Brain connectivity and high functioning autism: A promising path of research that needs refined models, methodological convergence, and stronger behavioral links

R

Bn

Ma

b

c

a

ARRA

KAARBfDEMRD

C

T

0d

Neuroscience and Biobehavioral Reviews 36 (2012) 604–625

Contents lists available at SciVerse ScienceDirect

Neuroscience and Biobehavioral Reviews

journa l h o me pa g e: www.elsev ier .com/ locate /neubiorev

eview

rain connectivity and high functioning autism: A promising path of research thateeds refined models, methodological convergence, and stronger behavioral links

arlies E. Vissersa,b, Michael X Cohena,c, Hilde M. Geurtsa,b,∗

Department of Psychology, Universiteit van Amsterdam, Weesperplein 4, 1018 XA Amsterdam, The NetherlandsDr Leo Kannerhuis, Amsterdam Clinic, Amsterdam, Paasheuvelweg 39-D, 1105 BG Amsterdam, The NetherlandsDepartment of Physiology, University of Arizona Arizona Health Sciences Center, 1501N. Campbell, Rm. 4104, PO Box 245051, Tucson, AZ 85724, United States

r t i c l e i n f o

rticle history:eceived 21 January 2011eceived in revised form 7 September 2011ccepted 16 September 2011

eywords:utismSDeview

a b s t r a c t

Here we review findings from studies investigating functional and structural brain connectivity inhigh functioning individuals with autism spectrum disorders (ASDs). The dominant theory regardingbrain connectivity in people with ASD is that there is long distance under-connectivity and local over-connectivity of the frontal cortex. Consistent with this theory, long-range cortico-cortical functional andstructural connectivity appears to be weaker in people with ASD than in controls. However, in con-trast to the theory, there is less evidence for local over-connectivity of the frontal cortex. Moreover,some patterns of abnormal functional connectivity in ASD are not captured by current theoretical mod-els. Taken together, empirical findings measuring different forms of connectivity demonstrate complex

rain connectivityMRITIEGEG

esting state

patterns of abnormal connectivity in people with ASD. The frequently suggested pattern of long-rangeunder-connectivity and local over-connectivity is in need of refinement.

© 2011 Elsevier Ltd. All rights reserved.

MN

ontents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6052. What do we know so far? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605

2.1. Current theories about brain (dis-)organization in ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6052.2. Long-range versus local connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605

2.2.1. The need for a widely accepted, uniform definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6052.2.2. The definition used in the current study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606

2.3. The way connectivity research may shed new light on findings from traditional MRI research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6062.4. Objective of the current review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606

3. Literature search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6064. Brain connectivity: What does it mean and how is it studied? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6075. Measures of ASD characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6086. Methods and results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608

6.1. Findings from task-related studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6086.1.1. Many studies report reduced long-range connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6086.1.2. Exceptions: Mixed results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
6.1.3. The relationship between functional connectivity during6.1.4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2. Findings from resting state studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author at: Brain and Cognition, Department of Psychology, Universiteiel.: +31 205256843; fax: +31 20639165.

E-mail address: [email protected] (H.M. Geurts).

149-7634/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.neubiorev.2011.09.003

task performance, behavior, and clinical symptoms . . . . . . . . . . . . . . . . 608 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609

t van Amsterdam, Weesperplein 4, 1018 XA Amsterdam, The Netherlands.

dx.doi.org/10.1016/j.neubiorev.2011.09.003

http://www.sciencedirect.com/science/journal/01497634

http://www.elsevier.com/locate/neubiorev

mailto:[email protected]

dx.doi.org/10.1016/j.neubiorev.2011.09.003

M.E. Vissers et al. / Neuroscience and Biobehavioral Reviews 36 (2012) 604–625 605

6.2.1. Many studies report reduced connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6096.2.2. Studies reporting mixed patterns of under- and over-connectivity during rest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6096.2.3. The relationship between resting state functional connectivity and clinical symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6096.2.4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615

6.3. Findings from DTI studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6156.3.1. Many studies report reduced structural connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6156.3.2. Exceptions: Reports of over-connectivity in ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6156.3.3. The relationship between white matter microstructure, behavior, and clinical symptoms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6196.3.4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619

6.4. Findings from EEG and MEG studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6196.4.1. Divergent findings of over- and under-connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6196.4.2. The relationship between functional connectivity measured with EEG or MEG, and clinical symptoms . . . . . . . . . . . . . . . . . . . . . . . . 6216.4.3. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621

7. What can we conclude about connectivity in ASD? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6217.1. The relationship between structural and functional connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6217.2. Do findings from different types of studies indicate similar abnormalities in connectivity for people with ASD? . . . . . . . . . . . . . . . . . . . . . . . . 6217.3. Do the findings support current theories about connectivity in ASD? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6227.4. Do different techniques show similar relationships between brain connectivity and behavior in ASD? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622

8. Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623. . . . . .

1

mit2AsTahAaMS

roit2odptbt

2

2

tiswlnobw

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. Introduction

Autistic spectrum disorders (ASDs) are neurobiological develop-ental disorders that are primarily characterized by impairments

n social and communicative skills, and by repetitive and stereo-ypical behaviors (American Psychiatric Association [APA], 1994,000). ASD includes a group of disorders such as autistic disorder,sperger syndrome, high functioning autism (HFA), and perva-ive developmental disorders not otherwise specified (PDD-NOS).he notion that ASD is a disorder of brain development is widelycknowledged, and an increasing number of brain imaging studiesave been conducted to investigate the neurobiological origin ofSD. Many of these studies implicate a collection of brain regionss being functionally or anatomically abnormal in ASD (e.g., Diartino et al., 2009; McAlonan et al., 2005; Schmitz et al., 2006;

tanfield et al., 2008).Most brain imaging studies of ASD focus on specific brain

egions, such as the amygdala (e.g., Kleinhans et al., 2009) or therbitofrontal cortex (e.g., Girgis et al., 2007). However, emerg-ng theories and empirical data implicate network activity, ratherhan specific regions, as being dysfunctional in autism (Just et al.,004). The purpose of this review is to overview recent studiesn functional and structural connectivity in ASD, as an attempt toraw a general conclusion on abnormalities in brain connectivity ineople with ASD. We will compare findings of structural and func-ional brain network abnormalities in ASD, discuss the relationshipetween brain connectivity and ASD symptoms, and suggest poten-ially important lines of future research.

. What do we know so far?

.1. Current theories about brain (dis-)organization in ASD

Development of MRI analysis techniques addressing connec-ivity across brain regions has led to a new branch of researchn autism. From studies applying such techniques it appears thattructural and functional brain connectivity is different in peopleith ASD compared to matched controls. These observations shed

ight on autism as a disorder of deficient integration and synchro-
ization of brain regions (Just et al., 2004). Since the introductionf the notion of disturbed connectivity in ASD, various studies haveeen conducted in order to investigate brain connectivity in peopleith ASD.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623

It has been proposed that in ASD, the brain is characterized bylocal over-connectivity and long distance under-connectivity of thefrontal cortex (Courchesne and Pierce, 2005). Initially, this the-ory was based on findings from a positron emission tomography(PET) study in ASD (Horwitz et al., 1988), postmortem microscopicstudies (e.g., Vargas et al., 2005), and anatomical and functionalMRI studies (e.g., Carper and Courchesne, 2005; Courchesne et al.,2001; Pierce et al., 2004). However, when the theory was postu-lated, only a handful of MRI studies on brain connectivity in peoplewith ASD were conducted (Barnea-Goraly et al., 2004; Just et al.,2004; Koshino et al., 2005). At present the number of studies onbrain connectivity in individuals with ASD has grown, as well asthe variety of techniques used to assess brain connectivity, creatingthe possibility to evaluate the theory.

Among the dominant views on the behavioral deficits in ASD isthe weak central coherence account (Happé and Frith, 2006), whichdescribes disturbed information processing an integration in ASD.Connections between brain regions have been proposed to underlieinformation integration, in order to realize behavior that is in accor-dance with environmental demands (Sporns et al., 2000). Based onthis notion, under-connectivity in ASD has been suggested to bea plausible neuroanatomical analogy of the pattern of behavioraldeficits in ASD described by the weak central coherence account(Just et al., 2004). Further, findings from microscopic studies thatreport atypical development of aspects of the frontal cortex, suchas the increased presence of minicolumns (Casanova et al., 2002)and enlarged frontal grey and white matter in children with ASD(Carper and Courchesne, 2005), gave rise to the notion of localover-connectivity in frontal regions in ASD (Courchesne and Pierce,2005). At present, it is unclear whether more recent findings fromconnectivity research support the notion of local over-connectivityand long-range under-connectivity in ASD.

2.2. Long-range versus local connectivity

2.2.1. The need for a widely accepted, uniform definitionAlthough the distinction between local and long-range con-

nectivity is often used in ASD research, agreement on whatdistinguishes local from long-range is lacking (Wilson et al., 2007).While some have described local connectivity as connectivity
among subregions of the frontal lobe (e.g., Courchesne and Pierce,2005), others have used the term to describe connectivity in localnetworks regardless of their location (Brock et al., 2002). Further,some have focused on short-range connections (U-fibers) to study

6 Biobeh

lLrt(ostwnrctca

2

siccbsrciftf

rlficbiet2

brrmwip

foaieap

2fi

atSvs

06 M.E. Vissers et al. / Neuroscience and

ocal connectivity (e.g., Shukla et al., 2011b; Sundaram et al., 2008).ong-range connectivity has been defined as connectivity betweenegions across different lobes (Courchesne and Pierce, 2005), or asemporal binding between local networks irrespective of locationBelmonte et al., 2004). It is important to recognize that definitionsf local and long-range connectivity are often not provided in mosttudies or are not explicitly mentioned to the reader. Moreover,he definitions that are used are not uniform across studies. Yet,ithout the use of an equal definition of local and long-range con-ectivity, results are difficult to compare across studies. For futureesearch, a coherent and explicit definition of local and long-rangeonnectivity in research on brain connectivity in ASD is essen-ial. Such a generally accepted definition will simplify and improveomparison of findings across different studies, in order to develop

better understanding of brain connectivity in ASD.

.2.2. The definition used in the current studyIn this review, we define long-range and local connectivity irre-

pective of lobe or other major brain subdivisions. This definitions broader than the initially proposed distinction of local over-onnectivity and long distance under-connectivity of the frontalortex (Courchesne and Pierce, 2005). However, the literature onrain connectivity in ASD also focuses on connectivity patterns out-ide frontal cortex, including other cortical areas and subcorticalegions. Therefore, we adopt a terminology that better facilitates aomparison of findings from the literature, as it will include all find-ngs without biasing the outcome towards a specific region (e.g.,rontal cortex). This terminology also allows us to examine whetherhe literature in general supports the initial theory focused on therontal cortex.

Local connectivity is defined as connectivity within a brainegion of approximately one cubic centimeter. This definition ofocal connectivity will include short range association fibers (Ubers), although we do not intend to equate the two. Biologi-ally, connectivity on this level may reflect neural communicationetween adjacent populations of neurons (Fries, 2005). One man-

festation of local connectivity that may be disturbed in ASD, forxample, is lateral inhibition, the process by which neural func-ional groups inhibit other neural functional groups (Casanova,006).

Long-range connectivity is defined as connectivity betweenrain regions that are separated by more than one centimeter. Longange connections represent interactions between distant brainegions that are likely to play a large role in perception and infor-ation processing (Sporns et al., 2000), which is atypical in peopleith ASD. Deficient long-range connectivity may be the neurolog-

cal basis for impaired information processing and integration ineople with ASD (Just et al., 2004).

The particular distinction made in this review was maderom a pragmatic point of view, as it distinguishes two typesf connectivity that may have different biological mechanismsnd consequences for behavior. However, we emphasize themportance of a more widely accepted definition across differ-nt connectivity studies in the field, which will make comparisonscross different studies easier and, thereby, will lead to strongerrogression of an understanding of brain connectivity in ASD.

.3. The way connectivity research may shed new light onndings from traditional MRI research

By “traditional”, we refer to standard MRI analyses in which mass-univariate analysis is applied in attempt to localize func-
ional or structural differences in discrete anatomical brain regions.tructural MRI (sMRI) studies in ASD have shown that theolume of the total brain, and regions such as the cerebral hemi-pheres, the cerebellum, the caudate nucleus, and the corpus
avioral Reviews 36 (2012) 604–625

callosum, is different in people with ASD than in healthy people (fora review, see Stanfield et al., 2008). Functional MRI (fMRI) studiesshow that recruitment of brain regions is often atypical in peoplewith ASD (for a review, see Williams and Minshew, 2007). This hasbeen shown for several domains that are known to be affected byASD, such as social interaction (for reviews, see Di Martino et al.,2009; Perkins et al., 2010), face-processing (e.g., Critchley et al.,2000; Schultz et al., 2000), imitation (e.g., Williams et al., 2006),and executive function (e.g., Schmitz et al., 2006).

However, due to their strong focus on spatial localizationand magnitude of activation, these traditional approaches do notreveal information about the temporal dynamics of brain activa-tion patterns, or the way spatially distant regions may form unifiednetworks. The fact that neural information processing is inherentlylinked to time stresses the importance of temporal dynamics ofbrain functioning (Cohen, 2011a). Further, cognition likely arisesfrom a complex interplay between different regions in the brain,and is not the result of activity in one specific region (Just et al.,2004; Sporns et al., 2004). To address the interaction between brainregions and the way this may fluctuate over time, research onbrain-connectivity is potentially more informative, as it providesinformation about ways brain regions or populations of neuronsco-activate or interact over time. Connectivity research could linkanatomical characteristics and functional characteristics of thebrain, which are typically investigated in isolation in traditionalfMRI studies (Sporns et al., 2000). For example, structural connec-tions show by which pathways brain regions may communicate,which is informative with respect to the functional specialization ofregions (Johansen-Berg and Rushworth, 2009). Information abouthow the brain is organized may also explain abnormalities in brainactivation patterns found in traditional fMRI research. Thus, con-nectivity research in ASD generates valuable additional knowledgeabout brain functioning in people with ASD.

