Review

Neuroanatomy of autism
David G. Amaral1, Cynthia Mills Schumann2 and Christine Wu Nordahl1
1
The M.I.N.D. Institute, Department of Psychiatry and Behavioral Sciences, University of California, Davis, 2825 50th Street,
Sacramento, CA 95817, USA
2
Department of Neurosciences, University of California, San Diego, 8110 La Jolla Shores Drive, Suite 201, La Jolla, CA 92037, USA

Autism spectrum disorder is a heterogeneous, behavio- an autism that is generally indistinguishable from early-
rally defined, neurodevelopmental disorder that occurs onset autism [7]. The possibility that there is early-onset
in 1 in 150 children. Individuals with autism have deficits versus regressive phenotypes of autism might have import-
in social interaction and verbal and nonverbal communi- ant implications for the types and time courses of neuro-
cation and have restricted or stereotyped patterns of pathology that one might expect to encounter.
behavior. They might also have co-morbid disorders
including intellectual impairment, seizures and anxiety. Where might one expect to see neuropathology?
Postmortem and structural magnetic resonance imaging In Figure 1, we summarize the major brain regions that
studies have highlighted the frontal lobes, amygdala and form the putative neural systems involved in the functions
cerebellum as pathological in autism. However, there is that are most impacted by the core features of autism.
no clear and consistent pathology that has emerged for Several brain regions have been implicated in social beha-
autism. Moreover, recent studies emphasize that the vior through experimental animal studies, lesion studies in
time course of brain development rather than the final human patients or functional imaging studies [8]. These
product is most disturbed in autism. We suggest that the include regions of the frontal lobe, the superior temporal
heterogeneity of both the core and co-morbid features cortex, the parietal cortex and the amygdala. Language
predicts a heterogeneous pattern of neuropathology in function is distributed throughout several cortical and
autism. Defined phenotypes in larger samples of children subcortical regions. Foremost for expressive language
and well-characterized brain tissue will be necessary for function is Broca’s area in the inferior frontal gyrus and
clarification of the neuroanatomy of autism. portions of the supplementary motor cortex. Wernicke’s
area is essential for receptive language function, and the
Introduction superior temporal sulcus plays a role in both language
Autism is a heterogeneous disorder with multiple causes processing and social attention [9]. Finally, the repetitive
and courses, a great range in the severity of symptoms, and or stereotyped behaviors of autism share many similarities
several associated co-morbid disorders. Increasingly, with the abnormal actions of obsessive-compulsive dis-
researchers refer to ‘the autisms’ rather than a single order that implicate regions such as the orbitofrontal
autism phenotype [1]. It would be surprising, therefore, cortex and caudate nucleus [10,11].
if the neuropathology of autism was identical across all The co-morbid disorders of autism are a matter of con-
affected individuals. cern for the interpretation of the neuroanatomy of autism.
Epilepsy, for example, is associated with pathology of the
The core and co-morbid features of autism cerebral cortex, amygdala, cerebellum and hippocampal
As initially described by Kanner [2], individuals with formation, all of which have also been implicated in aut-
autism have three core features: (i) impairments in reci- ism. Unfortunately, the majority of cases evaluated in
procal social interactions; (ii) an abnormal development earlier postmortem studies [12,13] involved autistic brains
and use of language; and (iii) repetitive and ritualized from individuals who had co-morbid seizure disorders.
behaviors and a narrow range of interests. In addition to
the core features of autism, there are common co-morbid Neuroanatomy of autism
neurological disorders [3]. The prevalence of mental retar- Experimental techniques
dation in idiopathic autism is 60% although, when the Structural magnetic resonance imaging (MRI) is a safe,
autism spectrum is taken as a whole, the number is closer relatively noninvasive tool for evaluating gross neuroana-
to 30% [4]. Epilepsy has long been associated with autism tomical changes related to autism. An ideal study would
although estimates of the occurrence of seizure disorder include a very large sample size (i.e. hundreds of subjects)
vary from 5% to 44% [5]. Anxiety and mood disorders are of well-characterized individuals of both genders [14],
also very common in autism [6]. imaged at birth and followed longitudinally at least into
There is also substantial heterogeneity in the onset of late childhood or early adolescence. The ideal study, how-
autism. Some children have signs of developmental delays ever, has not been carried out. Many have been hampered
within the first 18 months of life. However, 25%–40% of by small sample sizes and virtually all are limited to cross-
children with autism initially demonstrate near-normal sectional design (see Ref. [15] for a review). Furthermore,
development until 18–24 months, when they regress into the majority of imaging studies examine populations of
older and higher-functioning individuals. Because the
Corresponding author: Amaral, D.G. (dgamaral@ucdavis.edu). diagnosis of autism cannot reliably be made until around
0166-2236/$ – see front matter ß 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.tins.2007.12.005 Available online 6 February 2008 137
 Review Trends in Neurosciences Vol.31 No.3

Figure 1. Brain areas that have been implicated in the mediation of the three core behaviors that are impaired in autism: social behavior, language and communication, and
repetitive and stereotyped behaviors.

