Epilepsy & Behavior 11 (2007) 20–24

www.elsevier.com/locate/yebeh

Autism: The first firm finding = underconnectivity?

John R. Hughes *

Department of Neurology, University of Illinois Medical Center at Chicago, 912 South Wood Street, Chicago, IL 60612, USA

Received 14 December 2006; revised 13 March 2007; accepted 14 March 2007

Available online 24 May 2007

Abstract

In January 2005, J.R. Hughes and M. Melyn published an electroencephalographic study on autistic children and found 46% with
seizures and also a relatively high prevalence of 20% with epileptiform discharges but without any clinical seizures (Clin EEG Neurosci
2005;36:15–20). Because the discharges have always been viewed as focal events and the clinical seizures as requiring spread, the
conclusion from these data was that children with autism may have a deficiency of corticocortical fibers. Since that time many MRI
and functional MRI studies have been published confirming that one of the first findings in this devastating condition is underconnec-
tivity. Specific findings are the thinning of the corpus callosum and the reduced connectivity, especially with the frontal areas and also the
fusiform face area. Other studies involving positron emission tomography scans, magnetoencephalography, and perception have added
to the evidence of underconnectivity. One final point is the initial overgrowth of white matter in the first 2 years of life in autistic children,
followed later by arrested growth, resulting in aberrant connectivity; myelination of white matter will likely be significant in the etiology
of autism.
! 2007 Elsevier Inc. All rights reserved.

Keywords: Cerebral cortex; White matter; Magnetic resonance imaging; Functional magnetic resonance imaging; Fusiform face area; Corpus callosum;
Frontal area; Brain circuits; Myelination; Autism

1. Introduction on autism and autistic spectrum disorders that are consis-

In January 2005, J.R. Hughes and M. Melyn [1] pub-
lished an EEG study on autism, reporting seizures in
46%, as well as a relatively high prevalence of epileptiform
discharges (20%) in others who did not have a history of
clinical seizures. Although 20% may not seem high, this
value was significantly different from that of the control
group of matched children without autism. Because dis-
charges are usually regarded as an exquisitely focal finding
and the seizures require some spread from the focus, the
final conclusion from these data was that a clue to one
problem in autistic children may be a deficiency of cortico-
cortical fibers to account for this presumed lack of spread
from a focus. After that study, 19 studies—mainly MRI
and functional MRI (fMRI) studies—have been published
*
Fax: +1 312 996 4169.
E-mail address: jhughes@uic.edu
tent with the original conclusions from the EEG study.
The evidence is now clear that one major finding in autism
is the underconnectivity within the brain. This finding is
consistent with the behavior of these children, who tend
to concentrate on some object, rather than on any person,
but without any significant relationship to other sensory
modalities, as may be expected from disconnected cerebral
circuits.
2. Theoretical considerations and methods
2.1. Other theories
Some general conclusions relevant to autism are that
growth dysregulation [2] is likely involved and also that
mechanical factors from cortical folding influencing the
lamina [3] may also be implicated. A separate and different
view of this disorder is that females are better than males as

1525-5050/$ - see front matter ! 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.yebeh.2007.03.010
 J.R. Hughes / Epilepsy & Behavior 11 (2007) 20–24
‘‘empathizers,’’ dealing with people, and males are better
‘‘systemizers,’’ dealing with the rules governing the
behavior of things [4]. Autism is then viewed as an
‘‘extreme male brain’’ [5]. Another theory relates to
‘‘mirror neurons,’’ action-coding cells that are activated
both by passive observation of actions and by active execu-
tion of the same movements [6]. These neurons are consid-
ered to be reflected in mu waves (8–12 Hz) from central
scalp areas, known to appropriately attenuate in children
when either moving themselves or observing movement in
others. The conclusion was that such an observation—
execution matching system—exists even in the immature
brain but that such a circuit in autism may be associated
with a ‘‘faulty mirror neuron’’ system.