2.4. Objective of the current review

Due to the broad range of techniques and designs that have beenused in studies on connectivity in ASD, direct comparison acrossstudies is difficult. Furthermore, many studies address either localor long-range connectivity—but not both—in ASD. In the currentpaper we will provide a systematic review of 72 studies on struc-tural connectivity, task-related functional connectivity, and restingstate functional connectivity in people with ASD. Most of the stud-ies we review used MRI, although some EEG and MEG studies alsomet our inclusion criteria. The underlying theme of this reviewis to evaluate whether findings on connectivity in ASD supportthe account of local over-connectivity and long distance under-connectivity (e.g., Brock et al., 2002; Courchesne and Pierce, 2005),and to emphasize the importance of a common definition of localand long-range connectivity to be used in future research in thefield.

3. Literature search

Studies adhering to the following criteria were incorporated inthis review: (1) ASD participants without mental retardation, whomet diagnostic criteria as set forth in DSM-III-R, DSM-IV, or ICD-10,and for whom diagnoses were confirmed with standardized diag-nostic instruments, were the population under study; (2) a typicallydeveloping control group was included; (3) MRI, EEG or MEG wereused to investigate connectivity; (4) empirical findings on brain
connectivity were reported by the authors; (5) studies were pub-lished before June 1, 2011; (6) studies were published in an Englishpeer-reviewed journal. We found 72 studies located in PubMed thatmet these criteria. Search terms related to autism, connectivity,

Biobeh

Mn

4s

tctFsa

di2aptttostsbgd2

miWafsasuridyobnt

iipcascadc

pnaet

M.E. Vissers et al. / Neuroscience and

RI, EEG, MEG, and specific networks of interest (e.g., default-modeetwork), were used.

. Brain connectivity: What does it mean and how is ittudied?

Brain connectivity can be investigated by looking at similarity ofemporal characteristics of brain activity in multiple regions, alsoalled functional connectivity, or by looking at physical connec-ions between regions, also referred to as structural connectivity.or both types of connectivity, there is evidence that these mea-ures reflect means by which regions communicate (Johansen-Bergnd Rushworth, 2009; Sporns et al., 2000).

Brain regions show temporal correlations of activation patternsuring active processes such as perception and cognitive process-

ng (e.g., Sporns et al., 2000), but also during rest (Buckner et al.,008). Task related functional connectivity reveals informationbout what networks of brain regions are recruited in order torocess and integrate information, and to respond adequately toask demands. Resting-state functional connectivity is studied inhe absence of external stimulation: people are typically instructedo close their eyes and think of nothing in particular for a periodf 5–10 min (e.g., Assaf et al., 2010; Weng et al., 2010). Restingtate functional connectivity is useful to investigate (intrinsic) func-ional networks, for example the default-mode network (DMN, alsoometimes called the task-negative network), a constellation ofrain regions including the medial frontal cortex, posterior cin-ulate, and medial temporal areas, that become coherently activeuring task-irrelevant, undirected mental states (Buckner et al.,008).

In fMRI research, functional connectivity during task perfor-ance and during rest is often measured by correlated fluctuations

n the blood-oxygen-level-dependent (BOLD)-signal over time.hen the time courses of the activation patterns significantly devi-

te from statistical independence, these regions are assumed to beunctionally connected (Sporns et al., 2000). In most studies, analy-es are restricted to activation patterns in regions of interest (ROI),lthough some studies have applied more data-driven approachesuch as independent component analysis (Assaf et al., 2010), orsed self-organizing maps (Wiggins et al., 2011) in order to revealelevant regions. When studying the time courses of brain regionsn networks, factor analyses can provide meaningful indices of theegree of coordination of activity within a network. In factor anal-ses, ROIs within a network are grouped according to the similarityf the time course of their activation. It is assumed that the num-er of factors needed to explain the time course of activation in theetwork is inversely related to the degree of coordination withinhe network (e.g., Just et al., 2007).

Recently, regional homogeneity (ReHo) analyses have beenntroduced as a technique to study local patterns of connectivityn the brain (Zang et al., 2004). ReHo analyses investigate the tem-oral stability of the BOLD-response among neighboring voxels byorrelating their BOLD-time series. Thus, it is comparable to thenalysis of functional connectivity described above, but compari-on of time courses occurs on a finer spatial resolution (voxel-wiseomparisons instead of ROI-wise comparisons). These methods tossess functional connectivity can be applied to investigate stateependent functional connectivity regardless of the subjects’ spe-ific state (rest or active task performance).

Next to these state-dependent fluctuations in activationatterns in the brain, state-independent, interregional synchro-
ization of spontaneous low-frequency (<0.1 Hz) oscillations maylso reveal information about the way brain regions interact (Bluhmt al., 2007). Spontaneous low-frequency oscillations are thoughto be a meaningful aspect of the BOLD signal, and have been shown
avioral Reviews 36 (2012) 604–625 607

to be related to behavioral and neuroanatomical measures (for areview, see Fox and Raichle, 2007). When synchronization of low-frequency oscillations is investigated, data are usually low-passedfiltered at 0.1 Hz to regress possible task or condition related effectsout (e.g., Shih et al., 2010). However, in fractal analysis of the lowfrequency component of the BOLD-signal, low-pass filtering of thedata is not required (see Lai et al., 2010).

With EEG and MEG, functional connectivity can be measuredin a number of ways; predominant among them is examiningphase- or spectral-coherence between oscillations in distant brainregions. Electrical activation in the brain occurs in different fre-quency bands (from slow to fast, delta (<4 Hz); theta; alpha; beta;gamma (>30 Hz)). These frequency bands may selectively showsynchronization (i.e., phase locking) of activation between loca-tions. Moreover, phase locking in particular frequency bands maybe specific for different forms of connectivity. It has been proposedthat local synchronization is reflected by increased phase-lockingof activity in the gamma frequency band, whereas long-range con-nectivity is most clearly observable in the alpha, beta and thetafrequency bands (Von Stein and Sarnthein, 2000).

Structural connectivity in the living human brain can be inves-tigated using diffusion tensor magnetic resonance imaging (DTI),which provides a measure of the directional dependency of waterdiffusion in white matter fibers in the brain, called anisotropicdiffusion. The assumption is that more directionally dependentwater flow is indicative of the presence of more axons runningthrough the underlying white matter, and of greater structuralconnectivity.

The measure that is typically provided in DTI research is frac-tional anisotropy (FA), which indicates the directional dependencyof water diffusion in the brain. FA values are normalized, withvalues closer to 1 indicating highly directionally dependent waterflow. Other values that are commonly reported in DTI studies areaxial diffusivity, radial diffusivity, mean diffusivity, and apparentdiffusion coefficient. These measures may aid interpretation of themeaning of abnormal FA values and may reveal factors underlyingabnormalities in FA, such as diffusion along particular axes of thetensor (Alexander et al., 2007).

However, mere reports of measures such as FA values do notprovide us with information about differences between local andlong-range connectivity, as FA values may reflect either smallregions of white matter surrounded by grey matter, or white mat-ter in long fiber tracts connecting areas that are several centimetersapart. Thus, it is difficult to dissociate long- and short-range con-nectivity from FA values.

Probabilistic tractography can also be applied on DTI data, andmay be more informative with respect to local and long-rangeconnectivity. This method enables the prediction of trajectoriesof white matter pathways based on directionality of water diffu-sion in voxels. It provides an indication of the likely presence andorientation of axonal fibers running through particular regions inthe brain. DTI reveals patterns of connectivity similar to the pat-terns found in traditional, highly accurate tract-tracing methods(Johansen-Berg and Rushworth, 2009). Recently, the use of tract-based spatial statistics has increased in DTI research (e.g., Shuklaet al., 2011a; Weinstein et al., 2011). It allows for investigationof white matter microstructure in specific white matter pathwaysselectively (Smith et al., 2006).

It is important to note that findings from the methods discussedhere do not provide information about effective connectivity, i.e.,the direction of a connection. In order to obtain information aboutthe directionality of effects, models such as Granger causality(Granger, 1969) or structural equation modeling (SEM; Bollen,1989) are required. Until now, only a handful of studies applied
these models on brain connectivity data in people with ASD (e.g.,Shih et al., 2010).

6 Biobeh

5

bos

6

6

6

tubTn

imccKSbefec2a2

eieascKlwc

6

ianr(nb22

boco2im


. Measures of ASD characteristics

In this review we will also discuss findings on the relationshipetween brain connectivity and symptoms of ASD. A descriptionf the different measures that were used to assess ASD symptomeverity is provided in Table 1.

. Methods and results

.1. Findings from task-related studies

.1.1. Many studies report reduced long-range connectivityAn overview of the studies on functional connectivity during

ask performance can be found in Table 2. When using the termnder-connectivity, we refer to relatively reduced synchronizationetween brain regions in one group compared to another group.he term over-connectivity refers to relatively increased synchro-ization between regions in one group compared to another group.

During performance of a wide range of cognitive tasks, includ-ng those measuring cognitive control, visual attention, language,

emory, and theory of mind, long-range connectivity betweenortical regions was found to be reduced in individuals with ASDompared to controls (e.g., Agam et al., 2010; Just et al., 2004, 2007;ana et al., 2006, 2009; Kleinhans et al., 2008; Mason et al., 2008;olomon et al., 2009). Aberrant connectivity was mostly foundetween frontal and parietal regions (Damarla et al., 2010; Justt al., 2007; Kana et al., 2006; Liu et al., 2011), but also amongrontal, temporal, and occipital areas (Kana et al., 2009; Solomont al., 2009). Further, connectivity among subregions of the frontalortex was frequently found to be abnormal (e.g., Koshino et al.,008; Liu et al., 2011), often between the anterior cingulate cortexnd other frontal brain regions (e.g., Agam et al., 2010; Kana et al.,007).

Findings indicate that brain activity within cortical networks,specially in frontal and parietal cortices, is poorly coordinated inndividuals with ASD compared to healthy controls (e.g., Koshinot al., 2005; Solomon et al., 2009). Factor analyses showed thatctivity in frontal and parietal regions in ASD was grouped in moreub-networks and that those networks function in isolation whenompared to controls (e.g., Just et al., 2007; Kana et al., 2007;oshino et al., 2008). In line with this finding, synchronization of

ow frequency oscillations between different networks of regionsas found to be less differentiated in people with ASD than in

ontrols (Shih et al., 2011).

.1.2. Exceptions: Mixed resultsAlthough the majority of the studies on functional connectivity

ndicate under-connectivity in individuals with ASD, some report mixed pattern of both under- and over-connectivity. Local con-ectivity in ASD was shown to be reduced in frontal and parietalegions, but increased in temporal and parahippocampal regionsShukla et al., 2010b). Increased long-range connectivity, accompa-ied by reduced connectivity elsewhere in the brain, was observedetween the thalamus and several cortical regions (Mizuno et al.,006), and between the caudate and the motor cortex (Turner et al.,006) in people with ASD.

Three other studies exclusively report over-connectivityetween brain regions in ASD. However, the regions in whichver-connectivity was found are diverse, and concern task relatedhanges in the BOLD-signal as well as spontaneous low-frequency
scillations (Noonan et al., 2009; Shih et al., 2010; Welchew et al.,005). In general, these findings of over-connectivity have been
nterpreted to reflect diffuse and less specialized, rather thanore efficient, connectivity. Two studies report no differences in


connectivity in the brain of individuals with ASD compared tohealthy controls (Brieber et al., 2010; Lee et al., 2009).

6.1.3. The relationship between functional connectivity duringtask performance, behavior, and clinical symptoms

Findings on the relationship between task performance andaberrant functional connectivity in ASD show that there exist rela-tionships between brain functioning and cognitive processes inASD. Several studies that show reduced functional connectivitybetween brain regions in ASD also report poor task performancein ASD compared to controls (e.g., Agam et al., 2010; Just et al.,2007; Mostofsky et al., 2009). For example, people with ASD werefound to make more errors on an executive function task, and func-tional connectivity between fronto-parietal regions was found tobe reduced during task performance (Just et al., 2007). However, insome domains poor performance in ASD was found to be accom-panied by a more diffuse pattern of increased as well as reducedfunctional connectivity between regions, for example in visuomo-tor tasks (Mizuno et al., 2006; Mostofsky et al., 2009; Turner et al.,2006; Villalobos et al., 2005). Notably, reduced functional connec-tivity does not always lead to poorer task performance: for tasksconcerning imagery, memory, and visual perception, several stud-ies report reduced functional connectivity between brain regionsin ASD in the absence of atypical task performance by this group(Damarla et al., 2010; Kana et al., 2006, 2009; Koshino et al., 2005,2008; Liu et al., 2011). Further, reduced functional connectivity inseveral brain regions has been observed to be linked to lower IQlevels (Noonan et al., 2009; Shukla et al., 2010b).

In studies in which correlations between task performance andfunctional connectivity were assessed directly, functional connec-tivity between brain regions was found not to correlate to reactiontimes and error rates on a saccadic paradigm and cognitive controltask in people with ASD (Agam et al., 2010; Solomon et al., 2009).

In sum, the presence of observations of reduced (e.g., Just et al.,2007) as well as increased FC accompanying poor task perfor-mance in ASD (e.g., Noonan et al., 2009; Welchew et al., 2005),together with findings of reduced FC in ASD without an observ-able effect on task performance, indicate that there is no uniformeffect of abnormal brain connectivity on cognitive functioning.Rather, this relationship seems to be task- and network-specific.It is clear, however, that recruitment of functional networks duringtask performance is organized differently in people with ASD thanin healthy people.

For symptom severity, there is evidence for a relationshipbetween functional connectivity during task performance andsymptom severity on various autism diagnostic instruments.Social impairment measures (ADI-R) show a negative correla-tion with functional connectivity between various regions (e.g.,frontal–parietal regions, Just et al., 2007; inferior frontal cortex andpresupplementary motor area, Lee et al., 2009), such that weakerconnectivity was associated with more severe social impairments.Contrarily, functional connectivity between the right fusiform facearea and the left amygdala, and between the fusiform face area andthe right inferior frontal gyrus, correlated positively with the sever-ity of social impairments (ADI-R; ADOS; Kleinhans et al., 2008). Theseverity of repetitive, restricted behaviors shows positive corre-lations with functional connectivity between the frontal eye fieldand the dorsal anterior cingulate cortex in adults with ASD (ADI-R;Agam et al., 2010), indicating that stronger functional connectivitybetween these regions is associated with more repetitive behaviors.

These findings suggest that for some characteristics of ASD,
their severity is related to the extent to which functional connec-tivity during task performance is abnormal compared to controls.However, findings seem to be specific to different subscales of thediagnostic instruments and to regions under study.