2–3 years of age, the current field of MRI studies might be Differences in total brain volume
defining the end result of the pathology of autism rather Currently, one of the most prominent theories of the
than the etiological changes taking place during onset. neuropathology of autism is that the brain undergoes a
Whereas MRI provides a reliable method for studying period of precocious growth during early postnatal life
gross neuropathology in a large number of subjects over followed by a deceleration in age-related growth [16].
time, postmortem techniques provide a tool for under- The evidence for this early overgrowth comes from four
standing the underlying neurobiology of observed neuroa- studies of head circumference, a proxy for brain size, that
natomical abnormalities (e.g. if the brain is larger, are provide evidence for normal or smaller head circumference
there more neurons, fibers, axons, glia or synapses?). For at birth followed by an increase in the rate of growth
these studies, controlling for confounding factors by beginning at 12 months of age [16–19]. Existing MRI
excluding or segregating co-morbid conditions is essential. studies suggest that very young children with autism (ages
The use of postmortem techniques as a tool for studying the 18 months to 4 years) have a 5%–10% abnormal enlarge-
neuroanatomy of autism is still very much in its infancy, ment in total brain volume [19–21], but whether this
with fewer than 100 autism cases studied to date and a enlargement persists into later childhood and adolescence
mean sample size of 5 autism cases per study. is not as clear [20,22–26] (Figure 2).

138
 Review Trends in Neurosciences Vol.31 No.3

Figure 2. Percent difference between autism or autism spectrum groups and typical development with best-fit curves for (a) total brain volume, (b) gray matter and (c) white
matter, based on existing MRI literature. Only studies with direct statistical comparisons between control and autism cases are included. Differences that are statistically
significant ( p < .05) are depicted with a black outline around the colored circle. Earlier studies with samples that were later included in larger studies are not included. For
comparison purposes, when possible, only data from males are represented. With the exception of Sparks 2002 [21] and Schumann 2004 [25], total brain volume measures
include brainstem and cerebellum. For Hazlett 2005 [19], percent difference between autism and typically developing controls (not developmentally delayed controls) is
represented. For Schumann 2004 [25], percent difference between high-functioning autism (not low-functioning or Asperger’s syndrome) and typically developing controls
is represented. For Hazlett 2005 [19], the total gray matter difference is significant for left hemisphere only. For Hazlett 2006 [28], cerebral white matter, not total white
matter, is significantly larger than typically developing controls. Aylward 2002 [22]; Courchesne 2001 [20]; Hardan 2001 [46]; Hazlett 2005 [19]; Hazlett 2006 [28]; Herbert 2003
[23]; Lotspeich 2004 [27]; Palmen 2005 [24]; Piven 1996 [26]; Schumann 2004 [25]; Sparks 2002 [21].

Collectively, head circumference and MRI studies of observed in cross-sectional studies does not necessarily
total brain volume indicate a putative period of abnormal imply accelerated growth. The observation of presumed
brain growth beginning in the first year of life that results precocious brain growth awaits confirmation by a longi-
in a persistent enlargement at least through early child- tudinal MRI study. Furthermore, if the rate of brain
hood. It is important to emphasize that brain enlargement growth is indeed accelerated, additional studies will be