2.2. Methods (fMRI)
The next eight studies that are mentioned deal with
fMRI as a method to investigate autism, determining func-
tional neural connections. The first study used low-fre-
quency oscillations that were temporally correlated in
functionally related brain regions to quantify the degree
of functional connectivity by detecting similarities in the
oscillation, called ‘‘patterns with hierarchical clustering’’
[7]. The next study used a ‘‘time series to estimate fre-
quency-dependent correlational matrices’’ [8]. The next
study to determine functional connectivity used ‘‘concur-
rent spontaneous activity and spatial independent compo-
nent analysis’’ [9]. Two other studies used the decrease in
activity during various tasks [10], and one referred to a fre-
quent finding of decreased cerebral blood flow during a
given task [11]. Other studies used ‘‘network modulation
from rest to a given task’’ [12], measuring ‘‘absolute brain
state maps’’ from rest to a given activity [13] or ‘‘linear cor-
relational coefficients’’ [14] as a means of determining func-
tional connectivity.
Thus, examining the complex multivariate correlational
analyses in the rise and fall of the BOLD signal can help to
determine connectivity.
3. Decreased connectivity
3.1. MRI studies
3.1.1. Corpus callosum
The first group of studies discussed here used MRI to
determine underconnectivity. One of the most common
findings in autism has been the decrease in the size of
the corpus callosum. As early as 1993, Courchesne et al.
[15] reported the thinning of the corpus callosum as part
of white matter volume loss, in addition to the loss in the
parietal lobes. Also, the cortical volume loss in the same
parietal lobes was found to extend into the adjacent supe-
rior frontal and occipital lobes. The cortical volume loss
either could be secondary to disconnection or could result
in underconnectivity or possibly be unrelated. Eleven
years later, Chung et al. [16] confirmed that white matter
21
density is an index of neural connectivity and that
patients with autism exhibit less density in the genu,
rostrum, and splenium of the corpus callosum, as a hypo-
plasia, rather than as an atrophy. In 2006, Vidal et al. [17]
reported a significant reduction in both the splenium and
genu of the corpus callosum, and in the same year, Boger-
Megiddo et al. [18] confirmed that the corpora callosa
were disproportionately small, reflecting decreased inter-
hemispheric connectivity. Finally, Alexander et al. [19]
also reported small corpus callosum volumes, in addition
to slow processing speeds. For example, the latter patients
with small corpora callosa had significantly lower IQs
than controls.
3.1.2. Frontal area
Another area of interest often mentioned in autistic chil-
dren is the frontal area. Reduced connectivity was indi-
cated in the (fMRI) finding that bilateral inferior frontal
areas [20] were deficient in BOLD signal correlation with
the time series in the visual area. These findings were
viewed as compatible with the hypothesis of ‘‘mirror neu-
ron’’ defects mentioned previously. A general comment
was that a ‘‘dorsal stream connectivity’’ in autism was
not likely fully functional, based on BOLD signals during
visuomotor coordination, reporting reduced connectivity
with bilateral frontal areas. Another study demonstrated
significant localized gray matter reductions within the
frontostriatal circuits and within the parietal networks.
One other report on the frontal area, by Courchesne and
Pierce [21], reviewed different methods. Evidence was
found that connectivity within the frontal lobe was disorga-
nized and inadequately selective, whereas connectivity
between the frontal cortex and other systems was poorly
synchronized and weakly responsive based on inferences
from stimulation data. One final conclusion was that the
reduced long-distance corticocortical reciprocal activity
that appeared in their studies would impair frontal lobe
function. Also reported were decreases in volume beyond
the frontal area, namely, in ventral and superior temporal
gray matter, left internal capsule, fornices, and cerebellar
white matter [22]. The data were suggestive of abnormal
anatomy and connections of the limbic–striatal social brain
system in autistics.
3.1.3. Other studies
Other MRI studies have shown that minicolumnar
width and mean neuron and nucleolar cross sections are
smaller in autistic children. The findings suggest a bias
toward shorter connectivity fibers [23]. One other study
demonstrated changes in cerebellar gray and white matter
volume [24], and a final MRI study showed a decrease in
the gray matter in the right thalamus [25], although with
unexpected increases in the right fusiform gyrus, tempo-
ral–occipital region, and left frontal pole. The latter
increases may be explained by an additional finding of
increased total gray matter volume, which may reverse in
later years.
