903]

6. Ragland J, Gur R, Valdez J, Turetsky B, Elliott M, Kohler C, et al. Event-related fMRI of frontotemporal
activity during word encoding and recognition in schizophrenia. Am J Psychiatry 2004;161:1004–
1015. [PubMed: 15169688]
7. Fields R. The other half of the brain. Sci Am 2004;290:54–61. [PubMed: 15045754]
8. Crow T. Schizophrenia as a transcallosal misconnection syndrome. Schizophr Res 1998;30:111–114.
[PubMed: 9549773]
9. Shenton M, Dickey C, Frumin M, McCarley R. A review of MRI findings in schizophrenia. Schizophr
Res 2001;49:1–52. [PubMed: 11343862]
10. Kubicki M, McCarley R, Westin CF, Park HJ, Maier S, Kikinis R, et al. A review of diffusion tensor
imaging studies in schizophrenia. J Psychiatr Res 2007;41:15–30. [PubMed: 16023676]
11. Hasan K. Diffusion tensor eigenvalues or both mean diffusivity and fractional anisotropy are required
in quantitative clinical diffusion tensor MR reports: fractional anisotropy alone is not sufficient.
Radiology 2006;239:611–612. author reply 612-613. [PubMed: 16641362]

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 10

12. Voineskos A, O'Donnell L, Lobaugh N, Markant D, Ameis S, Niethammer M, et al. Quantitative

examination of a novel clustering method using magnetic resonance diffusion tensor tractography.
Neuroimage 2009;45:370–376. [PubMed: 19159690]
NIH-PA Author Manuscript

13. Kindlmann G, Ennis D, Whitaker R, Westin C. Diffusion tensor analysis with invariant gradients and
rotation tangents. IEEE Trans Med Imaging 2007;26:1483–1499. [PubMed: 18041264]
14. Shergill S, Kanaan R, Chitnis X, O'Daly O, Jones D, Frangou S, et al. A diffusion tensor imaging
study of fasciculi in schizophrenia. Am J Psychiatry 2007;164:467–473. [PubMed: 17329472]
15. Seok J, Park H, Chun J, Lee S, Cho H, Kwon J, et al. White matter abnormalities associated with
auditory hallucinations in schizophrenia: a combined study of voxel-based analyses of diffusion
tensor imaging and structural magnetic resonance imaging. Psychiatry Res 2007;156:93–104.
[PubMed: 17884391]
16. Szeszko P, Robinson D, Ashtari M, Vogel J, Betensky J, Sevy S, et al. Clinical and neuropsychological
correlates of white matter abnormalities in recent onset schizophrenia. Neuropsychopharmacology
2008;33:976–984. [PubMed: 17581532]
17. Hubl D, Koenig T, Strik W, Federspiel A, Kreis R, Boesch C, et al. Pathways that make voices: white
matter changes in auditory hallucinations. Arch Gen Psychiatry 2004;61:658–668. [PubMed:
15237078]
18. Whitford, TJ.; Kubicki, M.; Shenton, ME. The neuroanatomical underpinnings of schizophrenia and
their implications as to the cause of the disorder: a review of the structural MR and diffusion literature.
In: Shenton, ME.; Turetsky, B., editors. Neuropsychiatric Imaging. New York: Cambridge University
Press; in press
NIH-PA Author Manuscript

19. Song S, Sun S, Ramsbottom M, Chang C, Russell J, Cross A. Dysmyelination revealed through MRI
as increased radial (but unchanged axial) diffusion of water. Neuroimage 2002;17:1429–1436.
[PubMed: 12414282]
20. Hollingshead, A. Two factor of index of social position. New Haven: Yale Station; 1965.
21. Wilkinson, GS.; Wide Range, I. The Wide Range Achievement Test - Administration Manual.
Wilmingham: Wide Range Inc; 1993.
22. Kubicki M, Westin C, Nestor P, Wible C, Frumin M, Maier S, et al. Cingulate fasciculus integrity
disruption in schizophrenia: a magnetic resonance diffusion tensor imaging study. Biol Psychiatry
2003;54:1171–1180. [PubMed: 14643084]
23. Kubicki M, Park H, Westin C, Nestor P, Mulkern R, Maier S, et al. DTI and MTR abnormalities in
schizophrenia: analysis of white matter integrity. Neuroimage 2005;26:1109–1118. [PubMed:
15878290]
24. Kubicki M, Styner M, Bouix S, Gerig G, Markant D, Smith K, et al. Reduced interhemispheric
connectivity in schizophrenia-tractography based segmentation of the corpus callosum. Schizophr
Res 2008;106:125–131. [PubMed: 18829262]
25. First, MB.; Spitzer, RL.; Gibbon, M.; Williams, BJ. Structured Clinical Interview for DSM-IV-TR
Axis I Disorders, Research Version, Non-patient Edition (SCID-I/NP). New York: Biometrics
Research, New York State Psychiatric Institute; 2002.
26. O'Donnell LJ, Kubicki M, Shenton ME, Dreusicke MH, Grimson WE, Westin CF. A method for
NIH-PA Author Manuscript