139
Review Trends in Neurosciences Vol.31 No.3

needed to address whether there is an increased rate of sulcus, intraparietal sulcus and inferior frontal gyrus
normal neurodevelopmental processes or entirely abnor- [38,39]. Of the two studies on cortical thickness, one
mal processes unique to the disorder. reported increased cortical thickness over the entire
cerebral cortex, primarily driven by increases in parietal
Do increases in brain size equally involve gray and white and temporal cortices [40] in 8- to 12-year-olds, whereas
matter? the other study reported cortical thinning in frontal, par-
Another prominent theory, postulated by Herbert et al. ietal and temporal regions in adults [41].
[23], is that the abnormal brain enlargement observed in
children with autism is disproportionately accounted for by Postmortem cortical neuropathology
increased white matter, not gray matter. Indeed, two As discussed above, MRI studies have consistently found
studies of very young children (1.5–4 years) [19,20] show increases in brain size in younger children with autism
greater increases in white matter than gray matter followed by an abnormal growth pattern through adoles-
(Figure 2b,c). But again, whether these increases persist cence. What does this tell us about the neuropathology of
into later childhood and adolescence is less clear [24,27,28]. autism? If the brain is larger, are there too many neurons,
Existing evidence on gray matter volume suggests that glia, synapses and so forth? If the difference in brain size
although gray matter enlargement might be proportion- does not persist into adulthood, what neuropathological
ately smaller than white matter enlargement early in life, underpinnings account for this phenomenon of an abnor-
the enlargement might persist into adulthood (Figure 2b). mal growth trajectory? With the availability of more abun-
Four studies, collectively spanning early childhood dant, high-quality postmortem tissue and employing
through adolescence and adulthood, report 6%–12% enlar- modern neuroanatomical techniques such as stereological
gement of gray matter [20,24,27,28]. Additional imaging methods for counting neurons and in situ hybridization for
techniques, such as spectroscopy and T2 relaxometry, that evaluating expression levels of genes, answering these
can examine microstructural aspects of gray and white questions might be possible in the future.
matter, will be important to further explore underlying To date, only a few postmortem studies have been
etiologies of gray and white matter abnormalities. published on the cortical neuropathology of autism, and
One MRI technique that holds great promise in inves- quantitative studies are sparser yet. Besides early case
tigating white matter integrity is diffusion tensor imaging studies, Kemper and Bauman [13] were the first to carry
(DTI). Fractional anisotropy (FA) is a measure that reflects out a qualitative neuropathological investigation in a
the degree of isotropic movement of water in a given voxel; study that included six cases of autism (5/6 with mental
higher FA values reflect denser or more ordered brain retardation, 4/6 with seizures). The only area of consistent
structure. Two studies in older children and adults show abnormality in the cerebral cortex they identified was the
reductions in FA in cerebral white matter [29,30], with anterior cingulate cortex that appeared unusually coarse
consistent reductions reported in and near the genu of the and poorly laminated. Bailey et al. [12] qualitatively
corpus callosum. A recent DTI and volumetric study of the examined six cases of autism with mental retardation,
corpus callosum found a 14% reduction in corpus callosum four of which had seizures, compared to seven controls.
size was associated with reduced FA in the genu and Cortical dysgenesis was observed in four of the six autism
splenium [31]. However, the one DTI study of toddlers cases, including increased cortical thickness, high
[32] with autism reported differences in FA and other neuronal density, neurons present in the molecular layer,
diffusion-weighted variables, indicating an increase in and irregular laminar patterns. Ectopic gray matter and
white matter maturity. an increased number of neurons in the white matter were
observed in three of the autism cases.
Is there any regional specificity to gray and white matter
differences? Alterations of the columnar structure of the neocortex:
Although enlargements in gray and white matter have the minicolumn hypothesis
been reported in frontal, temporal, and parietal lobes, Increasing interest has been placed on the notion,
the largest and most consistent increases have been advanced by Casanova and colleagues [42–44], that there
reported in the frontal lobes [24,28,33,34]. Within the are an abnormal number and width of minicolumns (Box 1)
frontal lobes, no consistent pattern of regional specificity in individuals with autism. Only 14 cases of autism (9 of
has emerged. There is some evidence for greater increases which had seizures and at least 10 with mental retar-
in dorsolateral prefrontal and medial frontal cortex [34,35] dation) have been examined for minicolumnar pathology
but less consistent findings of no difference or decreases in in cortical layer III in three independent studies using
orbitofrontal cortex [36,37]. Some of these inconsistencies varying techniques [42–44]. The most consistent finding in
might be a result of differences in the definition and these studies is reduced intercolumnar width of the mini-
delineation of these frontal cortical regions. A perusal of columns (only layer III has been studied thus far) in
this literature emphasizes the need for the field of devel- dorsolateral prefrontal cortex or Brodmann’s area (BA)
opmental neuropathology to establish a systematic 9. These findings, coupled with increases in neuronal
approach to evaluating abnormal brain development. density on the order of 23% noted by Casanova et al.
Several MRI studies have examined other aspects of the [44], imply that there should be a greater number of
cerebral cortex, such as sulcal patterns, cortical shape and neurons in BA 9 of the autistic cortex. Given the narrower
cortical thickness. Abnormalities in cortical shape have neuropil area between columns, one would also predict a
been identified in the sylvian fissure, superior temporal decrease in the dendritic arborization of BA 9 neurons.