22 J.R. Hughes / Epilepsy & Behavior 11 (2007) 20–24
3.2. Functional MRI studies
3.2.1. Pre-2006 investigations
It should be made clear that these studies do not directly
assess connectivity, and conclusions are drawn from indi-
rect measurements. Most of these studies were published
in 2005 and 2006. Exceptions include the report from
2003 [26] involving patients performing finger movements,
which demonstrated abnormal variability and scatter of
functional maps, suggesting early-onset disturbances in
the development of cerebello-thalamo-cortical pathways
in autism. Two other exceptions to very recent publications
were reported in 2004. One showed less activation in
Broca’s left inferior frontal gyrus, but especially less syn-
chronization between various cortical areas, related to
‘‘neurobiological foundations of underconnectivity in aut-
ism’’ [27]. The patients were tested during sentence compre-
hension. The other study [28] showed reduced activation in
responses to the faces of strangers in the fusiform face area,
but also in the medial frontal lobe and the amygdala. The
general conclusion was that autistic children exhibit defects
in the systems that modulate the fusiform face area. How-
ever, it is impossible to distinguish whether the major issue
is an abnormal area affecting other regions or a primary
disorder of connectivity.
3.2.2. 2006 investigations of frontal areas
As in the MRI studies, the 2006 fMRI studies empha-
sized the frontal areas. Silk et al. [29] studied patients per-
forming a mental rotation task and reported a dysfunction
in frontostriatal networks in autism by showing less activa-
tion in the lateral and medial frontal, dorsolateral prefron-
tal, and anterior cingulate cortex and also in the caudate
nucleus. Turner et al. [30] studied patients during a visuo-
motor task and found that the circuits from BOLD signal
cross correlation with this task were less pronounced or
even absent in association with the orbitofrontal area and
caudate nuclei, as well as the associative, oculomotor,
and motor areas. These findings indicated an inefficiently
organized functional connectivity between these regions.
Similar results were reported by Just et al. [31], who studied
patients performing a Tower of London test to assess exec-
utive function. They reported lower synchronization
between the frontal and parietal areas, in addition to a
smaller corpus callosum, indicating reduced intracortical
connectivity or underconnectivity. Other reports, using
patients processing sentences with different imagery con-
tent, were that functional connectivity among cortical
regions, especially the language and spatial centers, is not
well synchronized, again indicating underconnectivity
[32]. Still another study [33] using a cognitive task reported
loosely connected anterior–posterior circuits as an indica-
tion of cortical underconnectivity. A similar conclusion
of ‘‘reduced corticocortical connectivity’’ was reached in
another report [34] of patients performing a visuomotor
task, even though the opposite was found for subcorti-
cal–cortical connectivity. The different areas implicated in
the latter studies may be a reflection of the short period
during which investigators have studied this problem, but
the emphasis on frontal areas seems to be clear. The pres-
ent section indicates that most of these studies were done
after 2005.
3.3. Other studies
One positron emission tomography (PET) study was
performed to demonstrate reduced functional connectivity
by showing less activation in the medial prefrontal, ante-
rior, and superior temporal cortex and the connections to
the extrastriate region [35]. A magnetoencephalography
(MEG) study revealed an abnormal left hemisphere
gamma oscillation (40 Hz), ‘‘augmenting connectivity theo-
ries of autism by demonstrating deficiencies in long-range
neural communication’’ [36]. In three studies involving per-
ception, similar conclusions were drawn. In one study [37],
increased amplitude of intrusive saccades and reduced
latency of target fixation led to the conclusion of faulty
functional connectivity in corticocerebellar networks. In
another investigation [38], autistic children were found to
be deficient in identifying complex texture-defined gratings,
leading to the conclusion of atypical neural connectivity. In
the third study [39], autistic children showed a lack of
‘‘attentional modulation,’’ especially for social stimuli,
and the conclusion of poor connectivity between extrastri-
ate and striate regions was drawn. One final study [40] con-
cluded that the developmental process was disrupted in
autistic children by showing ‘‘abnormal synthesis capacity’’
with respect to brain serotonin. It is clear that the latter
studies were not directly measuring connectivity, but pro-
duced data that were only consistent with underconnectivity.