clustering white matter fiber tracts. AJNR Am J Neuroradiol 2006;27:1032–1036. [PubMed:

16687538]
27. Zollei L, Learned-Miller E, Grimson E, Wells W. Efficient population registration of 3D data.
International Conference of Computer Vision. 2005
28. Ng, A.; Jordan, M.; Weiss, Y. On spectral clustering: analysis and an algorithm. Cambridge, MA:
MIT Press; 2002.
29. O'Donnell L, Westin C. White matter tract clustering and correspondence in populations. Med Image
Comput Comput Assist Interv 2005;8:140–147. [PubMed: 16685839]
30. Seal M, Yücel M, Fornito A, Wood S, Harrison B, Walterfang M, et al. Abnormal white matter
microstructure in schizophrenia: a voxelwise analysis of axial and radial diffusivity. Schizophr Res
2008;101:106–110. [PubMed: 18262770]
31. Song S, Sun S, Ju W, Lin S, Cross A, Neufeld A. Diffusion tensor imaging detects and differentiates
axon and myelin degeneration in mouse optic nerve after retinal ischemia. Neuroimage
2003;20:1714–1722. [PubMed: 14642481]

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 11

32. Field, A. Discovering Statistics using SPSS. 3. Thousand Oaks, CA: Sage Publications; 2009.
33. Fisher, RA. The design of experiments. Edinburgh: Oliver and Boyd; 1935.
34. Liddle PF. The symptoms of chronic schizophrenia: a re-examination of the positive-negative
NIH-PA Author Manuscript

dichotamy. British Journal of Psychiatry 1987;151:145–151. [PubMed: 3690102]

35. Burzynska A, Preuschhof C, Bäckman L, Nyberg L, Li S, Lindenberger U, et al. Age-related
differences in white matter microstructure: Region-specific patterns of diffusivity. Neuroimage
2010;49:2104–2112. [PubMed: 19782758]
36. Zhang Y, AT D, Hayasaka S, Jahng G, Hlavin J, Zhan W, et al. Patterns of age-related water diffusion
changes in the human brain by concordance and discordance analysis. Neurobiol Aging. in press.
37. Rosenberger G, Kubicki M, Nestor P, Connor E, Bushell G, Markant D, et al. Age-related deficits in
fronto-temporal connections in schizophrenia: a diffusion tensor imaging study. Schizophr Res
2008;102:181–188. [PubMed: 18504117]
38. Buchsbaum M, Friedman J, Buchsbaum B, Chu K, Hazlett E, Newmark R, et al. Diffusion tensor
imaging in schizophrenia. Biol Psychiatry 2006;60:1181–1187. [PubMed: 16893533]
39. Kanaan R, Shergill S, Barker G, Catani M, Ng V, Howard R, et al. Tract-specific anisotropy
measurements in diffusion tensor imaging. Psychiatry Res 2006;146:73–82. [PubMed: 16376059]
40. Price G, Cercignani M, Parker G, Altmann D, Barnes T, Barker G, et al. Abnormal brain connectivity
in first-episode psychosis: a diffusion MRI tractography study of the corpus callosum. Neuroimage
2007;35:458–466. [PubMed: 17275337]
41. Kanaan R, Kim J, Kaufmann W, Pearlson G, Barker G, McGuire P. Diffusion tensor imaging in
schizophrenia. Biol Psychiatry 2005;58:921–929. [PubMed: 16043134]
NIH-PA Author Manuscript