140
Review Trends in Neurosciences Vol.31 No.3

Box 1. Brain development and minicolumns

For over a century, neuroanatomists have remarked on the columnar each is a representative Golgi-stained section showing the extent of
structure of the neocortex [70]. The smallest column has come to be dendritic growth in this same cortical area over these same ages.
called the ‘minicolumn’ or ‘microcolumn’ [71,72]. The minicolumn Within the first year of life, there is a dramatic increase in dendritic
can be identified by the stacking of neuronal cell bodies, particularly growth. By 2 years of age, the minicolumns are spaced farther apart
in layers III and V of the neocortex (Figure Ie, adapted from [73]). with a lower cell density in a given region of cortex. Dendritic bundles
Mountcastle proposed that minicolumns are the basic functional unit and axonal fasicles that extend throughout several layers of the cortex
of the brain [74], although the significance of the vertical organization occupy the space between minicolumns [70–72].
of neurons has been a topic of much debate [71]. Casanova and colleagues [42–44] have posed the reasonable
Minicolumn formation has been associated with early stages of question of whether there is perturbation in the fundamental
cortical development when postmitotic neurons ascend in linear organization of minicolumns in the autistic brain. Preliminary data
arrays along a radial glial scaffolding [75]. Figure Ia–c depicts cell described in the text indicate aberrant columnar structure in layer III
body-stained sections of BA 9 at 1, 6 and 24 months of age [76]. Below with less space between cell body-defined minicolumns (Figure Id).

Figure I. Features of neocortical organization potentially altered in autism.

141
Review Trends in Neurosciences Vol.31 No.3

These neuropathological questions are ripe for analysis expression of GAD67 mRNA in the cerebellar Purkinje
using systematic stereological methods. cells of individuals with autism [53]. Whether the obser-
vations of lower Purkinje cell number and density actually
Neuropathology of the cerebellum reflect fewer Purkinje cells in the autistic brain awaits
Five MRI studies of individuals with autism at a wide confirmation with stereological neuron counting methods.
range of ages have found the cerebellum to be enlarged Although the postmortem finding of fewer Purkinje
relative to controls [45]. However, this increase in cerebel- cells appears to stand in stark contrast to the MRI finding
lar volume is generally proportional to total brain volume, of an enlarged cerebellum in autism, several factors
with one exception in which the difference exceeded that of make the two findings impossible to compare. Twenty-
the cerebrum [46]. Only one published study to date has two of the 24 brains examined in postmortem studies
examined children younger than age 3, and this study did came from individuals who also had mental retardation
not find a difference in cerebellar size between the children [12,13,51,52]. Almost half of the brains were from indi-
at risk for autism and typically developing controls [19]. viduals with epilepsy and some individuals who were
In contrast to the total cerebellum, the size of the vermis taking anticonvulsive medications that might themselves
appears to be slightly smaller in some individuals with damage Purkinje cells. By contrast, most of the MRI
autism [47,48]. All but one study on the size of the vermis studies were conducted with high-functioning individuals
report areal measurements from a single, midsagittal sec- with autism and typically excluded subjects with seizure
tion per subject, a measure that can be difficult to make disorders from the study. So, two very different cohorts
reliably. Results from the same laboratory indicate that of subjects are being studied with these different
the autistic vermis can either be smaller or larger than techniques.
controls [48], and size differences are likely because of the
heterogeneity in the autism phenotype [49]. In addition, Neuropathology of the amygdala
cerebellar vermal hypoplasia is not specific to autism, but The amygdala (Figure 3) in boys with autism appears to
commonly found in various developmental and psychiatric undergo an abnormal developmental time course that
disorders and/or mental retardation [47,50]. includes a period of precocious enlargement that persists
The cerebellum has also been of substantial interest in through late childhood [21,25]. Sparks et al. [21] found a
postmortem studies of autism. Of the 24 postmortem 13%–16% abnormal enlargement of the amygdala in young
cases of autism reported in the literature in which the children with autism (36–56 months of age). Recent studies
cerebellum was studied, 19 (or 79%) show decreased suggest that amygdala enlargement is associated with
density of Purkinje cells, particularly in the hemispheres more severe anxiety [54] and worse social and communi-
[12,13,51,52]. Interestingly, a recent study found 40% less cation skills [55].

Figure 3. Neuroanatomy of the human amygdala. (a) Lateral view of a three-dimensional reconstruction of an MRI (dashed line represents location of coronal slice in (b)),
(b) MRI coronal image with amygdala outlined (red), (c) coronal section of brain tissue (box around amygdala) and (d) Nissl-stained section of amygdala nuclei.
PAC = periamygdaloid cortex.