4. Emphasis on facial area
One of the major clinical findings in the autistic spec-
trum is the lack of apparent recognition of faces, as these
children tend to concentrate on things and not people. A
number of studies have provided evidence to account for
this phenomenon. In 2003, Frith [41] reported that autis-
tic children failed to activate the fusiform face area during
face perception tests, also demonstrating weak activation
of medial frontal cortex and the superior temporal gyrus.
The conclusion was that this finding was a marker for
abnormal connectivity, possibly from lack of neural prun-
ing. During the next year, Waiter et al. [25] reported an
unexpected increase in gray matter volume in the right
fusiform gyrus, as well as in the right temporal–occipital,
left frontal, and medial frontal cortex. The conclusion was
that this finding may ‘‘reflect an abnormally functioning
social cognitive neural network.’’ Finally, in 2006 Bird
et al. [39] reported a lack of attentional modulation of
face-selective regions, concluding abnormal connectivity.
Two other studies specifically involved fMRI. One was
a study that autistic children exhibit less fusiform activity
in response to strangers than to familiar faces. In general,
 J.R. Hughes / Epilepsy & Behavior 11 (2007) 20–24
they demonstrated a more limited network in response,
even to familiar faces, at the same time showing a
decrease in medial frontal lobe function. Their conclusion
was that the dysfunction in the face area may reflect
defects in systems that modulate the area, rather than a
defect in the face area itself. The last study [42] reviewed
the evidence that abnormal functional mechanisms are
involved in face recognition, resulting in ‘‘dysfunctional
reciprocal cortico-subcortical connections.’’
5. Increase in white matter
With considerable evidence indicating a decrease in the
networks involving various areas, it may be somewhat
surprising that an increase in white matter could be found.
The earliest evidence came from Frith’s study [41] showing
an increase in total brain volume during the first few years
as a marker of abnormal connectivity from lack of pruning.
In the next year (2004), Courchesne [43] was more specific
by pointing out that the overgrowth occurred during the
first 2 years of life, followed by abnormally slow or arrested
growth in the next few years. This disruption was viewed as
related to circuit formation, resulting in aberrant connec-
tivity. In the same year, Herbert et al. [44] spelled out that
the increased brain volume was an unexplained white mat-
ter enlargement, involving more the later or longer myeli-
nating regions. These changes were viewed as affecting
primarily intrahemispheric and corticocortical connections.
In the next year, Herbert [45] confirmed that a growth
spurt shortly after birth was primarily from white matter
and that neuroinflammation was often present to explain
impaired complex processing and underconnectivity. In
2006, Hendry et al. [46] reported that the increase in cere-
bral white matter, especially in the frontal and parieto-
occipital areas, suggested abnormalities in white matter tis-
sue water content. Courchesne et al. [47] later discussed the
early brain overgrowth and later arrest as a microstructural
maldevelopment resulting in short-distance overconnectiv-
ity that was ineffective. In addition, the conclusion was a
‘‘failure of long distance cortico-cortical coupling and a
reduction in frontal-posterior reciprocal connectivity.’’
6. Myelination
As white matter myelination generally reflects the pro-
gression of functional brain maturation [48], it is reason-
able to explore the process of myelination as a putative
etiology of autism. The time when this disorder usually
reveals itself clinically is between 2 and 3 years of age
[43], and, therefore, it is this same time interval that
may lead us to discover the etiology of autism. The onset
of myelination, as indicated by the expression of myelin
basic protein, is usually at 1 year of age, and the progres-
sion to ‘‘adultlike’’ staining occurs at nearly 2 years of age
[49]. Other data would indicate that adult values are
reached during the third year [50]. Thus, a crucial time
for myelination and also for the clinical expression of
23
autism is between the second and third years of life.