42. Beaulieu C. The basis of anisotropic water diffusion in the nervous system - a technical review. NMR
Biomed 15:435–455. [PubMed: 12489094]
43. Pakkenberg B. Total nerve cell number in neocortex in chronic schizophrenics and controls estimated
using optical disectors. Biol Psychiatry 1993;34:768–772. [PubMed: 8292680]
44. Uranova N, Vostrikov V, Vikhreva O, Zimina I, Kolomeets N, Orlovskaya D. The role of
oligodendrocyte pathology in schizophrenia. Int J Neuropsychopharmacol 2007;10:537–545.
[PubMed: 17313698]
45. Davis KL, Stewart DG, Friedman JI, Buchsbaum M, Harvey PD, Hof PR, et al. White matter changes
in schizophrenia: evidence for myelin-related dysfunction. Arch Gen Psychiatry 2003;60:443–456.
[PubMed: 12742865]
46. Roy K, Murtie J, El-Khodor B, Edgar N, Sardi S, Hooks B, et al. Loss of erbB signaling in
oligodendrocytes alters myelin and dopaminergic function, a potential mechanism for
neuropsychiatric disorders. Proc Natl Acad Sci U S A 2007;104:8131–8136. [PubMed: 17483467]
47. Rotarska-Jagiela A, Schönmeyer R, Oertel V, Haenschel C, Vogeley K, Linden D. The corpus
callosum in schizophrenia-volume and connectivity changes affect specific regions. Neuroimage
2008;39:1522–1532. [PubMed: 18096406]
48. Black D, Taber K, Hurley R. Metachromatic leukodystrophy: a model for the study of psychosis. J
Neuropsychiatry Clin Neurosci 2003;15:289–293. [PubMed: 12928504]
NIH-PA Author Manuscript

49. Weinberger, DR.; Marceno, S. Schizohprenia as a neurodevelopmental disorder. In: Hirsch, SR.;
Weinberger, DR., editors. Schizophrenia. 2. New York: Wiley-Blackwell; 2003. p. 326-348.
50. Maher B. Delusional thinking and perceptual disorder. Journal of Individual Psychology 1974;30:98–
113. [PubMed: 4857199]
51. Ford J, Gray M, Faustman W, Roach B, Mathalon D. Dissecting corollary discharge dysfunction in
schizophrenia. Psychophysiology 2007;44:522–529. [PubMed: 17565658]
52. Wolkin A, Choi S, Szilagyi S, Sanfilipo M, Rotrosen J, Lim K. Inferior frontal white matter anisotropy
and negative symptoms of schizophrenia: a diffusion tensor imaging study. Am J Psychiatry
2003;160:572–574. [PubMed: 12611842]
53. Maddah M, Kubicki M, Wells W, Westin C, Shenton M, Grimson W. Findings in schizophrenia by
tract-oriented DT-MRI analysis. Medical Image Computing and Computer Assisted Intervention
2008:917–924. [PubMed: 18979833]
54. Kyriakopoulos M, Vyas N, Barker G, Chitnis X, Frangou S. A diffusion tensor imaging study of
white matter in early-onset schizophrenia. Biol Psychiatry 2008;63:519–523. [PubMed: 17662964]

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 12

55. Kyriakopoulos M, Perez-Iglesias R, Woolley J, Kanaan R, Vyas N, Barker G, et al. Effect of age at
onset of schizophrenia on white matter abnormalities. Br J Psychiatry 2009;195:346–353. [PubMed:
19794204]
NIH-PA Author Manuscript

56. Konopaske G, Dorph-Petersen K, Sweet R, Pierri J, Zhang W, Sampson A, et al. Effect of chronic
antipsychotic exposure on astrocyte and oligodendrocyte numbers in macaque monkeys. Biol
Psychiatry 2008;63:759–765. [PubMed: 17945195]
57. Rey M, Schulz P, Costa C, Dick P, Tissot R. Guidelines for the dosage of neuroleptics. I:
Chlorpromazine equivalents of orally administered neuroleptics. Int Clin Psychopharmacol
1989;4:95–104. [PubMed: 2568378]
58. Woods S. Chlorpromazine equivalent doses for the newer atypical antipsychotics. J Clin Psychiatry
2003;64:663–667. [PubMed: 12823080]
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 13
NIH-PA Author Manuscript