M.E. Vissers et al. / Neuroscience and Biobehavioral Reviews 36 (2012) 604–625 609

Table 1Diagnostic instruments to assess ASD.

Instrument Description

Autism Diagnostic Interview – Revised(ADI-R; Lord et al., 1994)

The ADI-R is diagnostic instrument in the form of an interview, administrated by a trained clinician. The interview isadministrated from someone who knew the individual suspected of autism well during childhood. The interview covers thedevelopmental history of the individual. It addresses the child’s social and communication skills, social development andplay, the presence of repetitive, restricted behaviors, and general behavioral problems. In practice, higher scores areinterpreted to reflect higher severity of autism symptoms.

Autism Diagnostic ObservationSchedule (Generic) (ADOS; ADOS-G;Lord et al., 1989, 2000)

The ADOS-G is a diagnostic instrument for ASD. It has subscales that address social interaction, communication, play, andimaginative use of materials in children and adults that are suspected of ASD. Higher scores are indicative of the presence ofautism.

Social Responsiveness Scale (SRS;Constantino and Todd, 2005)

The SRS is a quantitative measure of autistic social impairment, its items concern reciprocal social behavior. This scale is nota diagnostic instrument; it can be completed by spouses or caretakers of the person of interest. Higher scores are indicativeof a more autism characteristics.

Childhood Autism Rating Scale (CARS;Schopler et al., 1988)

The CARS is a diagnostic instrument for ASD. It is composed of fifteen subscales on which the child’s behavior is rated (e.g.,emotional response, verbal communication). Scales are to be completed by clinicians or parents. Higher scores are indicativeof increased presence of autism symptoms.

rumenntal df autis

6

piKSc

tafFav(musicdS

crcffoTafh

6

6

iptiItfic

Gilliam Autism Rating Scale (GARS;Gilliam, 1995)

The GARS is an autism rating instsocial interaction, and developmeindicative of increased presence o

.1.4. DiscussionThe majority of studies on functional connectivity during task

erformance report long-range under-connectivity in the brain ofndividuals with ASD (e.g., Agam et al., 2010; Just et al., 2007;ana et al., 2006, 2009; Kleinhans et al., 2008; Mason et al., 2008;olomon et al., 2009), predominantly between frontal and otherortical regions, and between regions located in the frontal lobe.

There is some evidence for over-connectivity and for mixed pat-erns of over- and under-connectivity in ASD, which likely indicates

more diffuse network organization. Notably, these findings comerom studies in which data were analyzed in a nonstandard way.or example, two studies that report mixed patterns of reducednd increased functional connectivity during performance of aisuomotor task focused on subcortical as well as cortical regionsMizuno et al., 2006; Turner et al., 2006). This focus is unlike the

ajority of studies, which typically address cortical areas only andse tasks that rely more on frontal cortical areas. Further, othertudies reporting mixed findings of under- and over-connectivityn ASD investigated interregional correlations in task independentomponents of the BOLD-signal, rather than task-related time-omain covariance patterns (Noonan et al., 2009; Shih et al., 2010;hukla et al., 2010b).

In sum, the brain of individuals with ASD seems to beharacterized by long-range under-connectivity between corticalegions, primarily between the frontal cortex and other corti-al regions. However, there is suggestive evidence for increasedunctional connectivity between cortical and subcortical regions,or aberrant synchronization of the low-frequency hemodynamicscillations, and for abnormal local connectivity in people with ASD.he correlations between functional connectivity and behavioralbnormalities in ASD suggest that network connectivity is relevantor the expression of the disorder, although the findings reviewedere do not reveal how exactly these two are related.

.2. Findings from resting state studies

.2.1. Many studies report reduced connectivityWe use the same definitions of over- and under-connectivity as

n Section 6.1.1. When synchronization between a set of regions isresent in one group, but absent in another group, we will mentionhis explicitly. It should be noted that the resting state describedn Section 4 does not apply to one of the studies discussed here.
n particular, Cherkassky et al. (2006) examined connectivity ofhe default-mode network in individuals with ASD during lengthyxation trials (approximately 24 s) in between tasks. It is notlear whether the psychological state during long fixation trials is
t for children. It consists of four subscales (stereotyped behavior, communication,isturbances), which are to be filled out by clinicians or parents. Higher scores arem symptoms.

comparable to the typically induced resting state, which lastslonger and is not interspersed with tasks (Monk et al., 2009).

An overview of the studies on resting state functional connectiv-ity can be found in Table 3. The majority of studies on resting stateconnectivity indicate that the default-mode network is topographi-cally comparable in individuals with ASD and controls: for the mostpart it comprises similar regions (Assaf et al., 2010; Cherkasskyet al., 2006; although findings of a default-mode network compris-ing fewer regions in people with ASD have also been reported, seeKennedy and Courchesne, 2008). However, the regions within thedefault-mode network appear to be more weakly connected in indi-viduals with ASD compared to controls (Anderson et al., 2011; Assafet al., 2010; Cherkassky et al., 2006; Weng et al., 2010; Wigginset al., 2011). Studies on resting state connectivity have also reportedunder-connectivity among regions implicated in social and emo-tional processing, such as connectivity between the insula andregions involved in social and emotional processing (Ebisch et al.,2010). Endogenous low-frequency oscillations in various corticaland subcortical regions were found to be less correlated during restin individuals with ASD than in typical individuals (Lai et al., 2010).

6.2.2. Studies reporting mixed patterns of under- andover-connectivity during rest

Several other studies report both under-connectivity combinedwith over-connectivity in ASD between brain regions during rest(Monk et al., 2009; Paakki et al., 2010), or with atypical patterns ofconnectivity not seen in controls (Di Martino et al., 2011). Rest-ing state connectivity in individuals with ASD was found to beincreased between the posterior cingulate cortex, the right tempo-ral lobe and the right parahippocampal gyrus, but reduced betweenthe posterior cingulate cortex and the right superior frontal gyrus(Monk et al., 2009). Functional connectivity between striatal andcortical regions was observed between areas that showed no func-tional link in controls. These connections were found to be strongerand more diffuse in individuals with ASD (Di Martino et al., 2011).Also, local connectivity, predominantly located in the right hemi-sphere, was found to be characterized by a diffuse pattern ofover- and under-connectivity in individuals with ASD (Paakki et al.,2010).

6.2.3. The relationship between resting state functionalconnectivity and clinical symptoms

Increased presence of repetitive, restricted behaviors was found
to correlate positively with resting state functional connectivitybetween the posterior cingulate cortex and the right parahip-pocampal gyrus in adults with ASD (ADI-R; Monk et al., 2009), butnegatively with resting state functional connectivity between the

610

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625Table 2Functional connectivity (FC) during task performance.

Study Participants (nr, age and SD ofage)

Analysis Paradigm Results – local connectivity Results – long-rangeconnections

Results – relationship withbehavior

Just et al. (2004) 17 ASD age n/a (adults)17 TD age n/a (adults)

Typicala Sentence comprehension task ↓ FC between left frontaland temporal regions

• ASD: ↓ RT and, trend for ↑error rate

Koshino et al. (2005) 14 ASD 25.7 yrs (n/a)15 TD 29.8 yrs (n/a)

Typical n-back task (letters) ↓ FC, especially betweenfrontal regions• Right lateralization offrontal–parietal network(left in controls)

• No group difference forperformance

Villalobos et al. (2005) 8 ASD 28.4 yrs (8.9)8 TD 33.6 yrs (5.6)

Typical Visuomotor task ↓ FC between V1 andbilateral inferior frontalregions, right SFG, and PCL↓ FC in bilateral THA andright BG

• ASD: ↑ error rate

Welchew et al. (2005) 13 ASD 31.2 yrs (9.1)13 TD 25.6 yrs (5.1)

3-way MDS Facial affect processing-task ↑ FC between AMG, HC andPHG with the rest of thebrain

• ASD: ↓ explicitrecognition of fearful faces

Kana et al. (2006) 12 ASD 22.5 yrs (8.8)13 TD 20.3 yrs

Typical Imagery task ↓ FC between frontal andparietal regions

• No group difference inperformance

Mizuno et al. (2006) 8 ASD 28.4 yrs (n/a)8 TD 28.1 yrs (n/a)

Typical Visuomotor task • Temporal and occipital FCseen in controls is absent↑ Thalamo-cortical FC


Turner et al. (2006) 8 ASD 28.1 yrs (8.3)8 TD 28.6 yrs (7.2)

Typical Visuomotor task ↓ FC between the caudateand frontal, parietal andoccipital lobes↑ FC between bilateralcaudate and motor regions(diffuse pattern)


Just et al. (2007) 18 HFA 27.1 yrs (11.9)18 NC 24.5 yrs (9.9)

Typical Executive function task ↓ FC between frontal andparietal regions

• ASD: ↑ RT on difficulttrials• ASD: negative correlationbetween frontal–parietalFC and measure of autismsymptom severity (ADOS)

Kana et al. (2007) 12 ASD 26.8 yrs (7.7)12 TD 22.5 yrs (3.2)

Typical Go/No Go task ↓ FC between ACG, MCG,and INS, with right inferiorfrontal, middle frontal, andright inferior parietalregions

• ASD: ↑ effect of memoryload on error rate than TD

Kleinhans et al. (2008) 19 ASD 23.5 yrs (7.8)21 TD 25.1 yrs (7.6)

Typical 1-back task (houses and faces) ↓ FC between right FFA andseveral regions↓ Nr of regions in networkfunctionally connected toright FFA during faceprocessing

• ASD: correlation betweenFC between right FFA andleft AMG and measure ofsocial impairment (ADI-R)• ASD: correlation betweenFC between right FFA andright IFG and measure ofsocial impairment (ADOS)

Koshino et al. (2008) 11 HFA 24.5 yrs (10.2)11 TD 28.7 yrs (10.9)

Typical n-back task (faces) ↓ FC between left frontalregions↓ Size of frontal parietalnetwork

• No group difference intask performance

Mason et al. (2008) 18 ASD 26.5 yrs (n/a)18 TD 27.4 yrs (n/a)

Typical Causal inference drawing ↓ FC between left medialfrontal and right superiortemporal areas↓ FC left IFG and MTG, withleft medial frontal and rightsuperior temporal areas

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625

611

Kana et al. (2009) 12 ASD 24.6 yrs (6.9)12 TD 24.4 yrs (3.7)

Typical Mental state attribution ↓ FC between medialfrontal and orbitofrontalregions, with right MTGand STG, and with TPJ↓ FC of frontal withtemporal, parietal andoccipital regions↓ FC between temporal andoccipital regions

• No group difference intask performance• ASD: ↓ FC for viewingToM processing; TD: ↑ FCfor ToM processing

Lee et al. (2009) 12 ASD 10.17 yrs (1.57)12 TD 11.01 yrs (1.78)

Typical Go/No Go task – • ASD: negative correlationbetween left and right IFCand pSMA, and measure ofsocial impairment (ADI-R)

Mostofsky et al. (2009) 13 ASD 10.9 yrs (1.5)13 TD 10.5 yrs (1.4)

Typical Finger tapping task ↓ FC within motor circuits ↑ Errors for left hand motortask

Noonan et al. (2009) 10 ASD 23 yrs (9.9)10 TD 25.8 yrs (9.8)

Synchronization of LFB,task effects regressed out

Source memory task ↑ FC of LFB • ASD: ↓ recognition• ASD: negative correlationbetween FC between rightSTL and left MO seed andFS-IQ

Solomon et al. (2009) 22 ASD 15.2 yrs (1.7)23 TD 16.0 yrs (2.0)

Typical Cognitive control task ↓ FC between frontal,parietal and occipitalregions

• TD: negative correlationbetween error rate and FCbetween BA9 and SPC onincompatible responsetrials (not in ASD)• ASD: negative correlationbetween FC between leftBA9 and left BA40 andmeasure of ADHDsymptoms (DSM-IV)

Agam et al. (2010) 11 ASD 28 yrs (10)14 TD 27 yrs (8)

Typical Saccadic paradigm ↓ FC between FEF and dACC • ASD: ↑ antisaccade errors• ASD and TD: nocorrelation between dACCand FEF FC and error rateor latency• ASD: correlation betweenFC between FEF and leftdACC and measure of RRB(ADI-R)

Brieber et al. (2010) 15 ASD 16.42 yrs (2.28)15 TD 15.35 yrs (1.80)(14/group for FC analysis)

Typical Perceptual decision task – • No group difference intask performance

Damarla et al. (2010) 13 ASD 19 yrs (5.5)13 TD 22.1 yrs (4.25)

Typical Embedded figures task ↓ FC between frontal andposterior regions• ASD: correlation betweensize of CC and FC betweenfrontal and posteriorregions

• No group difference forperformance

Jones et al. (2010) 17 ASD 16.1 yrs (2.6)20 TD 17.1 yrs (2.1)


Verbal fluency task ↓ FC between LFB,strongest in frontal regionsand during rest-trials

Shih et al. (2010) 14 ASD 24.1 yrs (9.5)14 TD 24.2 yrs (8.4)


Semantic and letter decision tasks ↑ FC of LFB between theIFG, inferior parietal lobule,STS and frontal regions(frontal regions notinvolved for TD group)

612

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625

Table 2 (Continued)


Analysis Paradigm Results – local connectivity Results – long-rangeconnections

Results – relationship withbehavior

Shukla et al. (2010b) 26 ASD 13.7 yrs (0.6)29 TD 13.8 yrs (0.55)

Regional homogeneityanalyses, task effectsregressed out

Visual search task ↓ FC in superior parietaland anterior prefrontalregions↑ FC in temporal regionsand in the PHG

• TD: negative correlationbetween local FC in regionsthat show ↑ local FC inASD, and verbal IQ

Liu et al. (2011) 15 ASD 25.2 yrs (7.6)15 TD 26.3 yrs (8.2)

Typical Processing of 3D visual stimuli ↓ FC between medialfrontal and posteriorregions during local 3Dprocessing↓ FC between medialfrontal and other frontalregions during local 3Dprocessing

• No significant groupdifference for performance

Shih et al. (2011) 21 ASD 14.3 yrs (2.9)26 TD 14.3 yrs (2.8)


↑ FC between differentfunctional networks(rostral, mid, and caudal) ofposterior STS↓ Differentiation of spatialand temporal FC patternsbetween subregions of theSTS• No relationship betweenanatomical characteristicsof STS and FC (relationshippresent in TD)

• Correlation betweenmeasure of social andcommunicativeimpairments (ADOS), and ↓segregation of functionalnetworks

TD: typically developing; FC: functional connectivity; RT: response times; V1: primary visual cortex; SFG; superior frontal gyrus; PCL: paracentral lobule; THA: thalamus; BG: basal ganglia; MDS, multi-dimensional scaling; AMG:amygdala; HC: hippocampus; PHG: parahippocampal gyrus; 3D: 3-dimensional; ACG: anterior cingulate gyrus; MCG: middle cingulate gyrus; INS: insula; FFA: fusiform face area; IFG: inferior frontal gyrus; STS: superior temporalsulcus; MTG: middle temporal gyrus; STG: superior temporal gyrus; TPJ: temporoparietal junction; ToM; theory of mind; IFC: inferior frontal cortex; LFB: low-frequency BOLD-oscillatory activity; STL: superior temporal lobe;pSMA: presupplementary motor area; MO: middle occipital; FS-IQ: full-scale IQ; BA: Broca’s area; SPC: superior parietal cortex; dACC: dorsal anterior cingulate cortex; FEF: frontal eye field; RRB: restricted, repetitive behavior;and CC: corpus callosum.Note. If not otherwise stated, results are reported for participants with ASD when compared to the control group.

a Correlations between task-related changes in BOLD signal in different regions.