Consistent with the latter relationship, data have shown
that the most significant decrease in (periventricular)
white matter during maturation occurs during the first 2
years [51], and patients with developmental delay have,
in fact, exhibited a significant reduction in the content
of myelinated white matter [48]. As mentioned previously,
early brain overgrowth occurs, but later arrest is then
observed [47]. The factors that may play a significant role
in this maturation include the myelin-forming oligoden-
drocytes, and injury to their progenitors likely would con-
tribute to the pathogenesis of white matter development
[52]. One other factor that may be significant is hypogly-
cemia, as the latter condition inhibits oligodendrocyte
development and myelination, also triggering apoptotic
cell death in oligodendrocyte precursor cells [53]. This
reviewer is unaware of whether or not hypoglycemia is
common in autism. These latter considerations are only
clues to uncovering the etiology of autism.
In summary, the first firm finding in children with aut-
ism or the autistic spectrum is underconnectivity within
the brain. Both MRI and fMRI studies during the past 2
years have shown thinning of the corpus callosum and
reduced connectivity, especially in the frontal and fusiform
face areas, in addition to other circuits in the brain. The ini-
tial overgrowth of the white matter in the first 2 years is
often followed by arrested growth, resulting in aberrant
connectivity. The role of myelination of white matter is
likely to be significant. The goal for scientists is now to pro-
vide further evidence that underconnectivity is indeed pres-
ent with more definitive studies like postmortem
neuroanatomical analyses and to discover the reasons for
the development of this underconnectivity in the hope of
eliminating the causes of this devastating disorder.
References
[1] Hughes JR, Melyn M. EEG and seizures in autistic children and
adolescents: further findings with therapeutic implications. Clin EEG
Neurosci 2005;36:15–20.
[2] Akshoomoff N, Pierce K, Courchesne E. The neurobiological basis of
autism from a developmental perspective. Dev Psychopathol
2002;14(3):613–34.
[3] Hilgetag CC, Barbas H. Role of mechanical factors in the morphol-
ogy of the primate cerebral cortex. PloS Comput Biol 2006;2(3):22.
[4] Baron-Cohen S, Belmonte MK. Autism: a window onto the
development of the social and the analytic brain. Annu Rev Neurosci
2005;28:109–26.
[5] Baron-Cohen S, Knickmeyer RC, Belmonte MK. Sex differences in
the brain: implications for explaining autism. Science
2005;310:819–23.
[6] Lepage JF, Theoret H. EEG evidence for the presence of an action
observation–execution matching system in children. Eur J Neurosci
2006;23:2505–10.
[7] Cordes D, Haughton V, Carew JD, Arfanakis K, Maravilla K.
Hierarchical clustering to measure connectivity in fMRI resting-state
data. Magn Reson Imaging 2002;20:305–17.
[8] Achard S, Salvador R, Whitcher B, Suckling J, Bullmore E. A
resilient, low-frequency, small-world human brain functional network
with highly connected association cortical hubs. Neuroscience
2006;26:63–672.
24 J.R. Hughes / Epilepsy & Behavior 11 (2007) 20–24
[9] Van de Ven VG, Formisano E, Prvulovic D, Roeder CH, Linden DE.
Functional connectivity as revealed by spatial independent compo-
nent analysis of fMRI measurements during rest. Hum Brain Mapp
2004;22:165–78.
[10] Greicius MD, Krasnow B, Reiss AL, Menon V. Functional connec-
tivity in the resting brain: a network analysis of the default mode
hypothesis. Proc Natl Acad Sci USA 2003;22:165–78.
[11] Binder JR, Frost JA, Hammeke TA, Bellgowan PS, Rao SM, Cox
RW. Conceptual processing during the conscious resting state: a
functional MRI study. J Cogn Neurosci 1999;11:80–95.
[12] Jiang T, He Y, Zang Y, Weng X. Modulation of functional
connectivity during the resting state and the motor task. Hum Brain
Mapp 2004;22:63–71.
