Corpus callosum in maltreated children with posttraumatic stress
disorder: A diffusion tensor imaging study
Andrea P. Jackowskia,d, Heather Douglas-Palumberic, Marcel Jackowskie,
Lawrence Wina, Robert T. Schultza, Lawrence W. Staibb,
John H. Krystala, Joan Kaufmana,c,⁎
aChild Study Center, Yale University School of Medicine, New Haven, CT, United States
bDepartment of Diagnostic Radiology, Yale University School of Medicine, New Haven, CT, United States
cDepartment of Psychiatry, Yale University School of Medicine, New Haven, CT, United States
dLiNC, Universidade Federal de São Paulo, São Paulo, Brazil
eInstitute of Mathematics and Statistics, University of São Paulo, CEP 05508-090, São Paulo, Brazil
Received 8 August 2007; accepted 9 August 2007
Contrary to expectations derived from preclinical studies of the effects of stress, and imaging studies of adults with
posttraumatic stress disorder (PTSD), there is no evidence of hippocampus atrophy in children with PTSD. Multiple pediatric
studies have reported reductions in the corpus callosum — the primary white matter tract in the brain. Consequently, in the present
study, diffusion tensor imaging was used to assess white matter integrity in the corpus callosum in 17 maltreated children with
PTSD and 15 demographically matched normal controls. Children with PTSD had reduced fractional anisotropy in the medial and
posterior corpus, a region which contains interhemispheric projections from brain structures involved in circuits that mediate the
processing of emotional stimuli and various memory functions — core disturbances associated with a history of trauma. Further
exploration of the effects of stress on the corpus callosum and white matter development appears a promising strategy to better
understand the pathophysiology of PTSD in children.
© 2007 Elsevier Ireland Ltd. All rights reserved.
Keywords: Posttraumatic stress disorder; Imaging; DTI; Children
Child abuse is frequently associated with long-term
significant psychiatric sequelae, and currently little is
known about the mechanisms that initiate and maintain
the various forms of psychopathology associated with
early trauma. While not all abused children develop
difficulties, many experience a chronic course of
psychopathology, with posttraumatic stress disorder
(PTSD) one of the most common psychiatric sequelae
of child maltreatment (Molnar et al., 2001).
Preclinical studies suggest that stress early in life can
promote long-term changes in stress reactivity and brain
development (Kaufman et al., 2000). These studies
provide a valuable heuristic in understanding the
Available online at www.sciencedirect.com
Psychiatry Research: Neuroimaging 162 (2008) 256–261
⁎Corresponding author. Department of Psychiatry, Yale University
School of Medicine, New Haven, CT, United States. Tel.: +1 203 764
9949; fax: +1 203 764 9990.
E-mail address: firstname.lastname@example.org (J. Kaufman).
0925-4927/$ - see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.
pathophysiology of PTSD in adults, with many of the
biological alterations associated with early stress in
preclinical studies reported in adults with PTSD and
other stress-related disorders. The application of research
findings from these preclinical studies in understanding
the neurobiology of PTSD in children, however, is
is reduction in hippocampal volume (Bremner, 2006).
et al., 2001; De Bellis et al., 1999; De Bellis et al., 2002;
Tupler and De Bellis, 2006).Instead, children with PTSD
have been found in two independent samples to have
reduction in the medial and posterior region of the corpus
CC area has likewise been reported in psychiatric
inpatients with a history of maltreatment compared with
psychiatric and healthy controlswithouta history of early
childhood trauma, with half the children in the maltreat-
ment group meeting criteria for PTSD at discharge
(Teicher et al., 2004). It has also recently been
documented in adults with PTSD as well (Villarreal
et al., 2004).
To the best of our knowledge, there is only one
published structural MRI study in prepubescent non-
human primates subjected to early stress (Sanchez et al.,
above, this study also failed to find evidence of
hippocampal atrophy, and instead reported reductions
in the medial and posterior CC.