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625

613

Table 3Functional connectivity (FC) during resting state.

Study Participants (nr, age and SDof age)

Analysisa Results – local connectivity Results – long-range connections Results – relationship with behavior

Cherkassky et al. (2006) 57 ASD 24 yrs (10.6)57 TD 24 yrs (9.0)

Typical analysis, but datacollected during 24 sfixation trials

↓ FC between regions located inDMN, mainly in anterior–posteriorconnections

Kennedy and Courchesne (2008) 12 ASD 26.5 yrs (12.8)12 TD 27.5 yrs (10.9)

Typical Regional abnormality ofMPFC and left ANG

↓ FC in several regions of DMN↓ Number of regions in DMN

Monk et al. (2009) 12 ASD 26 yrs (5.93)12 TD 27 yrs (6.1)

Typical ↓ FC between PCC and right SFG↑ FC between PCC and righttemporal lobe↑ FC between PCC and right PHG

• ASD: negative correlation between FCbetween PCC and right SFG andmeasure of social interaction (ADI-R)• ASD: positive correlation between FCbetween PCC and right PHG andmeasure of RRB (ADI-R)

Assaf et al. (2010) 15 ASD 15.7 yrs (3.0)15 TD 17.1 yrs (3.6)

Typical ↓ FC in sub networks in the DMN,such as the MPFC and ACC

• ASD: negative correlation between FCof PrC and measure of symptomseverity on social, communication, andtotal scale (ADOS)• ASD: correlation between FC of MPFCand measure of autism symptoms(SRS)• ASD: correlation between FC of ACCand measure of autism symptoms(SRS)

Ebisch et al. (2010) 14 ASD 15.69 yrs (1.93)15 TD 15.95 yrs (1.65)

Typical ↓ FC between anterior andposterior insular cortices withemotional and sensory processingregions

Lai et al. (2010) • ASD: negative correlation between FCof LFB in right anterior INS andmeasure of social scores (ADI-R)• ASD: negative correlation between FCof LFB in bilateral splenial cortex andmeasure of symptom severity forcommunication (ADOS)

Paakki et al. (2010) 28 ASD 14.58 yrs (1.62)27 TD 14.59 yrs (1.51)

Regional homogeneityanalyses

Abnormal pattern of localconnectivity, primarilyright sided

Weng et al. (2010) 16 ASD 15 yrs (1.45)15 TD 16 yrs (1.44)

Typical ↓ FC in 9 of 11 areas of DMN • ASD: negative correlation between FCbetween PCC and several other regionsand measure of impairments in socialinteraction (ADI-R)• ASD: negative correlation between FCbetween PCC and dMPFC, medialtemporal lobes, and SFG, and measureof RRB (ADI-R)• ASD: negative correlation between FCbetween PCC and temporal lobes, andPCC and right PHG, and verbalcommunication ability (ADI-R)• ASD: negative correlation between FCbetween PCC and temporal lobes, rightPHG, and SFG, and non-verbalcommunication ability (ADI-R)

614

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625

Table 3 (Continued)

Study Participants (nr, age and SDof age)

Analysisa Results – local connectivity Results – long-range connections Results – relationship with behavior

Anderson et al. (2011) 53 ASD 22.4 yrs (7.2)39 TD 21.2 yrs (6.5)

Typical ↓ Interhemispheric FC betweensensorimotor cortices, frontalinsula, and superior parietallobules↓ FC greater in regions locatedfurther away from the midline

• Trend for correlation betweenSRS score and interhemispheric FC

Di Martino et al. (2011) 20 ASD 10.4 yrs (1.7)20 TD 10.9 yrs (1.6)

Typical ↑ and more diffuse FC betweenventral Str with INS and STG• Left ventral Str: positivelycorrelated with left STG (negativecorrelation in TD)• Dorsal caudate: positivelycorrelated with sensory processingregions (negative correlation in TD)↑ FC between dorsal PTA andvarious other regions (FC notpresent in TD)↑ FC between ventral rostral PTAand right STG and planumtemporale (FC not present in TD)↑ FC between brainstem andvarious regions (FC not present inTD)↓ FC between ventral rostral PTAand lateral occipital cortex

• Trend for a correlation betweenRRB-score (ADOS) and FC betweenright ventral rostral PTA and right STG• Trend for a negative correlationbetween RRB-score (ADOS) and FCbetween right rostral caudal PTA andright pons

Wiggins et al., 2011 39 ASD 14.0 yrs (2.08)41 TD 15.3 yrs (2.4)

Typical; SOM ↓ FC between right SFG andposterior regions

↓ Verbal cognitive functioning

TD: typically developing; FC: functional connectivity; DMN: default-mode network; MPFC: medial prefrontal cortex; ANG: angular gyrus; PCC: posterior cingulate cortex; SFG: superior frontal gyrus; PHG: parahippocampalgyrus; RRB: restricted, repetitive behavior; ACC: anterior cingulate cortex; PrC: precuneus; LFB: low-frequency BOLD-oscillatory activity; INS: insula; dMPFC: dorsal medial prefrontal cortex; Str: striatum; STG: superior temporalgyrus; PTA: putamen; and SOM: self-organizing maps.Note. If not otherwise stated, results are reported for participants with ASD when compared to the control group.

a Typical analysis = correlations between task-related changes in BOLD signal in different regions.

Biobeh

plearrrb(wwce

ccntsmcibw2ifocb

lcatrtu

6

svnoErenprp(

arprpcmrdi


osterior cingulate cortex, medial frontal cortex, medial temporalobes, and the superior frontal gyrus in adolescents with ASD (Wengt al., 2010). For social and communication impairments (ADI-Rnd ADOS), more impairments were found to be associated witheduced resting state functional connectivity between the poste-ior cingulate cortex and several other regions (Weng et al., 2010),educed functional connectivity within the right anterior insula,ilateral splenial cortices and several subnetworks of the DMNAssaf et al., 2010; Lai et al., 2010). For the SRS, increased scoresere shown to correlate with increased functional connectivityithin several sub-networks of the DMN, and increased functional

onnectivity between hemispheres (Anderson et al., 2011; Assaft al., 2010).

These findings indicate that for resting state fMRI, reduced socialompetences are accompanied by abnormal resting state functionalonnectivity between several brain regions of the default-modeetwork. Yet, while different subscales of measures of ASD symp-om severity (e.g., ADI-R; ADOS; SRS) all seem to address ASDymptoms, such as impaired social interaction and impaired com-unication, opposing correlations with resting state functional

onnectivity were found: some studies report that symptom sever-ty is positively correlated with increased functional connectivityetween subcomponents of the DMN (e.g., SRS; Assaf et al., 2010)hereas others report the opposite pattern (e.g., ADI-R; Monk et al.,

009; Weng et al., 2010). Also, for the presence of restricted, repet-tive behaviors, there is evidence for increased as well as reducedunctional connectivity between the posterior cingulate cortex andther brain regions, indicating that correlations between brainonnectivity and autism symptoms are highly specific to the com-ination of regions under study.

In sum, resting state functional connectivity was shown to corre-ate to several measures of ASD symptom severity, for example theorrelation between the severity of repetitive, restricted behavior,nd functional connectivity between the posterior cingulate cor-ex and other regions. For social and communicative impairments,esults are more difficult to interpret due to the variation in correla-ion patterns between functional connectivity and the instrumentsed to assess symptom severity.

.2.4. DiscussionIn sum, findings suggest reduced functional connectivity of

tate dependent and state independent brain activity in indi-iduals with ASD during rest. This holds for regions of intrinsicetworks such as the DMN (e.g., Weng et al., 2010), as well asther regions involved in social and emotional processing (e.g.,bisch et al., 2010). Reports of long-range over-connectivity duringest are uncommon, although striatal connectivity patterns are anxception here (Di Martino et al., 2011). With respect to local con-ectivity, the limited amount of evidence indicates a more diffuseattern of local connectivity in people with ASD, especially in theight hemisphere (Paakki et al., 2010). Diffuse patterns have beenroposed to reflect impaired differentiation of functional networksShih et al., 2011).

The presence of correlations between resting state connectivitynd behavioral characteristics of ASD shows that abnormalities inesting state connectivity in people with ASD are related to theresence of autism symptoms. With respect to the presence ofepetitive, restricted behaviors, functional connectivity with theosterior cingulate cortex seems to be of importance. For otherharacteristics of ASD, such as social and communicative impair-
ents, findings show abnormal functional connectivity in a broader
ange of regions. Notably, findings seem to be inconsistent acrossifferent diagnostic instruments, an issue that deserves attention

n future research.

avioral Reviews 36 (2012) 604–625 615

6.3. Findings from DTI studies

6.3.1. Many studies report reduced structural connectivityAn overview of the studies on structural connectivity can be

found in Table 4. When we refer to structural over- and under-connectivity, this should be interpreted as a relative difference inwhite matter integrity between groups.

The majority of findings from studies on structural connectivityshow that structural connectivity between several brain regions isweaker in individuals with ASD than in controls. Reports of under-connectivity in ASD mainly concern reduced FA of white matter infrontal and temporal regions (e.g., Barnea-Goraly et al., 2010), a col-lection of subcortical regions (e.g., Catani et al., 2008; Ingalhalikaret al., 2011; Noriuchi et al., 2010), and the corpus callosum (e.g.,Alexander et al., 2007; Jou et al., 2011; Lo et al., 2011). Further,there is evidence for reduced FA in the major pathways connect-ing different cortical lobes (fasciculi) in individuals with ASD (e.g.,Groen et al., 2011; Jou et al., 2011; Noriuchi et al., 2010; Sahyounet al., 2010).

Reduced values of FA in ASD are often found to be accom-panied by increased values of radial diffusivity (Alexander et al.,2007; Bloemen et al., 2010; Lee et al., 2007) and mean diffusivity(Alexander et al., 2007; Pugliese et al., 2009), and by reduced axialdiffusivity (Noriuchi et al., 2010).

Studies in which probabilistic tractography (e.g., Catani et al.,2008; Kumar et al., 2009; Pugliese et al., 2009; Thomas et al., 2011)or tract-based spatial statistics (e.g., Noriuchi et al., 2010; Shuklaet al., 2011a) were used, reveal differences between the trajectoriesof white matter pathways in individuals with ASD and healthy indi-viduals. The inferior longitudinal fasciculus (which connects thetemporal and occipital lobes), and the cingulum (which connectsthe cingulate cortex with medial temporal structures) were foundto be composed of more pathways in ASD than in controls, but werecharacterized by reduced values of FA (Kumar et al., 2009; Puglieseet al., 2009; Thomas et al., 2011). Similarly, the fasciculus that con-nects the limbic system with the orbito-frontal cortex (uncinatefasciculus) had an increased number of pathways, but reduced val-ues of FA in these tracts in people with ASD (Pugliese et al., 2009;Shukla et al., 2010a; Thomas et al., 2011). FA in white matter path-ways in the corpus callosum and forceps minor has been observedto be decreased (Jou et al., 2011; Noriuchi et al., 2010) and thesepathways have been shown to be composed of fewer fibers in peo-ple with ASD than in controls (Thomas et al., 2011).

6.3.2. Exceptions: Reports of over-connectivity in ASDSeveral studies report increased values of FA in ASD. Notably,

all studies with subjects older than 4 yrs found increased FA in par-ticular brain regions in individuals with ASD in combination withdecreased values of FA elsewhere in the brain. Increases of FA havebeen found in white matter in frontal (Cheng et al., 2010; Cheunget al., 2009; Ke et al., 2009), temporal (Cheng et al., 2010; Ke et al.,2009; Sahyoun et al., 2010), and occipital regions (Cheung et al.,2009), and also in some major white matter pathways (e.g., Cheunget al., 2009). Findings indicate reduced, or even absent, left later-alization of FA seen in typically developing individuals, in peoplewith ASD (Fletcher et al., 2010; Ke et al., 2009; Lange et al., 2010;Lo et al., 2011). Importantly, these findings of mixed patterns ofincreased and decreased values of FA all come from studies inves-tigating white matter in children (>4 yrs) and adolescents withASD.

However, two studies that focused on white matter integrity in
very young children (<4 yrs) with ASD suggest that early in devel-opment, the brain is characterized by a different pattern of whitematter abnormalities than the pattern found in older individu-als with ASD. White matter integrity in the corpus callosum was

616

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625

Table 4Structural connectivity.