[13] Waites AB, Stanislavsky A, Abbott DF, Jackson GD. Effect of prior
cognitive state on resting state networks measured with functional
connectivity. Hum Brain Mapp 2005;24:59–68.
[14] Bokde AL, Lopez-Bayo P, Meindl T, et al. Functional connectivity
of the fusiform gyrus during a face-matching task in subjects with
mild cognitive impairment. Brain 2006;129:1113–24.
[15] Courchesne E, Press GA, Yeung-Courchesne R. Parietal lobe
abnormalities with MR in patients with infantile autism. Am J
Roentgenol 1993;160:387–93.
[16] Chung MK, Dalton KM, Alexander AL, Davidson RJ. Less white
matter concentration in autism: D voxel-based morphometry. Neu-
roImage 2004;23:242–51.
[17] Vidal CN, Nicolson R, DeVito TJ, et al. Mapping corpus callosum
deficits in autism: an index of aberrant cortical connectivity. Biol
Psychiatry 2006;60:218–25.
[18] Boger-Megiddo I, Shaw DW, Friedman SD, et al. Corpus callosum
morphometrics in young children with autism spectrum disorder. J
Autism Dev Disord 2006;36:733–9.
[19] Alexander AL, Lee JE, Lazar M, et al. Diffusion tensor imaging of
the corpus callosum in autism. NeuroImage 2007;34:61–73.
[20] Villalobos ME, Mizuno A, Dahl BC, Kemmotsu N, Muller RA.
Reduced functional connectivity between V1 and inferior frontal
cortex associated with visuomotor performance in autism. NeuroIm-
age 2005;25:916–25.
[21] Courchesne E, Pierce K. Why the frontal cortex in autism might be
talking only to itself: local over-connectivity but long-distance
disconnection. Curr Opin Neurobiol 2005;15:225–30.
[22] McAlonan GM, Cheung V, Cheung C, et al. Mapping the brain in
autism: a voxel-based MRI study of volumetric differences and
intercorrelations in autism. Brain 2005;128:268–76.
[23] Casanova MF, van Kooten IA, Switala AE, et al. Minicolumnar
abnormalities in autism. Acta Neuropathol (Berl) 2006;112:287–303.
[24] Kates WR, Burnette CP, Eliez S, et al. Neuroanatomic variation in
monozygotic twin pairs discordant for the narrow phenotype for
autism. Am J Psychiatry 2004;161:539–46.
[25] Waiter GD, Williams JH, Murray AD, Gilchrist A, Perrett DI, Whiten
A. A voxel-based investigation of brain structure in male adolescents
with autistic spectrum disorder. NeuroImage 2004;22:619–25.
[26] Muller RA, Kleinhans N, Kemmotsu N, Pierce K, Courchesne E.
Abnormal variability and distribution of functional maps in autism:
an fMRI study of visuomotor learning. Am J Psychiatry
2003;160:1847–62.
[27] Just MA, Cherkassky VL, Keller TA, Minshew NJ. Cortical
activation and synchronization during sentence comprehension in
high-functioning autism: evidence of underconnectivity. Brain
2004;127:1811–21.
[28] Pierce K, Haist F, Sedaghat F, Courchesne E. The brain response to
personally familiar faces in autism: finding of fusiform activity and
beyond. Brain 2004;127:2703–16.
[29] Silk TJ, Rinehart N, Bradshaw JL, et al. Visuospatial processing and
the function of prefrontal-parietal networks in autism spectrum
disorders: a functional MRI study. Am J Psychiatry 2006;163:1440–3.
[30] Turner KC, Frost L, Linsenbardt D, McIlroy JR, Muller RA.
Atypically diffuse functional connectivity between caudate nuclei and
cerebral cortex in autism. Behav Brain Funct 2006;2:34.
[31] Just MA, Cherkassky VL, Keller TA, Kana RK, Minshew NJ.
Functional and anatomical cortical underconnectivity in autism:
evidence from an fMRI study of an executive function task
and corpus callosum morphometry. Cereb Cortex
2007;17:951–61.
[32] Kana RK, Teller TA, Cherkassky VL, Minshew NJ, Just MA.