Given prior results suggesting children, adolescents,
and adults with PTSD show atrophy of the medial and
we used diffusion tensor imaging (DTI) in the current
investigation to assess possible changes in myelination or
new technique that provides data on white matter
microstructure based on propertiesofdiffusion. Fractional
anisotropy (FA), a scale- and orientation-independent
measure of diffusion derived with DTI (Pierpaoli and
Basser, 1996), is the primary outcome measure examined
in the current report. FA is not a direct measure of
myelination, but a measure of anisotropy in water
diffusion, which increases with myelination (McGraw et
al., 2002; Snook et al., 2005). To the best of our
knowledge, this is the first investigation to utilize DTI in
maltreated children. Given that exposure to excessive
levels of stress hormones has been found to suppress glial
cell division critical for myelination(Lauder, 1983),it was
hypothesized that maltreated children with PTSD, com-
pared with normalcontrols,would have reducedFA in the
medial and posterior regions of the CC.
The sample included 32 children: 17 maltreated
children (7 males and 10 females) who met criteria for
PTSD secondary to intrafamilial abuse; and 15 demo-
graphically matched normal controls with no current or
lifetime history of psychiatric illness (7 males and
8 females). All the children within the PTSD group had
a history of protective services intervention for indicated
allegations of child maltreatment, and the absence of
maltreatment in the controls was verified by parent and
child reports, and review of State records. The children
mixed ethnically (25% Caucasian, 41% African Ameri-
can, 18% Hispanic, and 16% Biracial), and from low-
income families (Hollingshead socioeconomic status
(SES) score=30.8±11.7, range=14–50). The children
had a mean IQ of 93.2 (S.D.=14.4, range: 71–123), and
there were no differences between the two groups in age,
sex, race, SES, or IQ. In addition, all children were right
medication with central nervous system effects.
As previously described (Kaufman et al., 2006),
multiple informants and multiple methods were used to
assess children's maltreatment experiences. Only one
child experienced a single form of maltreatment; the
majority experienced three or more types: 71% had a
history of physical abuse, 29% sexual abuse, 47%
neglect, 59% emotional maltreatment, and 94% wit-
nessed domestic violence. Major depression was the
most common co-occurring diagnosis (41%) in this
sample, followed by other depressive diagnoses (30%),
oppositional defiant disorder (12%), and attention
deficit hyperactivity disorder (6%).
Both the Yale University Human Investigations
Committee and the Department of Children and Families
Institutional Review Board approved this study. The
majorityofchildrenfromthe present studywererecruited
from our ongoing studies of genetic and environmental
predictors of risk and resiliency in maltreated children
(Kaufman et al., 2006), and a subset of the children with
PTSD was recruited from a local child guidance agency.
The children's legal guardian provided written consent,
and all children provided written assent for study
participation. When the children's legal guardian was
not their birth parent, written assent was also obtained
from the birth parent if the birth parent was available.
257A.P. Jackowski et al. / Psychiatry Research: Neuroimaging 162 (2008) 256–261
2.2.1. Clinical evaluation
The Schedule for Affective Disorders and Schizo-
phrenia for School Aged Children — Present and
Lifetime Version (K-SADS-PL) (Kaufman et al., 1997),
a semi-structured psychiatric interview, was adminis-
tered to each child and one guardian. A number of
standardized and well-validated clinical symptomatolo-
gy rating scales were also administered and psychiatric
diagnoses were derived using best estimate procedures
as described previously (Pine et al., 2005).
2.2.2. MRI acquisition
Imaging data were acquired on a GE Signa 1.5T
scanner. DTI was performed using echo-planar imaging.
Thirty-five coronal slices were obtained using the
following parameters: matrix 128×128, TE=minimum,
TR=11000 ms, FOV=24 cm, NEX=1, no gap,
resolution=1.875 mm×1.875 mm×5 mm, across six
runs. The diffusion-sensitizing gradients were applied
along six non-collinear directions, with a b-value of
1000 s/mm2. T2-weighted images were also obtained
for each run. It has been suggested that if DTI voxel
should only be performed for the thickest white matter
tracts (Smith et al., 2007). Since the corpus callosum is a
thick fiber tract characterized by tightly packed bundles
that run in a similar direction, the acquisition parameters
utilized in the current investigation provide adequate
voxel resolution and minimal artifact effects (Oouchi
et al., 2007). Structural MR images were acquired using
a 3D SPGR coronal acquisition series (TR=18 ms,
TE=5 ms, flip angle=30°, NEX=2, matrix size=
256×160, FOV=24 cm, 124 slices, thickness=1.5 mm).