142
Review Trends in Neurosciences Vol.31 No.3

Schumann et al. [25] examined the amygdala in 8- to Where do we go from here?

18-year-old boys and also found evidence of an altered Given the enormous heterogeneity in the manifestation of
developmental time course. The amygdala was enlarged core features of autism spectrum disorders, varied occur-
by 15% in 8- to 12-year-old boys with autism relative to rence of significant co-morbid syndromes and generally
typically developing controls, but did not differ in 13- to small sample sizes of both MRI and postmortem studies
18-year-old boys. Whereas the amygdala in typically devel- of autism, it is remarkable that any significant group
oping boys increased in size by 40% from 8 to 18 years of differences have been detected. It would not be surprising,
age, the same growth trajectory did not occur in the boys therefore, if more clear-cut pathology emerges once distinct
with autism. Although these findings await confirmation phenotypes of the disorder are considered as factors in the
from a longitudinal study, the amygdala appears initially analyses.
to be larger than normal in children with autism, but does Attempts at defining the neuroanatomy of autism are
not undergo the same preadolescent age-related increase obviously in their infancy. The field will benefit both from
in volume that takes place in typically developing boys. the application of classical techniques such as the Golgi
Studies focused primarily on older adolescents, adults or a method as well as modern molecular neuroanatomical
wide age range of subjects have found no difference [56] or procedures such as in situ hybridization and single-cell
even smaller [57–59] amygdala volumes in individuals PCR. It is imperative that more systematic and quantitat-
with autism relative to age-matched controls. ive studies become the norm rather than the exception. All
Kemper and Bauman [13] were the first to report of these studies will rely, however, on the acquisition and
abnormalities in the microscopic organization of the amyg- sharing of higher-quality and a higher number of post-
dala. Qualitative observations in six postmortem cases of mortem brain specimens. In this regard, consolidated
autism ages 9–29 years (5/6 with mental retardation, 4/6 national efforts such as the Autism Tissue Program should
with seizure disorder) indicated that neurons in certain be supported.
nuclei of the amygdala of autism cases appeared unusually It is also important that noninvasive imaging tech-
small and more densely packed than in age-matched con- niques be applied to larger populations of better-pheno-
trols. Schumann and Amaral [60] carried out a design- typed individuals. Imaging studies that start at the
based stereological study to estimate the number and size earliest possible age and are longitudinal rather than
of neurons in the amygdala (Figure 3) in nine autism cases cross-sectional should be encouraged. Because autism is
10–44 years of age without seizure disorder compared to likely to involve both abnormal brain structure and con-
ten typically developing age-matched male controls. The nections, techniques that noninvasively probe brain con-
autism group had significantly fewer neurons in the total nectivity and function should also be promoted. One has
amygdala and in the lateral nucleus than the controls. the sense that the real secrets of the neuropathology of
They did not find increased neuronal density or decreased autism have yet to be uncovered. This makes the search
size of neurons as Kemper and Bauman [13] had reported. even more critical and exciting.
If the decreased number of neurons in the amygdala is
found to be a reliable characteristic of autism, what might Acknowledgements
account for this finding? Two possible hypotheses are: (i) The authors have been supported by NIH grants R37 MH057502 and
fewer neurons were generated during early development, MH41479. C.W.N. was a Fellow of the Autism Research Training
or (ii) a normal or even excessive number of neurons was Program (T32 MH07134). We would like to thank the following
colleagues who read earlier versions of this manuscript and provided
generated initially, which would be consistent with MRI
helpful advice: Margaret Bauman, Gene Blatt, Manuel Casanova, Joseph
findings of a larger amygdala in early childhood, but some Piven, Sally Ozonoff and two anonymous reviewers.
of these have subsequently been eliminated during adult-
hood. Unfortunately, there is currently no evidence to
support or reject either of these possibilities. References
1 Geschwind, D.H. and Levitt, P. (2007) Autism spectrum disorders:
developmental disconnection syndromes. Curr. Opin. Neurobiol. 17,
Other brain regions 103–111
The investigation of other brain regions that might be 2 Kanner, L. (1943) Autistic disturbances of affective contact. Nervous
affected in autism (Figure 1) is limited. There is little Child 2, 217–250
MRI evidence and no postmortem evidence for abnormal- 3 DiCicco-Bloom, E. et al. (2006) The developmental neurobiology of
autism spectrum disorder. J. Neurosci. 26, 6897–6906
ities in the thalamus [61–63] and basal ganglia [64–67].
4 Fombonne, E. (2006) Past and future perspectives on autism
There is evidence for enlargement of the caudate nucleus epidemiology. In Understanding Autism from Basic Neuroscience to
that might be correlated with the presence of repetitive Treatment (Moldin, S.O. and Rubenstein, J.L.R., eds), pp. 25–48,
and ritualistic behaviors in adolescents and young adults Taylor and Francis
[64,65]. There is also evidence, although somewhat incon- 5 Tuchman, R. and Rapin, I. (2002) Epilepsy in autism. Lancet Neurol. 1,
352–358
sistent (see Ref. [45] for a review), for abnormalities within
6 Lecavalier, L. (2006) Behavioral and emotional problems in young
the hippocampus, both in volume [25] and more recently in people with pervasive developmental disorders: relative prevalence,
shape [68,69]. In postmortem studies, Kemper and Bau- effects of subject characteristics, and empirical classification.
man [13] reported increased cell packing density and J. Autism Dev. Disord. 36, 1101–1114
smaller neurons in the hippocampus in all autism cases 7 Werner, E. and Dawson, G. (2005) Validation of the phenomenon of
autistic regression using home videotapes. Arch. Gen. Psychiatry 62,
examined. However, this finding was not replicated by 889–895
Bailey et al. [12], who reported increased cell packing 8 Adolphs, R. (2001) The neurobiology of social cognition. Curr. Opin.
density in only one of the five cases examined. Neurobiol. 11, 231–239