Sentence comprehension in autism: thinking in pictures with
decreased functional connectivity. Brain 2006;129:2484–93.
[33] Cherkassky VL, Kana RK, Keller TA, Just MA. Functional
connectivity in a baseline resting-state network in autism. NeuroRe-
port 2006;17:1687–90.
[34] Mizuno A, Villalobos ME, Davies MM, Dahl BC, Muller RA.
Partially enhanced thalamocortical functional connectivity in autism.
Brain Res 2006;1104:160–74.
[35] Castelli F, Frith C, Happe F, Frith U. Autism, Asperger syndrome
and brain mechanisms for the attribution of mental states to
animated shapes. Brain 2002;125:1839–49.
[36] Wilson TW, Rojas DC, Reite ML, Teale PD, Rogers SJ. Children
and adolescents with autism exhibit reduced MEG steady-state
gamma responses. Biol Psychiatry 2007;in press.
[37] Nowinski CV, Minshew NJ, Luna B, Takarae Y, Sweeney JA.
Oculomotor studies of cerebellar function in autism. Psychiatry Res
2005;127:11–9.
[38] Bertone A, Mottron L, Jelenic P, Faubert J. Enhanced and
diminished visuo-spatial information processing in autism depends
on stimulus complexity. Brain 2005;128:2430–41.
[39] Bird G, Catmur C, Silani G, Frith C, Frith U. Attention does not
modulate neural responses to social stimuli in autism spectrum
disorders. NeuroImage 2006;31:1614–24.
[40] Chugani DC, Muzik O, Behen M, et al. Developmental changes in
brain serotonin synthesis capacity in autistic and nonautistic children.
Ann Neurol 1999;45:287–95.
[41] Frith C. What do imaging studies tell us about the neural basis of
autism? Novartis Found Symp 2003;251:149–66.
[42] Brambilla P, Hardan AY, di Nemi SU, et al. The functional
neuroanatomy of autism. Funct Neurol 2004;19:9–17.
[43] Courchesne E. Brain development in autism: early overgrowth
followed by premature arrest of growth. Ment Retard Dev Disabil
Res Rev 2004;10:106–11.
[44] Herbert MR, Ziegler DA, Makris N, et al. Localization of white
matter volume increase in autism and developmental language
disorder. Ann Neurol 2004;55:530–40.
[45] Herbert MR. Large brains in autism: the challenge of pervasive
abnormality. Neuroscientist 2005;11:417–40.
[46] Hendry J, DeVito T, Gelman N, et al. White matter abnormalities in
autism detected through transverse relaxation time imaging. Neuro-
Image 2006;29:1049–57.
[47] Courchesne E, Redcay E, Morgan JT, Kennedy DP. Autism at the
beginning: microstructural and growth abnormalities underlying the
cognitive and behavioral phenotype of autism. Dev Psychopathol
2005;17:577–97.
[48] Pujol J, Lopez-Sala A, Sebastian-Galles N, et al. Delayed myelina-
tion in children with developmental delay detected by volumetric
MRI. NeuroImage 2004;22:897–903.
[49] Haynes RL, Borenstein NS, Desilva TM, et al. Axonal development
in the cerebral white matter of the human fetus and infant. J Comp
Neurol 2005;484:156–67.
[50] Schneider JF, Il’yasov KA, Hennig J, Martin E. Fast quantitative
diffusion-tensor imaging of cerebral white matter from the neonatal
period to adolescence. Neuroradiology 2004;46:258–66.
[51] Zhang L, Thomas KM, Davidson MC, Casey BJ, Heier LA, Ulug
AM. MR quantitation of volume and diffusion changes in the
developing brain. Am J Neuroradiol 2005;26:45–9.
[52] Back SA. Perinatal white matter injury: the changing spectrum of
pathology and emerging insights into pathogenetic mechanisms.
Ment Retard Dev Disabil Res Rev 2006;12:129–40.
[53] Yan H, Rivkees SA. Hypoglycemia influences oligodendrocyte
development and myelin formation. NeuroReport 2006;17:55–9.