2.2.3. Image analysis
The corpus was outlined at the midsagittal slice and
segmented into seven sections with a semi-automated
program generated using the guidelines delineated by
Witelson (1989). The seven corpus regions were defined
on the structural data, and brought into the DTI space to
obtain the region of interest FA assessments. ICC inter-
rater reliability was N0.90 for all corpus sections.
The DTI data were analyzed using a custom software
package, part of the Yale BioImage Suite (Papademetris
et al., 2006) using an encoding gradient set containing
six directions based on an energy minimization
procedure (Papadakis et al., 1999). Each diffusion-
weighted series was registered to a single T2 image
volume. A 12-parameter affine registration was per-
formed for each individual SPGR to the same T2 image.
Computation of FAwas performed for each voxel within
each region of the CC.
2.2.4. Statistical analyses
for normality using the Shapiro–Wilk test. As standard-
ized transformations failed to normalize skewed mea-
indicated. The associations of DTI measures with age,
sex, IQ, SES, and ratings of motion were also examined,
but were not included in the model examining group
differences as they did not relate to the FA values. As
groups differed on total white matter volume, with the
PTSD group being significantly smaller (F=5.3,
Pb0.03), we corrected for white matter volume using it
as a covariate within the multiple analysis of covariance
(MANCOVA). MANCOVA was used to test for the
effects of diagnosis and gender on the FA of the seven
corpus regions. Since the MANCOVA was significant
(Wilks' Lambda, F=3.10, Pb0.04), follow-up univariate
analyses of covariance (ANCOVA) were conducted.
3.1. Group differences on DTI measures
Table 1 presents group means, effect sizes (Cohen'sd)
and univariate comparisons for corpus DTI data.
Fractional anisotropy assessments of the corpus callosum
PTSD N=17Controls N=15F-value statisticP-valueCohen's d effect size
Region 1 — Rostrum
Region 2 — Genu
Region 3 — Rostral body
Region 4 — Ant. midbody
Region 5 — Post. midbody
Region 6 — Isthmus
Region 7 — Splenium
DTI measures are adjusted for total white matter volume (mean±SE).
aRaw scores are presented in the table, rank transformed scores were used in the analyses.
258A.P. Jackowski et al. / Psychiatry Research: Neuroimaging 162 (2008) 256–261
Compared with controls, maltreated children with PTSD
had significantly reduced FA in two of the four predicted
corpus regions (anterior and posterior midbody), and a
trend toward significantly reduced FA in a third region
(splenium) (Fig. 1).
3.2. Correlation between DTI measures and clinical
FA measures in regions 2 (Rho=−0.40, Pb0.02), 4
(Rho=−0.43, Pb0.02), and 7 (Rho=−0.38, Pb0.03)
of the corpus callosum correlated significantly with total
anxiety scores obtained using the Screen for Child
Anxiety and Related Disorders (Birmaher et al., 1997).
FAvalues in these same regions also correlated with the
Panic and Separation Anxiety subscale scores of this
measure (Rho range: −0.48 to −0.52, Pb0.01, all
comparisons), but were not related to depression or
in the CC observed in this study may be due to reduced
myelination or other subtle alterations in axonal structure
(e.g., neurofilaments, microtubules). CC myelination oc-
childhood into early adulthood (Giedd et al., 1996).
Consequently, it has been suggested that different regions
the effects early trauma (Teicher et al., 2002, 2004).
The medial and posterior portions of the CC contain
interhemispheric projections from the auditory cortices,
cortices. The CC also includes connections from the
inferior parietal lobe to the contralateral inferior parietal
lobe, superior temporal sulcus, cingulate, retrosplenial
cortex, and parahippocampal gyrus (Pandya and Seltzer,
1986). Several of the regions with interhemispheric
projections through the medial and posterior CC have
connections with prefrontal cortical areas, and are
involved in circuits that mediate the processing of
disturbances associated with PTSD.