Analysis Results – directionality of waterdiffusion

Results – tractography Results – relationship with behavior

Barnea-Goraly et al. (2004) 7 ASD 14.6 yrs (3.4)9 TD 13.4 yrs (2.8)

Typicala ↓ FA values in frontal and temporalregions

Keller et al. (2006) 34 ASD 18.9 yrs (7.3)31 TD 18.9 yrs (6.2)

Typical ↓ FA in CC and in regions surroundingCC

Ben Bashat et al. (2007) 7 ASD n/a (range 1.8–3.3 yrs)18 TD n/a (range 0.3–23 yrs)

Developmental curves werebased on TD data, to whichvalues for ASD group werethen compared

↑ FA in CC and in regions surroundingCC, predominantly in left HS

Alexander et al. (2007) 43 ASD 16.23 yrs (6.7)34 TD 16.44 yrs (5.97)

Typical ↓ FA in regions of CC↑ MD in total CC↑ Dr for all regions of CC, drivesdifferences in FA and MD

• AD: correlation between FA and PIQ,negative correlation between MD andPIQ• ASD: no correlation between FA orMD and autism traits (ADOS-G; SRS)• ASD and TD: negative correlationbetween Dr and processing speed(WPSI)• ASD and TD: correlation between Dr,FA (inverse), and MD in CC and autismtraits (SRS)

Lee et al. (2007) 43 ASD 16.2 yrs (6.7)34 TD 16.4 yrs (6.0)

Typical ↓ FA in STG and TSt↑ General Dr↑ MD in right TSt and STG

Catani et al. (2008) 15 ASD 31 yrs (9)16 TD 35 yrs (11)

Tractography ↓ Cerebellar FA, especially in right SCPand in right sICF

• ASD: negative correlation betweenFA in left CP and measure of socialinteraction (ADI-R)

Sundaram et al. (2008) 50 ASD 4.80 yrs (2.43)16 TD 8.84 yrs (3.45)

Typical ↓ FA for short range fibers in frontallobe↑ ADC in frontal lobe, for long andshort range connections↓ Fiber length

Thakkar et al. (2008) 12 ASD 30 yrs (11)12 TD 27 yrs (8)

Typical ↓ FA in WM underlying ACC • ASD: negative correlation betweenFA in left subgenual right ACC andmeasure of RRB (ADI-R)

Cheung et al. (2009) 13 ASD n/a (range 6–14 yrs)14 TD n/a (range 6–14 yrs)

Typical ↓ FA in several frontal and temporalregions↑ FA in right SLF and left occipital lobe

• ASD: negative correlation betweenFA in fronto-striatal temporal regionsand posterior CC, and measure of socialinteraction (ADI-R)• ASD: negative correlation betweenFA in fronto-striatal regions andposterior CC, and measure ofcommunication (ADI-R)• ASD: negative correlation betweenFA in various posterior regions andmeasure of RRB (ADI-R)• ASD: correlation between FA in LPGand measure of RRB (ADI-R)

Ke et al. (2009) 12 ASD 8.75 yrs (2.26)12 TD 9.40 yrs (2.07)

Typical ↓ FA in frontal and temporal regions inleft HS↑ FA in structures in frontal andtemporal regions in right HS

• ASD: correlation between FA in rightfrontal lobe and measure of autismsymptoms (CARS)• ASD: no correlation between FA andscore on ADI-R

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625

617

Kumar et al. (2009) 32 ASD 5.0 yrs (range 2.5–8.9)16 TD 5.5 yrs (range 2.5–8.6yrs)

Tractography, TBSS ↓ FA in right UNC, right Cg, right ARC,left ARC, and CC

↓ Length of left UNC↑ Length and density of rightUNC↑ Length and density in CC↑ Density of left Cg

• ASD: correlation between fiberdensity and volume of right and leftUNC and measures of stereotypicbehavior and social isolation (GARS)• ASD: correlation between fiberlength and density in CC and measureof communication (GARS)

Pugliese et al. (2009) 24 ASD 23 yrs (12)42 NC 25 yrs (10)

Tractography ↓ FA in IFOF bilaterally, and in rightUNC↑ MD in ILF bilaterally, right Cg andIFOF

↑ Density of Cg and ILF↓ Density of right UNC

Barnea-Goraly et al. (2010) 13 ASD 10.5 yrs (2.0)11 TD 9.6 yrs (2.1)

Typical, TBSS ↓ FA in several frontal parietal andtemporal lobes↓ Da in several regions

• ASD: no correlation between FA andADOS or ADI-R

Bloemen et al. (2010) 13 ASD 39 yrs (9.8)13 TD 37 yrs (9.6)

Typical ↓ FA in frontal, medial, temporal andsuperior frontal–parietal regions↑ Dr in majority of regions withreduced FA

Cheng et al. (2010) 25 ASD 13.71 (2.54)25 TD 13.51 (2.20)

Typical, TBSS ↓ FA in the right posterior limb of the IC↑ FA in the frontal lobe, right CG, INS,right STG, and bilateral middle CP↓ Dr for regions where FA wasincreased

Fletcher et al. (2010) 10 ASD 14.25 yrs (1.92)10 TD 13.36 yrs (1.34)

Typical ↓ Lateralization of FA towards left HScompared to controls↑ MD in ARC↑ Dr, drives increase in MD

Lange et al. (2010) 30 ASD 15.78 (5.6)30 TD 15.79 yrs (5.5)

Typical; NLPC ↓ InterHS asymmetry in STG↓ FA (left), and Da in STG↑ Right Da, Dr, and MD in TSt• WM characteristics in STG and TStcan be used to successfully classify ASDand TD participants

Noriuchi et al. (2010) 7 ASD 13.96 yrs (2.68)7 TD 13.36 yrs (2.74)

Typical; TBSS ↓ FA in ACC, DLPFC, right TPJ, SLF, andanterior CC↓ Da in left STS, left TPJ, and right ACC↑ Da in cerebellar vermis lobules

• ASD: negative correlation betweenFA in DLPFC region and SRS score

Sahyoun et al. (2010) 12 ASD 13.3 yrs (2.5)9 for DTI12 TD 13.3 yrs (2.1)

Typical ↓ FA in frontal regions↓ FA in the left IFOF↓ FA in right PFOF↑ FA in temporal white matter and inthe UNC

• Negative correlation between FA inregions implicated in the task and RTon visual and semantic conditions ofPPS task different for ASD and TD group

Shukla et al. (2010a) 26 ASD 12.6 yrs (0.6)24 TD 13.0 yrs (0.6)

↓ Whole-brain FA↑ Whole-brain MD and Dr↓ FA and ↑ Dr in CC↓ FA and ↑ MD and Dr in IC↓ FA in middle CP

• No correlation between DTI measuresand ADOS or ADI scores

618

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625

Table 4 (Continued)


Analysis Results – directionality of waterdiffusion

Results – tractography Results – relationship with behavior

Groen et al. (2011) 17 ASD 14.4 yrs (1.6)25 TD 15.5 yrs (1.8)

Typical ↑ Global MD↓ FA in SLF and ILF↓ FA in left corona radiata• Differences in FA disappear whencontrolling for age and IQ↑ MD in corona radiata, the IC, middleCP, thalamus and thalamic radiations,SLF, ILF, FOF, and parts of the CC

Ingalhalikar et al. (2011) 45 ASD 10.5 yrs (2.5)30 TD 10.3 yrs (2.5)

Typical; NLPC ↓ FA in right IC and EC, leftSLF, and inferior occipital WM↑ MD in occipital gyrus, superiortemporal WM, and insular cortex

• Negative correlation (weak) betweenabnormality score (of WMcharacteristics) and symptom severity(SRS; SCQ) in ASD

Jou et al. (2011) 10 ASD10 TDGeneral age: 13.5 yrs (4.0)

Typical; Tractography ↓ FA in anterior radiation and body ofCC

↓ FA in left SLF, bilateral IFOF,right SLF, and bilateral ILF

Lo et al. (2011) 15 ASD 15.2 yrs (1.0)15 TD 15.0 yrs (0.8)

Typical • Absence of typical leftwardasymmetry in ARC, UNC, and Cg↓ FA in callosal fibers to orbitofrontallobes, IFG, and STG

Shukla et al. (2011a) 26 ASD 12.8 yrs (0.6)24 TD 13.0 yrs (0.6)

Typical; TBSS ↓ FA, combined with ↑ MD and Dr in IC,ILF, IFOF, SLF, splenium of CC, CST,anterior thalamic radiation, Fo-Ma andright SCR↓ FA and ↑ Dr in genu and body of CC↑ MD and Dr in UNC and EC

Shukla et al. (2011b) 26 ASD 12.6 yrs (0.6)24 TD 13.0 yrs (0.6)

Typical; TBSS ↓ FA (marginally significant) in allshort distance WM fibers↓ FA in short distance WM fibers infrontal lobe↑ MD in short distance WM fibers infrontal, temporal, and parietal lobes↑ Dr in short distance WM fibers infrontal, temporal, and parietal lobes

Thomas et al. (2011) 12 ASD 28.5 yrs (9.7)18 HFA 22.4 yrs (4.1)

Typical, Tractography ↓ nr of streamlines in and sizeof in Fo-Mi and CC↑ nr of streamlines and size ofILF, IFOF, and UNC in the leftHS (lateralization absent in TD)

• Negative correlation between nr ofstreamlines in and size of Fo-Mi andpresence of RRB (ADI; ADOS)

Weinstein et al. (2011) 22 ASD 3.2 yrs (1.1)28 TD 3.6 yrs (1.2)

Typical, TBSS, tractography ↑ FA and ↓ Dr in genu and midbody ofCC, left SLF, and right Cg

TD: typically developing; FA: fractional anisotropy; CC: corpus callosum; HS: hemisphere; PIQ: performance IQ; MD: mean diffusivity; Dr: radial diffusivity; WPSI: Wechsler Processing Speed Index; STG: superior temporal gyrus;TSt: temporal stem; SCP: superior cerebellar peduncle; sICF: short intracerebellar fibers; CP: cerebellar peduncle; ADC: apparent diffusion coefficient; WM: white matter; ACC: anterior cingulate cortex; RRB: restricted, repetitivebehaviors; SLF: superior longitudinal fasciculus; LPG: left precentral gyrus; UNC: uncinate fasciculus; Da: axial diffusivity; Cg: cingulum; ARC: arcuate fasciculus; ILF: inferior longitudinal fasciculus; IFOF: inferior fronto-occipitalfasciculus; IC: internal capsule; CSQ: social communication questionnaire; INS: insula; NLPC: non-linear pattern classification; FOF: fronto-occipital fasciculus; CG: cingulate gyrus; DLPFC: dorsolateral prefrontal cortex; STS:superior temporal sulcus; TPJ: temporal parietal junction; PFOF: posterior fronto-occipital fasciculus; RT: response time; PPS: pictorial problem solving task; EC: external capsule; IFG: inferior frontal gyrus; CST: corticospinaltract; SCR: superior corona radiate; Fo-Ma: forceps major; and Fo-Mi: forceps-minor.Note. If not otherwise stated, results are reported for participants with ASD when compared to the control group.

a Calculation of normalized FA values for separate voxels.

Biobeh

sc

6b

ttwf2cm

bsleplfiTaTtifrsi2

rAtales(nTwra

6

rKbctsiSttlSirobc


electively found to be higher in young children with ASD than inontrols (Ben Bashat et al., 2007; Weinstein et al., 2011).

.3.3. The relationship between white matter microstructure,ehavior, and clinical symptoms

Only a handful of studies investigated the relationship betweenask performance and brain connectivity. Response times on a pic-orial problem solving task were found to correlate negatively withhite matter integrity in regions that were active during task per-

ormance, both for controls and people with ASD (Sahyoun et al.,010). In line with this finding, processing speed (Wechsler pro-essing speed index; Wechsler, 1997) was related to greater whiteatter integrity (Alexander et al., 2007).There are several reports of the presence of a relationship

etween symptom severity and white matter integrity. Reducedocial and communicative competence in people with ASD is corre-ated with reduced FA values in the left cerebellar peduncle (Catanit al., 2008), fronto-striatal regions, temporal regions, the posteriorart of the corpus callosum (Cheung et al., 2009), and the dorso-

ateral prefrontal cortex (Noriuchi et al., 2010), and with increasedber length and density in the corpus callosum (Kumar et al., 2009).he degree of repetitive, restricted behaviors was found to be neg-tively correlated to FA in various regions (Cheung et al., 2009;hakkar et al., 2008; Thomas et al., 2011) and to the number ofracts in the forceps-minor (Thomas et al., 2011). Further, a pos-tive correlation between increased fiber length in the uncinateasciculus and the corpus callosum, and symptom severity has beeneported (Kumar et al., 2009). However, there are also findings thathow that in people with ASD, ASD characteristics are not signif-cantly related to white matter characteristics (Alexander et al.,007; Barnea-Goraly et al., 2010; Shukla et al., 2010a).

In sum, reduced white matter integrity has frequently beeneported to be related to greater symptom severity in people withSD, while increased white matter integrity is related to faster

ask performance in people with ASD and controls. Further, therere several reports of a positive relationship between the fiberength and density, and the presence of autism symptoms. How-ver, there also exist reports of a lack of a relationship betweenymptom severity and white matter integrity in people with ASDe.g., Alexander et al., 2007; Barnea-Goraly et al., 2010), that can-ot be attributed to a difference in age or diagnostic instrument.hus, although there is evidence for a relationship between reducedhite matter integrity and higher symptom severity in ASD, more

esearch is desirable to investigate why these patterns may bebsent in some situations.

.3.4. DiscussionThe majority of studies on structural connectivity report

educed FA (Alexander et al., 2007; Barnea-Goraly et al., 2010;eller et al., 2006; Thakkar et al., 2008), which suggests that therain of subjects with ASD is characterized by weaker structuralonnections between brain regions. In two studies in which a dis-inction was made between FA in long and short fibers (U-fibers),hort fiber tracts in the frontal cortex were found to be weaker inndividuals with ASD compared to controls (Shukla et al., 2011b;undaram et al., 2008). Findings concerning long association fibershat run between cortical regions within hemispheres, and fibershat connect both hemispheres, predominantly show decreasedong-range connectivity (e.g., Keller et al., 2006; Kumar et al., 2009;hukla et al., 2010a). Thus, findings from DTI research in ASD arendicative of impaired long-range and local connectivity. In general,
educed integrity of white matter is related to increased presencef ASD symptoms, although findings of a lack of a relationshipetween white matter integrity and ASD symptoms complicate thisonclusion.
avioral Reviews 36 (2012) 604–625 619

Studies incorporated in this review have examined people ofvarious ages, which on the one hand impeded direct comparison offindings, but on the other hand allowed us to qualitatively exam-ine effects across age groups. The pattern of white matter integrityabnormalities in very young children with ASD is clearly differentfrom that of older people with the disorder. In very young chil-dren with ASD (age < 4 yrs), studies exclusively report increasedFA in individuals with ASD (Ben Bashat et al., 2007; Weinsteinet al., 2011). Evidence from studies with older children (age > 4 yrs)shows that this pattern reverses at an early stage in develop-ment: the majority of studies investigating brain connectivity inolder children report reduced FA values, mainly in frontal regions(e.g., Barnea-Goraly et al., 2010; Ke et al., 2009; Sundaram et al.,2008). The occasional findings of increased FA in regions in peo-ple with ASD, accompanied by reduced FA elsewhere in the brain,all stem from studies on children between the age of 4 and 16.There are reports of increased number of tracts in white matterstructures in adults with ASD (Pugliese et al., 2009; Thomas et al.,2011), but these tracts were always characterized by decreasedvalues of FA, reflecting more voluminous but less effective path-ways in ASD. Supporting the effect of age observed in the reviewedstudies, there is evidence that the effect of age on white matterintegrity is different between healthy people and people with ASD(Shukla et al., 2011a). To conclude, the effect of age on white mattermicrostructure in ASD seems to be a highly interesting venue forfuture research. More specifically, it will be interesting to investi-gate whether young children with ASD also show increased whitematter integrity in regions other than the corpus callosum, andwhether the shift from structural over-connectivity at young age,to structural under-connectivity later in life, indeed occurs in indi-viduals with ASD.