143
Review Trends in Neurosciences Vol.31 No.3

9 Redcay, E. (2008) The superior temporal sulcus performs a common 37 Hardan, A.Y. et al. (2006) Magnetic resonance imaging study of the
function for social and speech perception: implications for the orbitofrontal cortex in autism. J. Child Neurol. 21, 866–871
emergence of autism. Neurosci. Biobehav. Rev. 32, 123–142 38 Levitt, J.G. et al. (2003) Cortical sulcal maps in autism. Cereb. Cortex
10 Atmaca, M. et al. (2007) Volumetric MRI study of key brain 13, 728–735
regions implicated in obsessive-compulsive disorder. Prog. 39 Nordahl, C.W. et al. (2007) Cortical folding abnormalities in autism
Neuropsychopharmacol. Biol. Psychiatry 31, 46–52 revealed by surface-based morphometry. J. Neurosci. 27, 11725–11735
11 Whiteside, S.P. et al. (2004) A meta-analysis of functional 40 Hardan, A.Y. et al. (2006) An MRI study of increased cortical thickness
neuroimaging in obsessive-compulsive disorder. Psychiatry Res. 132, in autism. Am. J. Psychiatry 163, 1290–1292
69–79 41 Hadjikhani, N. et al. (2005) Anatomical differences in the mirror
12 Bailey, A. et al. (1998) A clinicopathological study of autism. Brain 121, neuron system and social cognition network in autism. Cereb. Cortex
889–905 16, 1276–1282
13 Kemper, T.L. and Bauman, M.L. (1993) The contribution of 42 Buxhoeveden, D.P. et al. (2006) Reduced minicolumns in the frontal
neuropathologic studies to the understanding of autism. Neurol. cortex of patients with autism. Neuropathol. Appl. Neurobiol. 32,
Clin. 11, 175–187 483–491
14 Giedd, J.N. et al. (1999) Brain development during childhood and 43 Casanova, M.F. et al. (2002) Minicolumnar pathology in autism.
adolescence: a longitudinal MRI study. Nat. Neurosci. 2, 861–863 Neurology 58, 428–432
15 Brambilla, P. et al. (2003) Brain anatomy and development in autism: 44 Casanova, M.F. et al. (2006) Minicolumnar abnormalities in autism.
review of structural MRI studies. Brain Res. Bull. 61, 557–569 Acta Neuropathol. 112, 287–303
16 Courchesne, E. et al. (2003) Evidence of brain overgrowth in the first 45 Minshew, N.J. et al. (2005) Neurologic aspects of autism. In Handbook
year of life in autism. JAMA 290, 337–344 of Autism and Pervasive Developmental Disorders (Vol. 1) (Volkmar,
17 Dawson, G. et al. (2007) Rate of head growth decelerates and F.R. et al., eds), In pp. 473–514, John Wiley & Sons
symptoms worsen in the second year of life in autism. Biol. 46 Hardan, A.Y. et al. (2001) Posterior fossa magnetic resonance imaging
Psychiatry 61, 458–464 in autism. J. Am. Acad. Child Adolesc. Psychiatry 40, 666–672
18 Dementieva, Y.A. et al. (2005) Accelerated head growth in early 47 Kaufmann, W.E. et al. (2003) Specificity of cerebellar vermian
development of individuals with autism. Pediatr. Neurol. 32, abnormalities in autism: a quantitative magnetic resonance imaging
102–108 study. J. Child Neurol. 18, 463–470
19 Hazlett, H.C. et al. (2005) Magnetic resonance imaging and head 48 Courchesne, E. et al. (1994) Cerebellar hypoplasia and hyperplasia in
circumference study of brain size in autism: birth through age 2 infantile autism. Lancet 343, 63–64
years. Arch. Gen. Psychiatry 62, 1366–1376 49 Piven, J. et al. (1997) An MRI study of autism: the cerebellum revisited.
20 Courchesne, E. et al. (2001) Unusual brain growth patterns in early Neurology 49, 546–551
life in patients with autistic disorder: an MRI study. Neurology 57, 50 Okugawa, G. et al. (2003) Smaller cerebellar vermis but not
245–254 hemisphere volumes in patients with chronic schizophrenia. Am. J.
21 Sparks, B.F. et al. (2002) Brain structural abnormalities in young Psychiatry 160, 1614–1617
children with autism spectrum disorder. Neurology 59, 184–192 51 Palmen, S.J. et al. (2004) Neuropathological findings in autism. Brain
22 Aylward, E.H. et al. (2002) Effects of age on brain volume and head 127, 2572–2583
circumference in autism. Neurology 59, 175–183 52 Ritvo, E.R. et al. (1986) Lower Purkinje cell counts in the cerebella of