In the current investigation, however, it is impossible
to determine if the CC alterations are a function of PTSD
per se, or maltreatment, as all subjects had PTSD
secondary to physical abuse, sexual abuse, or exposure
to domestic violence. Teicher and colleagues (1997), in
their study of maltreated inpatients, did not find the
presence of a PTSD diagnosis to contribute in predicting
CC area above and beyond the effects of neglect,
although the study was not fully powered to investigate
this issue (Teicher et al., 2002). There is preliminary
evidence to suggest that sociopathy is associated with
increased CC area, and that this effect is independent of
childhood history of abuse (Raine et al., 2003), but
further work in this area is warranted. The use of two
(e.g., maltreatment vs. no maltreatment) by two (e.g.,
PTSD vs. no PTSD) research designs in carefully
characterized samples will help to determine the
specificity of CC changes observed in association with
trauma, PTSD, and other psychiatric diagnoses.
It is possible that CC changes observed in PTSD
exposure to stressful events that predispose individuals to
35 identical twin pairs that were discordant for combat
exposure (Gilbertson et al, 2002). Of the 35 combat-
exposed twins, 12 developed PTSD. Combat-exposed
to have smaller hippocampal volumes than combat-
exposed twins without PTSD and their unexposed co-
twins. Given that the identical twins that were not exposed
to those of their combat-exposed co-twins who developed
PTSD, and significantly smaller than the hippocampi of
combat-exposed men who did not develop PTSD, the
authors concluded that the reduced hippocampal volume
represented a pre-existing vulnerability factor rather than a
consequence of traumaexposure.This interpretationhas to
Fig. 1. Subregions of corpus callosum. Overlay of the seven corpus
callosum regions onto a fractional anisotropy (FA) map. Circle:
callosal areas (anterior and posterior midbody) of reduced FA in
maltreated children with PTSD compared with controls.
259A.P. Jackowski et al. / Psychiatry Research: Neuroimaging 162 (2008) 256–261
be viewed with caution, however, as the combat-exposed
veterans who developed PTSD and their non-combat-
exposed co-twins were significantly more likely to have a
history of alcohol dependence than the combat-exposed
veterans who did not develop PTSD and their non-combat
exposed co-twins. Moreover, childhood histories of sexual
veterans who developed PTSD and their non-combat-
exposed co-twins than in than the other twin pairs. More
in individuals with PTSD are a result of genetic and
environmental factors. Additional use of twin designs (De
Geus et al., 2007) and other molecular genetic approaches
(Canli et al., 2006; Hariri and Weinberger, 2003) will help
to untangle these effects.
There is emerging preclinical and clinical evidence to
suggest that the neurobiological effects of stress vary at
different developmental periods, with white matter
changes more prominent than subcortical gray (e.g.,
hippocampal) matter changes in juvenile cohorts.
Developmental differences in the neurobiological cor-
relates of PTSD across the life cycle have raised
questions as to whether PTSD is one and the same
disorder in children and adults. With minimum changes,
however, the criteria required to diagnose PTSD can be
applied to children as young as 6 years of age (Stover
and Berkowitz, 2005). Emerging data on neurobiolog-
ical correlates and treatment response of depressed
children, adolescents, and adults likewise demonstrate
similarities and differences across the life cycle, with
pharmacological treatments (e.g. tricyclic antidepres-
sants) that are highly effective in adults proving to be no
more effective than placebo in children with depression
(Kaufman et al., 2004). Developmental differences in
the neural underpinning of PTSD will likewise hamper
the ability to extrapolate from adult treatment studies in
guiding intervention strategies for children with PTSD,
highlighting the need for further work in this area.
Although promising, the significance of our results is
limited by the small sample studied, and the poor
resolution of the 1.5 T DTI data collected, which pro-
hibited tract- tracing and examination of the circuits
affected by the FA changes in the corpus. Smaller and
isotropic voxel size will certainly be beneficial for fiber
tracking as well as for the voxel-wise analytic tech-
niques that have become available and can enhance the
quality of future work in this area.
Further exploration of the effects of stress on the
development of the CC, corticolimbic circuitry, and
white matter tracts appears a promising strategy to better
understanding the etiology and pathophysiology of
PTSD in children. Developmental differences in the
neurobiological correlates of PTSD across the life cycle
highlight the need for more work in this area.