Figure 1.
A summary of the image processing procedures. Firstly, whole-brain tractography was
NIH-PA Author Manuscript

performed on each subject's Diffusion-Weighted Image (Panel A), and the resultant fibers
combined for all subjects. Fibers with similar shapes and spatial positions were then grouped
together into one of 400 fiber-clusters (Panel B). The fiber-clusters constituting the Corpus
Callosum were then identified for a randomly selected subject (Panel C), and subsequently
subdivided into six segments on the basis of which cortical regions the fiber-clusters projected
(Panel D). The fiber-clusters constituting each of the six Corpus segments were then
automatically extracted for all subjects, and the average FA and Mode of these fiber-clusters
calculated. As illustrated in Panel D, the six Corpus segments were: CC1 (red) – Frontal fibers
(defined as fibers projecting anterior to the SMA), CC2 (blue) – Premotor fibers (defined as
fibers projecting to the SMA or premotor areas), CC3 (green) – Sensorimotor fibers (defined
as fibers projecting to the primary motor or primary sensory cortices), CC4 (pink) – Parietal
fibers (defined as fibers projecting to the superior or inferior parietal lobules), CC5 (orange) –
Occipital fibers (defined as fibers projecting posterior to the parieto-occipital sulcus) and CC6
(yellow) – Temporal fibers (defined as fibers projecting ventrally to the temporal cortices).
The Corpus segmentations of four randomly selected subjects are displayed in Panel E.
NIH-PA Author Manuscript

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 14
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Figure 2.
Scatterplots illustrating the variations in FA (Panel A) and Mode (Panel B), between groups
(SZ in red, Controls in blue), and across Corpus segments (CC1 through CC6). The black bars
represent CC segment means, and the asterisks represent significant between-group differences
in diffusivity.

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 15
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Figure 3.
Scatterplots illustrating the variations in FA (Panel A), Radial Diffusivity (Panel B) and Axial
Diffusivity (Panel C) between groups (SZ in red, Controls in blue) for CC1 (frontal fibers).
The black bars represent CC segment means, and the asterisks represent significant between-
group differences in diffusivity.

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 16
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Figure 4.
Scatterplots illustrating the relationships between SZ patients' Reality Distortion score (i.e.,
the sum of the PANSS-Hallucinations and Delusions subscales) and their FA (Panel A) and
Radial Diffusivity (Panel B) in CC1. The black lines represent the line-of-best-linear-fit for
the data. The red dotted lines show the mean values of FA and Radial Diffusivity exhibited by
the healthy controls.
NIH-PA Author Manuscript

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.

 Whitford et al. Page 17

Table 1
Demographic details of the subject sample – means and standard deviations are provided.
NIH-PA Author Manuscript

Variable SZ patients (n=19) Controls (n=19) SZ vs. Controls

Age (yrs) 37.9 (9.60) 37.0 (12.2) t(36)=0.27, p=0.79

Handedness (1=RH, -1=LH) 0.71 (0.25) 0.79 (0.15) t(30)=0.98, p=0.33
Parental SES 2.63 (1.07) 2.22 (1.16) t(35)=1.12, p=0.27
Premorbid IQ (WRAT-3 Reading) 94.8 (11.4) 102.4 (9.9) t(26)=1.80, p=0.08

Medication dosage (CPZ equivalent1) 334 (288) - -

Medication type2 2 typical, 13 atypical, 1 both, 3 unmedicated - -

Age-at-onset (yrs) 23.5 (5.71) - -

PANSS-Positive 22.4 (8.76) - -
PANSS-Negative 21.9 (10.4) - -
PANSS-General 43.8 (15.5) - -

1
Medication dosages: Patients' medication dosages at time of clinical interview were converted into chlorpromazine (CPZ) equivalents on the basis
of the Rey et al. (57) conversion for typical antipsychotics and the Woods (58) conversion for atypical antipsychotics.
2
Medication type: At the time of interview, two patients were taking typical antipsychotic medications (fluphenazine or haloperidol), thirteen were
NIH-PA Author Manuscript

taking atypical antipsychotics (clozapine, risperidone or olanzapine), one patient was taking both typical and atypical antipsychotics, and three patients
were unmedicated.
NIH-PA Author Manuscript

Biol Psychiatry. Author manuscript; available in PMC 2011 July 1.