6.4. Findings from EEG and MEG studies

6.4.1. Divergent findings of over- and under-connectivityAn overview of EEG and MEG studies on functional connec-

tivity can be found in Table 5. We incorporated EEG and MEGstudies in which connectivity was investigated by examining syn-chronization of brain activity in distant brain regions. Two studiesusing MEG and six studies using EEG were incorporated in thisreview.

We use the same definitions of under- and over-connectivity asin Section 6.1.1.

In people with ASD there was increased synchronizationof gamma band activity during a sentence processing task(Braeutigam et al., 2008) and during rest, especially between tem-poral and other brain regions (Sheikhani et al., 2010). Increasedconnectivity in the gamma, as well as the beta band, has also beenobserved in the parietal cortex, although in this study it was accom-panied by reduced connectivity between frontal regions in thesefrequency bands (Perez Velazquez et al., 2009). Findings concern-ing connectivity in lower frequency bands (delta and theta) areheterogeneous. During sleep and rest, connectivity in the deltaand theta band was reduced in mid-frontal regions, and betweenfrontal and occipital regions (Barttfeld et al., 2010), but increased inoccipital areas (Léveillé et al., 2010). Interhemispheric connectiv-ity in the theta band was found to be impaired in people with ASD(Isler et al., 2010), while connectivity in the theta frequency bandin people with ASD has been observed to be increased betweenfrontal regions, accompanied by weaker alpha band connectivity
between distant brain regions (Murias et al., 2007). Further, thereis evidence for generally reduced functional connectivity betweenbrain regions in delta, beta, and theta frequency bands (Coben et al.,2008).

620

M.E.

Vissers

et al.

/ N

euroscience and

Biobehavioral R

eviews

36 (2012)

604–625

Table 5EEG and MEG studies on functional connectivity (FC).


Method Paradigm Results Results – relationship with behavior

Murias et al. (2007) 18 ASD 22.7 yrs (4.4)18 TD 24.9 yrs (6.8)

EEG Resting state ↑ Within frontal regions (�)↓ Long-range connectivity betweendifferent cortical regions (�)

Braeutigam et al. (2008) 11 ASD (28–52 yrs)11 TD (25–54 yrs)

MEG Sentence processing-task ↑ (�)

Coben et al. (2008) 20 ASD n/a (age range 6–11yrs)20 TD n/a (age range 6–11 yrs)

EEG Resting state ↓ Between electrodes at short-mediumdistance and at long distance (�, �, and�)↓ Between HS

Perez Velazquez et al. (2009) 15 ASD 10.8 yrs (3.4)16 TD 11.1 yrs (2.6)

MEG Executive function tasks ↓ Overall↓ Between prefrontal regions (� and �)↑ Within parietal cortices (� and �)

• ASD: ↑ perseverative errors and ↑distraction errors in card sorting task

Barttfeld et al. (2010) 10 ASD 23.8 yrs (7.6)10 TD 25.3 yrs (6.54)

EEG Resting state ↓ In mid-frontal regions (�)↓ Between mid-frontal and occipitalregions (�)↑ In left frontal regions (�)

• Negative correlation betweensynchronization likelihood atmid-frontal electrodes (�) and ADOSscores• Correlation between focalsynchronization likelihood at lateralfrontal electrodes (�) and ADOS scores

Isler et al. (2010) 9 ASD 7.8 yrs (0.57)11 TD 6.8 yrs (0.76)

EEG Long latency visual evokedpotential

↓ Between hemispheres (�)

Léveillé et al. (2010) 9 ASD 21.1 yrs (4.0)13 TD 21.5 yrs (4.3)

EEG Recordings during REM-sleep ↓ In right frontal regions (� and �)↑ Occipital lobes for electrode pairs atlong and short distances (� and �)

Sheikhani et al. (2010) 17 ASD 9.2 yrs (1.6)11 TD 9.5 yrs (2.8)

EEG Resting state ↑ Between distant electrodes (�),especially between temporal and otherbrain regions

TD: typically developing; FC: functional connectivity; �: theta; �: alpha; �: gamma; �: delta; �: beta; HS: hemisphere; and REM: rapid eye movement.Note. If not otherwise stated, results are reported for participants with ASD when compared to the control group.

Biobeh

6w

tnaIw2

6

vFb(ebildficdaaa

tuuFsMittEgMtfa

7

7c

rcocaecw(ticwrta


.4.2. The relationship between functional connectivity measuredith EEG or MEG, and clinical symptoms

Only one study investigated the relationship between symp-oms of ASD and abnormal brain connectivity. Results showed aegative correlation between connectivity between mid-frontalnd other electrodes during rest, and autism characteristics (ADOS).ncreased focal synchronization at lateralized frontal electrodes

as related to more autism symptoms (ADOS; Barttfeld et al.,010).

.4.3. DiscussionEEG and MEG studies on functional connectivity do not pro-

ide a clear message on functional connectivity in people with ASD.or example, the increased connectivity in the gamma frequencyand in people with ASD may indicate increased local connectivityVon Stein and Sarnthein, 2000). However, the decreased coher-nce in gamma band activity between distant brain regions, foundy Sheikhani et al. (2010), is not in line with this notion. Find-

ngs on connectivity in lower frequency bands that may mediateong-range connectivity (Von Stein and Sarnthein, 2000) concernifferent brain regions and are not uniform. Also, when comparingndings on specific regions, findings are inconsistent. For example,onnectivity between frontal regions has both been observed to beecreased (e.g., Léveillé et al., 2010; Perez Velazquez et al., 2009)s well as increased in people with ASD (Murias et al., 2007). Thus,lthough these studies consistently report differences between ASDnd controls, the direction of those differences is inconsistent.

The lack of uniformity among findings is likely attributable tohe different focus of the studies discussed in this section. Studiessually focus on one or more particular frequency bands, a partic-lar cognitive state or process, and participants of a particular age.or example, age may have an effect on brain-connectivity mea-ured with EEG (Micheloyannis et al., 2009). Furthermore, EEG andEG studies in which tasks were administered, as well as stud-

es in which subjects rested or slept, were discussed together dueo their small number. The cognitive state may have influencedhe patterns of connectivity. Therefore, the field is in need of moreEG and MEG studies on functional brain connectivity, in order toain more insight into characteristics of functional brain activation.ore EEG and MEG research will provide knowledge on the fine

emporal characteristics of synchronization of brain activity in ASD,or example of difference between abnormalities in connectivitycross frequency bands.

. What can we conclude about connectivity in ASD?

.1. The relationship between structural and functionalonnectivity

In healthy people, structural and functional connectivity duringest, measured through the BOLD-signal, appear to be positivelyorrelated (Damoiseaux and Greicius, 2009). However, both typesf connectivity can exist independently. For example, functionalonnectivity was observed between regions that did not show

structural connection (Damoiseaux and Greicius, 2009; Honeyt al., 2009). This observation may indicate that brain regionsan also interact via indirect routes (Honey et al., 2009) or viaeak structural connections with high functional significance

Damoiseaux and Greicius, 2009). Structural connectivity and func-ional connectivity, as measured with EEG, were shown to correlaten several studies. Both in young adults and in elderly, functionalonnectivity between distant brain regions was found to correlate
ith white matter integrity in pathways connecting these brain
egions (Cohen, 2011b; Teipel et al., 2009). To our knowledge,his type of research has only been conducted in people withoututism. It is possible that measures of structural and functional

avioral Reviews 36 (2012) 604–625 621

connectivity are not related in a similar way in people with ASD asthey are in people without a brain disorder (Bassett and Bullmore,2009; Rubinov and Sporns, 2010).

7.2. Do findings from different types of studies indicate similarabnormalities in connectivity for people with ASD?

The reviewed evidence suggests the answer to this question is“to a high degree” for task-related and resting state MRI studies, “tosome degree” for DTI studies, and “no” for EEG and MEG studies.

Findings from MRI studies on functional connectivity show thatduring task performance, there is reduced functional connectivitybetween brain regions in people with ASD. Reduced connectivitywas most often observed between frontal and parietal regions (e.g.,Just et al., 2007; Kana et al., 2006), but was also seen between otherbrain regions (e.g., Agam et al., 2010; Just et al., 2004; Kana et al.,2009; Kleinhans et al., 2008; Mason et al., 2008; Solomon et al.,2009). Findings from resting state MRI studies are in line withfindings during task performance, indicating under-connectivitybetween regions of the DMN (e.g., Assaf et al., 2010; Cherkasskyet al., 2006; Weng et al., 2010), which are predominantly locatedin frontal and parietal regions (Broyd et al., 2009).

Although findings from functional connectivity MRI studiesgenerally converge, fMRI studies that focused on low-frequencycomponents of brain connectivity, local patterns of synchroniza-tion, or subcortical regions, yield a more complex pattern offindings. For example, local patterns of connectivity in people withASD were different during rest and task performance (Paakki et al.,2010; Shukla et al., 2010b, but for a discussion of possibly rele-vant methodological differences between these studies, see Shuklaet al., 2010b). Also, connectivity of the low-frequency BOLD-signalhas both been found to be increased (e.g., Noonan et al., 2009) anddecreased (Jones et al., 2010; Lai et al., 2010) in people with ASD.Further, findings on connectivity between subcortical and corticalregions appear to be different from cortico-cortical connectivitypatterns, and also seem to be state dependent: while there wasincreased subcortico-cortical connectivity between people withASD during task performance (e.g., Mizuno et al., 2006), there wasdecreased connectivity between some of these regions during rest(e.g., Di Martino et al., 2011; Ebisch et al., 2010).

Findings from DTI research are partly in line with the findingsfrom functional imaging research. The large amount of evidencefor reduced white matter integrity in people with ASD supportsthe notion of long-range under-connectivity also observed infMRI research. However, while reduced white matter integritywas most frequently observed in frontal and temporal regions(Alexander et al., 2007; Barnea-Goraly et al., 2010; Keller et al.,2006; Thakkar et al., 2008), functional under-connectivity was mostoften observed in frontal and parietal regions (e.g., Just et al., 2007;Kana et al., 2006). Thus, the regions in which under-connectivitywas most frequently observed in people with ASD are not com-pletely the same in DTI and fMRI connectivity research. Withrespect to local connectivity, DTI research indicates impaired localconnectivity in the frontal cortex, which is in line with a selection offindings from fMRI studies in ASD. However, the pattern of abnor-malities in structural brain connectivity in young children withASD shows over-connectivity instead of under-connectivity (e.g.,Weinstein et al., 2011). Unfortunately these data cannot be com-pared to data from functional connectivity studies in this particularage group. Thus, while DTI research support findings of reducedlong-range connectivity in ASD, the results with respect to the local-ization of the effects, and results with respect to local patterns of
connectivity, are not entirely compatible.
The majority of findings from EEG and MEG do not confirm thepattern of general reduced functional connectivity that emergesfrom MRI studies. Instead, studies show different and sometimes

6 Biobeh

cMwfaMrc

iftMdepunplr

svlasaata

7i

tMofisttl(a

paanabsbt(r

toPcS

r


ontradictory outcomes. The diversity in findings from EEG andEG studies may be explained by the different paradigms thatere used to investigate connectivity in ASD, by the fact that dif-

erent frequency bands were investigated across studies, and byge differences between samples. In sum, findings from EEG andEG studies do not seem to be compatible with the majority of

esults from MRI research, but do confirm that ASD is a disorderharacterized by atypical patterns of brain connectivity.

The differences in outcomes between the MRI and electrophys-ological (EEG and MEG) studies may be partly attributable to theact that electrophysiological techniques are characterized by highemporal, but low spatial resolution (especially EEG), compared to

RI research. This way, these methods will inherently provide aifferent representation of what happens in the brain (for a morextensive discussion, see Cohen, 2011a). EEG and MEG researchrovides highly specific information about time, but cannot informs about local connectivity patterns or physical pathways that con-ect regions. On the other hand, the divergent findings concerningarticular frequency bands from EEG and MEG research in ASD, will

ikely not be detected by MRI research, since MRI lacks the temporalesolution to distinguish between adjacent frequency bands.

From this review it appears that findings from functional andtructural MRI research on connectivity in ASD are highly con-ergent. However, results from fMRI research concerning theow-frequency component of the BOLD-signal, and results from EEGnd MEG research, are still heterogeneous. This pattern demon-trates that spatial characteristics of brain connectivity in ASD havelready been well characterized compared to the temporal char-cteristics of brain connectivity in ASD, and shows that the fineemporal characteristics of brain connectivity in ASD deserve morettention in future research.

.3. Do the findings support current theories about connectivityn ASD?

The reviewed evidence suggests that the answer to this ques-ion is “yes” for findings on long-range connectivity from functional

RI research and DTI research, “no” for findings on local patternsf connectivity from functional MRI and DTI research, and “no” forndings from EEG and MEG research. Thus, the findings from MRItudies on functional and structural connectivity are in line withhe idea of long-range under-connectivity between the frontal cor-ex and other brain regions in ASD, but the findings concerningocal patterns of connectivity do not support the notion of increasedfrontal) local connectivity in ASD proposed before (Courchesnend Pierce, 2005).