The authors thank the children and families, the staff
at the State Department of Children and Families, and
the clinicians at Clifford Beers Clinic that facilitated the
completion of this work. The authors also acknowledge
Deborah Lipschitz, M.D., and Damion Grasso, M.A.,
for their help in the collection of the clinical data and
derivation of best estimate psychiatric diagnoses. This
research was funded by grants from the NIH:
1R01MH65519-01 (JK), P50 AA-12870-04 (JHK,
JK); support from the National Center for Posttraumatic
Stress Disorder, Veterans Administration, West Haven,
Connecticut, and Yale University General Clinical
Research Center Grant (M01RR06022).
Birmaher, B., Khetarpol, S., Brent, D., Cully, M., Balach, L.,
Kaufman, J., Neer, S.M., 1997. The Screen for Child Anxiety
Related Emotional Disorders (SCARED): scale construction and
psychometric properties. Journal of the American Academy of
Child and Adolescent Psychiatry 36, 545–553.
Bremner, J.D., 2006. The relationship between cognitive and brain
changes in posttraumatic stress disorder. Annals of the New York
Academy of Sciences 1071, 80–86.
Canli, T., Qiu, M., Omura, K., Congdon, E., Haas, B.W., Amin, Z.,
Herrmann, M.J., Constable, R.T., Lesch, K.P., 2006. Neural
correlates of epigenesis. Proceedings of the National Academy of
Sciences of the United States of America 103 (43), 16033–16038
(Epub 12006 Oct 16010).
Carrion, V.G., Weems, C.F., Eliez, S., Patwardhan, A., Brown, W.,
Ray, R.D., Reiss, A.L., 2001. Attenuation of frontal asymmetry in
pediatric posttraumatic stress disorder. Biological Psychiatry 50,
Hall, J., Moritz, G., 2002. Brain structures in pediatric maltreat-
ment-related posttraumatic stress disorder: a sociodemographically
matched study. Biological Psychiatry 52, 1066–1078.
Boring, A.M., Frustaci, K., Ryan, N.D., 1999. Developmental
de Geus, E.J., van't Ent, D., Wolfensberger, S.P., Heutink, P.,
Hoogendijk, W.J., Boomsma, D.I., Veltman, D.J., 2007. Intrapair
differences in hippocampal volume in monozygotic twins
discordant for the risk for anxiety and depression. Biological
Psychiatry 61 (9), 1062–1071.
Giedd, J.N., Snell, J.W., Lange, N., Rajapakse, J.C., Casey, B.J.,
Kozuch, P.L., Vaituzis, A.C., Vauss, Y.C., Hamburger, S.D.,
Kaysen, D., Rapoport, J.L., 1996. Quantitative magnetic resonance
imaging of human brain development: ages 4–18. Cerebral Cortex
Hariri, A.R., Weinberger, D.R., 2003. Imaging genomics. British
Medical Bulletin 65, 259–270.
260A.P. Jackowski et al. / Psychiatry Research: Neuroimaging 162 (2008) 256–261
Kaufman, J., Birmaher, B., Brent, D., Rao, U., Flynn, C., Moreci, P., Download full-text
Williamson, D., Ryan, N., 1997. Schedule for affective disorders
and schizophrenia for school-age children-present and lifetime
version (K-SADS-PL): initial reliability and validity data. Journal
of the American Academy of Child and Adolescent Psychiatry 36,
Kaufman, J., Plotsky, P., Nemeroff, C., Charney, D., 2000. Effects of
early adverse experience on brain structure and function: clinical
implications. Biological Psychiatry 48, 778–790.
Kaufman, J., Blumberg, H.P., Young, C.M., 2004. Neurobiology of
early onset mood disorders, In: Charney, D.S., Nestler, E.J. (Eds.),
Neurobiology of Mental Illness, 2nd ed. Oxford University Press,
Kaufman, J., Yang, B.Z., Douglas-Palumberi, H., Grasso, D.,
Lipschitz, D., Houshyar, S., Krystal, J.H., Gelernter, J., 2006.
Brain-derived neurotrophic factor-5-HTTLPR gene interactions
and environmental modifiers of depression in children. Biological
Psychiatry 59, 673–680.
Lauder, J.M., 1983. Hormonal and humoral influences on brain
development. Psychoneuroendocrinology 8, 121–155.