23 Herbert, M.R. et al. (2003) Dissociations of cerebral cortex, subcortical four autistic subjects: initial findings of the UCLA-NSAC Autopsy
and cerebral white matter volumes in autistic boys. Brain 126, Research Report. Am. J. Psychiatry 143, 862–866
1182–1192 53 Yip, J. et al. (2007) Decreased GAD67 mRNA levels in cerebellar
24 Palmen, S.J. et al. (2005) Increased gray-matter volume in medication- Purkinje cells in autism: pathophysiological implications. Acta
naive high-functioning children with autism spectrum disorder. Neuropathol. 113, 559–568
Psychol. Med. 35, 561–570 54 Juranek, J. et al. (2006) Association between amygdala volume and
25 Schumann, C.M. et al. (2004) The amygdala is enlarged in children but anxiety level: magnetic resonance imaging (MRI) study in autistic
not adolescents with autism; the hippocampus is enlarged at all ages. children. J. Child Neurol. 21, 1051–1058
J. Neurosci. 24, 6392–6401 55 Munson, J. et al. (2006) Amygdalar volume and behavioral
26 Piven, J. et al. (1996) Regional brain enlargement in autism: a development in autism. Arch. Gen. Psychiatry 63, 686–693
magnetic resonance imaging study. J. Am. Acad. Child Adolesc. 56 Haznedar, M.M. et al. (2000) Limbic circuitry in patients with autism
Psychiatry 35, 530–536 spectrum disorders studied with positron emission tomography and
27 Lotspeich, L.J. et al. (2004) Investigation of neuroanatomical magnetic resonance imaging. Am. J. Psychiatry 157, 1994–2001
differences between autism and Asperger syndrome. Arch. Gen. 57 Aylward, E.H. et al. (1999) MRI volumes of amygdala and hippocampus
Psychiatry 61, 291–298 in non-mentally retarded autistic adolescents and adults. Neurology
28 Hazlett, H.C. et al. (2006) Cortical gray and white brain tissue volume 53, 2145–2150
in adolescents and adults with autism. Biol. Psychiatry 59, 1–6 58 Nacewicz, B.M. et al. (2006) Amygdala volume and nonverbal social
29 Barnea-Goraly, N. et al. (2004) White matter structure in autism: impairment in adolescent and adult males with autism. Arch. Gen.
preliminary evidence from diffusion tensor imaging. Biol. Psychiatry Psychiatry 63, 1417–1428
55, 323–326 59 Pierce, K. et al. (2001) Face processing occurs outside the fusiform
30 Keller, T.A. et al. (2007) A developmental study of the structural ‘face area’ in autism: evidence from functional MRI. Brain 124,
integrity of white matter in autism. Neuroreport 18, 23–27 2059–2073
31 Alexander, A.L. et al. (2007) Diffusion tensor imaging of the corpus 60 Schumann, C.M. and Amaral, D.G. (2006) Stereological analysis of
callosum in autism. Neuroimage 34, 61–73 amygdala neuron number in autism. J. Neurosci. 26, 7674–7679
32 Ben Bashat, D. et al. (2007) Accelerated maturation of white matter in 61 Hardan, A.Y. et al. (2006) Abnormal brain size effect on the thalamus in
young children with autism: a high b value DWI study. Neuroimage 37, autism. Psychiatry Res. 147, 145–151
40–47 62 Haznedar, M.M. et al. (2006) Volumetric analysis and three-
33 Carper, R.A. et al. (2002) Cerebral lobes in autism: early hyperplasia dimensional glucose metabolic mapping of the striatum and
and abnormal age effects. Neuroimage 16, 1038–1051 thalamus in patients with autism spectrum disorders. Am. J.
34 Herbert, M.R. et al. (2004) Localization of white matter volume Psychiatry 163, 1252–1263
increase in autism and developmental language disorder. Ann. 63 Tsatsanis, K.D. et al. (2003) Reduced thalamic volume in high-
Neurol. 55, 530–540 functioning individuals with autism. Biol. Psychiatry 53, 121–129
35 Carper, R.A. and Courchesne, E. (2005) Localized enlargement of the 64 Hollander, E. et al. (2005) Striatal volume on magnetic resonance
frontal cortex in early autism. Biol. Psychiatry 57, 126–133 imaging and repetitive behaviors in autism. Biol. Psychiatry 58,
36 Girgis, R.R. et al. (2007) Volumetric alterations of the orbitofrontal 226–232
cortex in autism. Prog. Neuropsychopharmacol. Biol. Psychiatry 31, 65 Sears, L.L. et al. (1999) An MRI study of the basal ganglia in autism.
41–45 Prog. Neuropsychopharmacol. Biol. Psychiatry 23, 613–624