The currently dominant theory of brain connectivity in peo-le with ASD holds that there is long-range under-connectivitynd local over-connectivity of the frontal cortex (Courchesnend Pierce, 2005). Findings from MRI studies on functional con-ectivity and structural connectivity partly support this accountnd show that there is indeed reduced long-range connectivityetween the frontal cortex and other brain regions in ASD. Thesetudies demonstrated reduced long-range connectivity, especiallyetween frontal and parietal regions located in the frontal cor-ex (fMRI research), and between frontal and temporal regionsDTI research), although reduced connectivity between other brainegions has been reported as well.

Current findings on local connectivity from structural and func-ional MRI studies in ASD do not support the notion of localver-connectivity in frontal regions as proposed by Courchesne andierce (2005), but rather contradict the notion of enhanced local
onnectivity in ASD in frontal as well as other brain regions (e.g.,hukla et al., 2010b, 2011b; Sundaram et al., 2008).
Findings from EEG and MEG research do not provide convergingesults that support the notion of long-range under-connectivity


and local over-connectivity, but rather show that in people withASD, brain regions interact in a different way than they do in typi-cally developing people.

There are some aspects of brain connectivity that seem relevantfor ASD that are not addressed by current theory, such as connectiv-ity between cortical and subcortical regions, or differences betweenconnectivity patterns in different frequency bands. Also, findingsfrom MEG and EEG research seem to indicate that abnormalities inconnectivity in ASD are more complex than the pattern suggestedby the account of local over-connectivity and long-range under-connectivity (Courchesne and Pierce, 2005). From the reviewedevidence, it becomes clear that the theory only covers a selectivepart of findings from the field. Hence, even though the theory oflong-range under-connectivity and local over-connectivity of thefrontal cortex in ASD has been valuable as it generated a large bodyof research, the reviewed findings point out that this theory needsto be revised and refined in order to explain the actual pattern ofabnormalities in brain connectivity in ASD.

7.4. Do different techniques show similar relationships betweenbrain connectivity and behavior in ASD?

In functional MRI research during task performance and duringrest, and in EEG research, various correlations between impairedfunctional connectivity and presence of autism symptoms havebeen shown. However, the direction of the correlation betweenfunctional connectivity and behavior often varies per region understudy and the subscale of the diagnostic instrument. Yet, there iscoherent evidence for a link between the presence of repetitive,restricted behaviors, and functional connectivity between the pos-terior cingulate cortex and other brain regions. From DTI studiesit appears that reduced white matter integrity, which is frequentlyobserved in individuals with ASD, is related to increased presence ofautism symptoms, such as social and communicative impairments,and to reduced processing speed on cognitive tasks.

Together, these findings indicate that it is informative to studybrain connectivity in ASD: abnormalities in the way brain regionscommunicate are related to abnormalities that are typical of thedisorder. However, at present most findings on this relationshipcome from post-hoc correlation analyses. For future research,it will be interesting to conduct hypothesis-driven analysis, forexample by constructing empirically testable models on the rela-tionship between symptoms of ASD and brain connectivity. Thereis a relatively large number of studies reporting a relationshipbetween presence of repetitive, restricted behaviors, and functionalconnectivity in individuals with ASD. Especially, the functional con-nectivity of posterior and anterior cingulate cortex with other brainareas seems to relate to the presence of repetitive, restricted behav-iors in people with ASD (e.g., Agam et al., 2010; Monk et al., 2009;Weng et al., 2010). A model regarding the contribution of connec-tivity strength between cingulate cortex and other regions to thepresence of restricted repetitive behaviors will be useful in clarify-ing the role of this particular brain region within the brain. Anotherpromising model would be a model of different characteristics ofwhite matter in relation to ASD symptoms. From the current find-ings, it appears that abnormalities in the number of tracts, thelength of tracts, and FA values of tracts differ in their relationwith ASD symptoms. A single model that captures the relation-ship between abnormal length, number, and FA of pathways, anddifferent behavioral characteristics of ASD, might generate empir-ical studies that contribute to a more complete understanding ofstructural brain abnormalities in people with ASD. This type of
research will provide a more direct assessment of the relationshipbetween neurobiological and behavioral measures than the mea-sures of correlation that have been provided until now (Kievit et al.,2011), and will thereby be more informative with respect to the

Biobeh

rdt

8

ltAtbRag

bintc

cecmdcofaDild

er2snfeoci

cocctiwtrmK

ucir2v


elationship between brain and behavior in ASD than the currentlyiffuse patterns of correlations provided by studies incorporated inhis review.

. Future directions

The fact that the account of long-range over-connectivity andocal under-connectivity (Courchesne and Pierce, 2005) falls shorto describe the actual pattern of abnormalities in connectivity inSD stresses the importance of future research. More studies using

echniques that have not been applied frequently in research onrain connectivity in ASD are desirable, such as studies that applyeHo analyses, DTI tractography or tract-based spatial statistics,nd studies using EEG and MEG. Also, it seems relevant to investi-ate subcortico-cortical connectivity in ASD more extensively.

As discussed in Section 2.2, agreement on the distinctionetween long-range and local connectivity is lacking. This is an

mportant complication when comparing different studies on con-ectivity. A clear definition of local versus long-range connectivityhat is generally accepted in the field seems essential to enableomparison of findings across future studies.

In order to interpret the variation and the consistency in out-omes between current studies, it is essential to investigate theffects of age on brain connectivity in ASD. For example, age-relatedhanges in cortical thickness (Wallace et al., 2010), and in whiteatter integrity (Shukla et al., 2011a), have been observed to be

ifferent in people with ASD than in controls. Moreover, structuralonnectivity studies in people with ASD show a different patternf abnormalities in young children than in older individuals. Forunctional connectivity during rest, there is some evidence for thebsence of age related increase in functional connectivity in theMN seen in controls (Wiggins et al., 2011). Cross-sectional stud-

es, in which similar designs and analysis techniques are used, andongitudinal studies, will provide more reliable knowledge aboutevelopmental patterns of brain connectivity in ASD.

Further, more knowledge on the relationship between differ-nt forms of connectivity in individuals with ASD is needed. Thiselationship has been investigated in healthy people (e.g., Cohen,011b; Honey et al., 2009; Teipel et al., 2009), but at present notudy investigated the relationship between different forms of con-ectivity in people with ASD. Possibly, the way structural and

unctional connectivity are related in people with ASD is differ-nt from healthy people (Bassett and Bullmore, 2009). Knowledgen the relationship between structural connectivity and functionalonnectivity in ASD will help to interpret the meaning of differencesn abnormalities in different forms of connectivity.

Another interesting direction for future research is the appli-ation of models that enable the investigation of directionalityf effects, such as Granger causality (Granger, 1969), or dynamicausal modeling (DCM; Friston et al., 2011). These models will espe-ially be of interest in functional connectivity research in ASD, ashe use of such models will provide information about differencesn the functional roles of regions in healthy people and people

ith ASD, and the way these differences shape connectivity. Also,he application of these methods is desirable in research on theelationship between behavioral characteristics of ASD and abnor-alities in brain connectivity (for a more extensive discussion, see

ievit et al., 2011).We selectively examined results from high functioning individ-

als with ASD (see criterion 1, Section 3). Hence, our conclusionsannot be generalized to the whole ASD population. Various stud-
es have shown an effect of intelligence, and especially of mentaletardation, on brain connectivity (e.g., Chiang et al., 2009; Yu et al.,008). This shows that brain connectivity in low functioning indi-iduals with ASD is a complex issue to investigate, as abnormalities
avioral Reviews 36 (2012) 604–625 623

in brain connectivity in these individuals may be related to mentalretardation, as well as ASD. Currently, (in vivo) brain connectiv-ity in low functioning individuals with ASD has been investigatedless frequently than in high functional individuals with ASD. There-fore, the way brain connectivity in low functioning individuals withASD is different from healthy controls, and from high functionalindividuals with ASD, is an important question that deserves moreattention in future research.

In this review, we showed that abnormal patterns of brain con-nectivity in ASD are more complex and diffuse than previouslythought. There is evidence for reduced long-range connectivitybetween the frontal cortex and other brain regions in ASD, as pro-posed by one of the dominant theories of brain connectivity inthe field. However, contradictory to this theory, there is also evi-dence for impaired local connectivity in several regions of the brain,including the frontal cortex. Further, several abnormalities in brainconnectivity in ASD are not included in the currently dominanttheory in the field. In future research, it will be important to fur-ther characterize the complex pattern of brain connectivity in ASD,and to assess the relationship between brain connectivity and ASDsymptoms in a hypothesis-driven manner. Further, the introduc-tion of a widely accepted definition of long-range and local brainconnectivity in ASD, which is still lacking in the field, will contributeto a more complete understanding of brain connectivity in ASD andthe way it is related to behavioral characteristics of the disorder.

References

Agam, Y., Joseph, R.M., Barton, J.J., Manoach, D.S., 2010. Reduced cognitive con-trol of response inhibition by the anterior cingulate cortex in autism spectrumdisorders. NeuroImage 52, 336–347.

Alexander, A.L., Lee, J.E., Lazar, M., Boudos, R., DuBray, M.B., Oakes, T.R., Miller, J.N., Lu,J., Jeong, E.K., McMahon, W.M., Bigler, E.D., Lainhart, J.E., 2007. Diffusion tensorimaging of the corpus callosum in Autism. NeuroImage 34, 61–73.

American Psychiatric Association, 1994. DSM-IV, Diagnostic and Statistical Manualof Mental Disorders, 4th ed. American Psychiatric Association, Washington, DC.

American Psychiatric Association, 2000. DSM-IV-TR, Diagnostic and Statistical Man-ual of Mental Disorders, 4th ed., text revision. American Psychiatric Association,Washington, DC.

Anderson, J.S., Druzgal, T.J., Froehlich, A., DuBray, M.B., Lange, N., Alexander, A.L.,Abildskov, T., Nielsen, J.A., Cariello, A.N., Cooperrider, J.R., Bigler, E.D., Lainhart,J.E., 2011. Decreased interhemispheric functional connectivity in autism. Cereb.Cortex 21, 1134–1146.

Assaf, M., Jagannathan, K., Calhoun, V.D., Miller, L., Stevens, M.C., Sahl, R., O’Boyle,J.G., Schultz, R.T., Pearlson, G.D., 2010. Abnormal functional connectivity ofdefault mode sub-networks in autism spectrum disorder patients. NeuroImage53, 247–256.

Barnea-Goraly, N., Kwon, H., Menon, V., Eliez, S., Lotspeich, L., Reiss, A.L., 2004. Whitematter structure in autism: preliminary evidence from diffusion tensor imaging.Biol. Psychiatry 55, 323–326.

Barnea-Goraly, N., Lotspeich, L.J., Reiss, A.L., 2010. Similar white matter aberrationsin children with autism and their unaffected siblings: a diffusion tensor imagingstudy using tract-based spatial statistics. Arch. Gen. Psychiatry 67, 1052–1060.

Barttfeld, P., Wicker, B., Cukier, S., Navarta, S., Lew, S., Sigman, M., 2010. A big-worldnetwork in ASD: dynamical connectivity analysis reflects a deficit in long-rangeconnections and an excess of short-range connections. Neuropsychologia 49,254–263.

Bassett, D.S., Bullmore, E.T., 2009. Human brain networks in health and disease. Curr.Opin. Neurol. 22, 340–347.

Belmonte, M.K., Allen, G., Beckel-Mitchener, A., Boulanger, L.M., Carper, R.A., Webb,S.J., 2004. Autism and abnormal development of brain connectivity. J. Neurosci.24, 9228–9231.

Ben Bashat, D., Kronfeld-Duenias, V., Zachor, D.A., Ekstein, P.M., Hendler, T., Tarrasch,R., Even, A., Levy, Y., Ben Sira, L., 2007. Accelerated maturation of white matter inyoung children with autism: a high b value DWI study. NeuroImage 37, 40–47.

Bloemen, O.J., Deeley, Q., Sundram, F., Daly, E.M., Barker, G.J., Jones, D.K., VanAmelsvoort, T.A., Schmitz, N., Robertson, D., Murphy, K.C., Murphy, D.G., 2010.White matter integrity in Asperger syndrome: a preliminary diffusion tensormagnetic resonance imaging study in adults. Autism Res. 3, 203–313.

Bluhm, R.L., Miller, J., Lanius, R.A., Osuch, E.A., Boksman, K., Neufeld, R.W., Theberge,J., Schaefer, B., Williamson, P., 2007. Spontaneous low-frequency fluctuations inthe BOLD signal in schizophrenic patients: anomalies in the default network.
Schizophr. Bull. 33, 1004–1012.
Bollen, K.A., 1989. Structural Equations with Latent Variables. Wiley, New York.Braeutigam, S., Swithenby, S.J., Bailey, A.J., 2008. Contextual integration the unusual

way: a magnetoencephalographic study of responses to semantic violation inindividuals with autism spectrum disorders. Eur. J. Neurosci. 27, 1026–1036.