Molnar, B.E., Buka, S.L., Kessler, R.C., 2001. Child sexual abuse and
subsequent psychopathology: results from the National Comor-
bidity Survey. American Journal of Public Health 91, 753–760.
Oouchi, H., Yamada, K., Sakai, K., Kizu, O., Kubota, T., Ito, H.,
Nishimura, T., 2007. Diffusion anisotropy measurement of brain
white matter is affected by voxel size: underestimation occurs in
areas with crossing fibers. American Journal of Neuroradiology
Pandya, D.N., Seltzer, B., 1986. The topography of commisural fibers.
In: Lepore, F., Ptito, M., Jasper, H.H. (Eds.), Two Hemispheres —
One Brain: Functions of the Corpus Callosum, vol 17. Alan R.
Liss, Inc., New York, pp. 47–74.
Papadakis,N.G., Xing, D., Huang, C.L.H.,Hall, L.D., Carpenter, T.A.,
1999. A comparative study of acquisition schemes for diffusion
tensor imaging using MRI. Journal of Magnetic Resonance 137,
Papademetris, X., Jackowski,M., Rajeevan, N., Constable,R.T., Staib,
L., 2006. Bioimage Suite, Yale Department of Diagnostic Imaging.
Pierpaoli, C., Basser, P.J., 1996. Toward a quantitative assessment of
diffusion anisotropy. Magnetic Resonance in Medicine 36 (6),
Pine, D.S., Mogg, K., Bradley, B.P., Montgomery, L., Monk, C.S.,
McClure, E., Guyer, A.E., Ernst, M., Charney, D.S., Kaufman, J.,
2005. Attention bias to threat in maltreated children: implications
for vulnerability to stress-related psychopathology. American
Journal of Psychiatry 162, 291–296.
Raine, A., Lencz, T., Taylor, K., Hellige, J.B., Bihrle, S., Lacasse, L.,
Lee, M., Ishikawa, S., Colletti, P., 2003. Corpus callosum
abnormalities in psychopathic antisocial individuals. Archives of
Journal Psychiatry 60 (11), 1134–1142.
Sanchez, M.M., Hearn, E.F., Do, D., Rilling, J.K., Herndon, J.G.,
1998. Differential rearing affects corpus callosum size and
cognitive function of rhesus monkeys. Brain Research 812, 38–49.
Smith, S.M., Johansen-Berg, H., Jenkinson,M., Rueckert, D., Nichols,
T.E., Miller, K.L.,Robson,M.D., Jones, D.K., Klein, J.C., Bartsch,
A.J., Behrens, T.E., 2007. Acquisition and voxelwise analysis of
multi-subject diffusion data with tract-based spatial statistics.
Nature Protocols 2, 499–504.
Teicher, M.H., Andersen, S.L., Polcari, A., Andersen, S.L., Dumont,
N., Ackerman, E., 2002. Developmental neurobiology of child-
hood stress and trauma. Psychiatric Clinics of North America 25,
Teicher, M.H., Dumont, N.L., Ito, Y., Vaituzis, C., Giedd, J.N.,
Andersen, S.L., 2004. Childhood neglect is associated with
reduced corpus callosum area. Biological Psychiatry 56, 80–85.
Teicher, M.H., Ito, Y., Glod, C.A., et al., 1997. Preliminary evidence
for abnormal cortical development in physically and sexually
abused children using EEG coherence and MRI. Annals of the
New York Academy of Sciences 821, 160–175.
Tupler, L.A., De Bellis, M.D., 2006. Segmented hippocampal volume
in children and adolescents with posttraumatic stress disorder.
Biological Psychiatry 59, 523–529.
Villarreal, G., Hamilton, D.A., Graham, D.P., Driscoll, I., Qualls, C.,
Petropoulos, H., Brooks, W.M., 2004. Reduced area of the corpus
callosumin posttraumatic stressdisorder. PsychiatryResearch131,
Witelson, S.F., 1989. Handand sex differences in the isthmusand genu
of the human corpus callosum. A postmortem morphological
study. Brain 112, 799–835.
261A.P. Jackowski et al. / Psychiatry Research: Neuroimaging 162 (2008) 256–261