144
Review Trends in Neurosciences Vol.31 No.3

66 Hardan, A.Y. et al. (2003) Motor performance and anatomic magnetic 71 Jones, E.G. (2000) Microcolumns in the cerebral cortex. Proc. Natl.
resonance imaging (MRI) of the basal ganglia in autism. J. Child Acad. Sci. U. S. A. 97, 5019–5021
Neurol. 18, 317–324 72 Rockland, K.S. and Ichinohe, N. (2004) Some thoughts on cortical
67 Langen, M. et al. (2007) Caudate nucleus is enlarged in high-functioning minicolumns. Exp. Brain Res. 158, 265–277
medication-naive subjects with autism. Biol. Psychiatry 62, 262–266 73 Lübke, J. and Feldmeyer, D. (2007) Excitatory signal flow and
68 Dager, S.R. et al. (2007) Shape mapping of the hippocampus in young connectivity in a cortical column: focus on barrel cortex. Brain
children with autism spectrum disorder. AJNR Am. J. Neuroradiol. 28, Struct. Funct. 212, 3–17
672–677 74 Mountcastle, V.B. (1997) The columnar organization of the neocortex.
69 Nicolson, R. et al. (2006) Detection and mapping of hippocampal Brain 120, 701–722
abnormalities in autism. Psychiatry Res. 148, 11–21 75 Rakic, P. (1988) Specification of cerebral cortical areas. Science 241,
70 DeFelipe, J. (2005) Reflections on the structure of the cortical 170–176
minicolumn. In Neocortical Modularity and the Cell Minicolumn 76 Conel, J. (1939) The Postnatal Development of the Human Cerebral
(Casanova, M., ed.), pp. 57–92, Nova Science Publishers Cortex. Harvard University Press

Endeavour
The quarterly magazine for the history and
philosophy of science.

You can access Endeavour online on

ScienceDirect, where you’ll find book reviews,
editorial comment and a collection of beautifully
illustrated articles on the history of science.

Featuring:

Information revolution: William Chambers, the publishing pioneer by A. Fyfe

Does history count? by K. Anderson
Waking up to shell shock: psychiatry in the US military during World War II by H. Pols
Deserts on the sea floor: Edward Forbes and his azoic hypothesis for a lifeless deep ocean by T.R. Anderson and T. Rice
‘Higher, always higher’: technology, the military and aviation medicine during the age of the two world wars by C. Kehrt
Bully for Apatosaurus by P. Brinkman

Coming soon:

Environmentalism out of the Industrial Revolution by C. Macleod

Pandemic in print: the spread of influenza in the Fin de Siècle by J. Mussell
Earthquake theories in the early modern period by F. Willmoth
Science in fiction - attempts to make a science out of literary criticism by J. Adams
The birth of botanical Drosophila by S. Leonelli

And much, much more. . .

Endeavour is available on ScienceDirect, www.sciencedirect.com

145