6 Biobeh

B

B

B

B

C

C

C

C

C

C

C

C

C

CC

C

C

C

C

D

D

D

D

E

F

F

F

F

G

G

G


rieber, S., Herpertz-Dahlmann, B., Fink, G.R., Kamp-Becker, I., Remschmidt, H., Kon-rad, K., 2010. Coherent motion processing in autism spectrum disorder (ASD):an fMRI study. Neuropsychologia 48, 1644–1651.

rock, J., Brown, C.C., Boucher, J., Rippon, G., 2002. The temporal binding deficithypothesis of autism. Dev. Psychopathol. 14, 209–224.

royd, S.J., Demanuele, C., Debener, S., Helps, S.K., James, C.J., Sonuga-Barke, E.J.,2009. Default-mode brain dysfunction in mental disorders: a systematic review.Neurosci. Biobehav. Rev. 33, 279–296.

uckner, R.L., Andrews-Hanna, J.R., Schacter, D.L., 2008. The brain’s default network:anatomy, function, and relevance to disease. Ann. N. Y. Acad. Sci. 1124, 1–38.

arper, R.A., Courchesne, E., 2005. Localized enlargement of the frontal cortex inearly autism. Biol. Psychiatry 57, 126–133.

asanova, M.F., 2006. Neuropathological and genetic findings in autism: the signif-icance of a putative minicolumnopathy. Neuroscientist 12, 435–441.

asanova, M.F., Buxhoeveden, D.P., Switala, A.E., Roy, E., 2002. Minicolumnar pathol-ogy in autism. Neurology 58, 428–432.

atani, M., Jones, D.K., Daly, E., Embiricos, N., Deeley, Q., Pugliese, L., Curran, S.,Robertson, D., Murphy, D.G.M., 2008. Altered cerebellar feedback projectionsin Asperger syndrome. NeuroImage 41, 1184–1191.

heng, Y., Chou, K.H., Chen, I.Y., Fan, Y.T., Decety, J., Lin, C.P., 2010. Atypical devel-opment of white matter microstructure in adolescents with autism spectrumdisorders. NeuroImage 50, 873–882.

herkassky, V.L., Kana, R.K., Keller, T.A., Just, M.A., 2006. Functional connectivity ina baseline resting-state network in autism. Neuroreport 17, 1687–1690.

heung, C., Chua, S.E., Cheung, V., Khong, P.L., Tai, K.S., Wong, T.K., Ho, T.P., McAlo-nan, G.M., 2009. White matter fractional anisotropy differences and correlatesof diagnostic symptoms in autism. J. Child Psychol. Psychiatry 50, 1102–1112.

hiang, M.C., Barysheva, M., Shattuck, D.W., Lee, A.D., Madsen, S.K., Avedissian, C.,Klunder, A.D., Toga, A.W., McMahon, K.L., de Zubicaray, G.I., Wright, M.J., Srivas-tava, A., Balov, N., Thompson, P.M., 2009. Genetics of brain fiber architecture andintellectual performance. J. Neurosci. 29, 2212–2224.

oben, R., Clarke, A.R., Hudspeth, W., Barry, R.J., 2008. EEG power and coherence inautistic spectrum disorder. Clin. Neurophysiol. 119, 1002–1009.

ohen, M.X., 2011a. It’s about time. Front. Hum. Neurosci. 5, 2.ohen, M.X., 2011b. Error-related medial frontal theta activity predicts cingulate-

related structural connectivity. NeuroImage 55, 1373–1383.onstantino, J.N., Todd, R.D., 2005. Intergenerational transmission of subthreshold

autistic traits in the general population. Biol. Psychiatry 57, 655–660.ourchesne, E., Karns, C.M., Davis, H.R., Ziccardi, R., Carper, R.A., Tigue, Z.D., Chisum,

H.J., Moses, P., Pierce, K., Lord, C., Lincoln, A.J., Pizzo, S., Schreibman, L., Haas,R.H., Akshoomoff, N.A., Courchesne, R.Y., 2001. Unusual brain growth patternsin early life in patients with autistic disorder: an MRI study. Neurology 57,245–254.

ourchesne, E., Pierce, K., 2005. Why the frontal cortex in autism might be talk-ing only to itself: local over-connectivity but long-distance disconnection. Curr.Opin. Neurobiol. 15, 225–230.

ritchley, H.D., Daly, E.M., Bullmore, E.T., Williams, S.C., Van Amelsvoort, T., Robert-son, D.M., Rowe, A., Phillips, M., McAlonan, G., Howlin, P., Murphy, D.G., 2000.The functional neuroanatomy of social behaviour: changes in cerebral bloodflow when people with autistic disorder process facial expressions. Brain 123,2203–2212.

amarla, S.R., Keller, T.A., Kana, R.K., Cherkassky, V.L., Williams, D.L., Minshew, N.J.,Just, M.A., 2010. Cortical underconnectivity coupled with preserved visuospatialcognition in autism: evidence from an fMRI study of an embedded figures task.Autism Res. 3, 273–279.

amoiseaux, J.S., Greicius, M.D., 2009. Greater than the sum of its parts: a review ofstudies combining structural connectivity and resting-state functional connec-tivity. Brain Struct. Funct. 213, 525–533.

i Martino, A., Kelly, C., Grzadzinski, R., Zuo, X.N., Mennes, M., Mairena, M.A., Lord, C.,Castellanos, F.X., Milham, M.P., 2011. Aberrant striatal functional connectivityin children with autism. Biol. Psychiatry 69, 847–856.

i Martino, A., Ross, K., Uddin, L.Q., Sklar, A.B., Castellanos, F.X., Milham, M.P., 2009.Functional brain correlates of social and nonsocial processes in autism spectrumdisorders: an activation likelihood estimation meta-analysis. Biol. Psychiatry 65,63–74.

bisch, S.J., Gallese, V., Willems, R.M., Mantini, D., Groen, W.B., Romani, G.L., Buite-laar, J.K., Bekkering, H., 2010. Altered intrinsic functional connectivity of anteriorand posterior insula regions in high-functioning participants with autism spec-trum disorder. Hum. Brain Mapp. 32, 1013–1028.

letcher, P.T., Whitaker, R.T., Tao, R., DuBray, M.B., Froehlich, A., Ravichandran, C.,Alexander, A.L., Bigler, E.D., Lange, N., Lainhart, J.E., 2010. Microstructural con-nectivity of the arcuate fasciculus in adolescents with high-functioning autism.NeuroImage 51, 1117–1125.

ox, M.D., Raichle, M.E., 2007. Spontaneous fluctuations in brain activity observedwith functional magnetic resonance imaging. Nat. Rev. Neurosci. 8, 700–711.

ries, P., 2005. A mechanism for cognitive dynamics: neuronal communicationthrough neuronal coherence. Trends Cogn. Sci. 9, 474–480.

riston, K.J., Li, B., Daunizeau, J., Stephan, K.E., 2011. Network discovery with DCM.NeuroImage 56, 1201–1221.

illiam, J., 1995. Gilliam Autism Rating Scale: Examiner’s Manual. Pro-Ed, Austin,TX.

irgis, R.R., Minshew, N.J., Melhem, N.M., Nutche, J.J., Keshavan, M.S., Hardan, A.Y.,2007. Volumetric alterations of the orbitofrontal cortex in autism. Prog. Neu-ropsychopharmacol. Biol. Psychiatry 31, 41–45.

ranger, C.W.J., 1969. Investigating causal relations by econometric models andcross-spectral methods. Econometrica 37, 424–438.


Groen, W.B., Buitelaar, J.K., van der Gaag, R.J., Zwiers, M.P., 2011. Pervasivemicrostructural abnormalities in autism: a DTI study. J. Psychiatry Neurosci. 36,32–40.

Happé, F., Frith, U., 2006. The weak coherence account: detail-focused cognitive stylein autism spectrum disorders. J. Autism Dev. Disord. 36, 5–25.

Honey, C.J., Sporns, O., Cammoun, L., Gigandet, X., Thiran, J.P., Meuli, R., Hagmann,P., 2009. Predicting human resting-state functional connectivity from structuralconnectivity. Proc. Natl. Acad. Sci. U.S.A. 106, 2035–2040.

Horwitz, B., Rumsey, J.M., Grady, C.L., Rapoport, S.I., 1988. The cerebral metaboliclandscape in autism. Intercorrelations of regional glucose utilization. Arch. Neu-rol. 45, 749–755.

Ingalhalikar, M., Parker, D., Bloy, L., Roberts, T.P., Verma, R., 2011. Diffusion basedabnormality markers of pathology: toward learned diagnostic prediction of ASD.NeuroImage 57, 918–927.

Isler, J.R., Martien, K.M., Grieve, P.G., Stark, R.I., Herbert, M.R., 2010. Reduced func-tional connectivity in visual evoked potentials in children with autism spectrumdisorder. Clin. Neurophysiol. 121, 2035–2043.

Johansen-Berg, H., Rushworth, M.F., 2009. Using diffusion imaging to study humanconnectional anatomy. Annu. Rev. Neurosci. 32, 75–94.

Jones, T.B., Bandettini, P.A., Kenworthy, L., Case, L.K., Milleville, S.C., Martin, A., Birn,R.M., 2010. Sources of group differences in functional connectivity: an investi-gation applied to autism spectrum disorder. NeuroImage 49, 401–414.

Jou, R.J., Jackowski, A.P., Papademetris, X., Rajeevan, N., Staib, L.H., Volkmar, F.R.,2011. Diffusion tensor imaging in autism spectrum disorders: preliminary evi-dence of abnormal neural connectivity. Aust. N.Z. J. Psychiatry 45, 153–162.

Just, M.A., Cherkassky, V.L., Keller, T.A., Kana, R.K., Minshew, N.J., 2007. Functionaland anatomical cortical underconnectivity in autism: evidence from an fMRIstudy of an executive function task and corpus callosum morphometry. CerebralCortex 17, 951–961.

Just, M.A., Cherkassky, V.L., Keller, T.A., Minshew, N.J., 2004. Cortical activation andsynchronization during sentence comprehension in high-functioning autism:evidence of underconnectivity. Brain 127, 1811–1821.

Kana, R.K., Keller, T.A., Cherkassky, V.L., Minshew, N.J., Just, M.A., 2006. Sentencecomprehension in autism: thinking in pictures with decreased functional con-nectivity. Brain 129, 2484–2493.

Kana, R.K., Keller, T.A., Cherkassky, V.L., Minshew, N.J., Just, M.A., 2009. Atypicalfrontal–posterior synchronization of Theory of Mind regions in autism duringmental state attribution. Soc. Neurosci. 4, 135–152.

Kana, R.K., Keller, T.A., Minshew, N.J., Just, M.A., 2007. Inhibitory control in high-functioning autism: decreased activation and underconnectivity in inhibitionnetworks. Biol. Psychiatry 62, 198–206.

Ke, X., Tang, T., Hong, S., Hang, Y., Zou, B., Li, H., Zhou, Z., Ruan, Z., Lu, Z., Tao, G.,Liu, Y., 2009. White matter impairments in autism, evidence from voxel-basedmorphometry and diffusion tensor imaging. Brain Res. 1265, 171–177.

Keller, T.A., Kana, R.K., Just, M.A., 2006. A developmental study of the structuralintegrity of white matter in autism. Neuroreport 18, 23–27.

Kennedy, D.P., Courchesne, E., 2008. The intrinsic functional organization of the brainis altered in autism. NeuroImage 39, 1877–1885.

Kievit, R.A., Romeijn, J.W., Waldorp, L.J., Wicherts, J.M., Scholte, H.S., Borsboom, D.,2011. Mind the gap: a psychometric approach to the reduction problem. Psychol.Inq. 22, 67–87.

Kleinhans, N.M., Johnson, L.C., Richards, T., Mahurin, R., Greenson, J., Dawson, G.,Aylward, E., 2009. Reduced neural habituation in the amygdala and socialimpairments in autism spectrum disorders. Am. J. Psychiatry 166, 467–475.

Kleinhans, N.M., Richards, T., Sterling, L., Stegbauer, K.C., Mahurin, R., Johnson, L.C.,Greenson, J., Dawson, G., Aylward, E., 2008. Abnormal functional connectivity inautism spectrum disorders during face processing. Brain 131, 1000–1012.

Koshino, H., Carpenter, P.A., Minshew, N.J., Cherkassky, V.L., Keller, T.A., Just,M.A., 2005. Functional connectivity in an fMRI working memory task in high-functioning autism. NeuroImage 24, 810–821.

Koshino, H., Kana, R.K., Keller, T.A., Cherkassky, V.L., Minshew, N.J., Just, M.A., 2008.fMRI investigation of working memory for faces in autism: visual coding andunderconnectivity with frontal areas. Cereb. Cortex 18, 289–300.

Kumar, A., Sundaram, S.K., Sivaswamy, L., Behen, M.E., Makki, M.I., Ager, J., Janisse, J.,Chugani, H.T., Chugani, D.C., 2009. Alterations in frontal lobe tracts and corpuscallosum in young children with autism spectrum disorder. Cereb. Cortex 20,2103–2113.

Lai, M.C., Lombardo, M.V., Chakrabarti, B., Sadek, S.A., Pasco, G., Wheelwright, S.J.,Bullmore, E.T., Baron-Cohen, S., Suckling, J., 2010. A shift to randomness of brainoscillations in people with autism. Biol. Psychiatry 68, 1092–1099.

Lange, N., Dubray, M.B., Lee, J.E., Froimowitz, M.P., Froehlich, A., Adluru, N., Wright,B., Ravichandran, C., Fletcher, P.T., Bigler, E.D., Alexander, A.L., Lainhart, J.E.,2010. Atypical diffusion tensor hemispheric asymmetry in autism. Autism Res.3, 350–358.

Lee, J.E., Bigler, E.D., Alexander, A.L., Lazar, M., DuBray, M.B., Chung, M.K., Johnson,M., Morgan, J., Miller, J.N., McMahon, W.M., Lu, J., Jeong, E.K., Lainhart, J.E., 2007.Diffusion tensor imaging of white matter in the superior temporal gyrus andtemporal stem in autism. Neurosci. Lett. 424, 127–132.

Lee, P.S., Yerys, B.E., Della Rosa, A., Foss-Feig, J., Barnes, K.A., James, J.D., VanMeter,J., Vaidya, C.J., Gaillard, W.D., Kenworthy, L.E., 2009. Functional connectivity ofthe inferior frontal cortex changes with age in children with autism spectrum
disorders: a fcMRI study of response inhibition. Cereb. Cortex 19, 1787–1794.
Léveillé, C., Barbeau, E.B., Bolduc, C., Limoges, E., Berthiaume, C., Chevrier, E., Mottron,L., Godbout, R., 2010. Enhanced connectivity between visual cortex and otherregions of the brain in autism: a REM sleep EEG coherence study. Autism Res. 3,280–285.

Biobeh

L

L

L

L

L

M

M

M

M

M

M

M

N

N

P

P

P

P

P

R

S

S

S

S

S

S

S


iu, Y., Cherkassky, V.L., Minshew, N.J., Just, M.A., 2011. Autonomy of lower-levelperception from global processing in autism: evidence from brain activationand functional connectivity. Neuropsychologia 49, 2105–2111.

o, Y.C., Soong, W.T., Gau, S.S., Wu, Y.Y., Lai, M.C., Yeh, F.C., Chiang, W.Y., Kuo, L.W.,Jaw, F.S., Tseng, W.Y., 2011. The loss of asymmetry and reduced interhemisphericconnectivity in adolescents with autism: a study using diffusion spectrum imag-ing tractography. Psychiatry Res. 192, 60–66.

ord, C., Risi, S., Lambrecht, L., Cook Jr., E.H., Leventhal, B.L., DiLavore, P.C., Pick-les, A., Rutter, M., 2000. The autism diagnostic observation schedule-generic:a standard measure of social and communication deficits associated with thespectrum of autism. J. Autism Dev. Disord. 30, 205–223.

ord, C., Rutter, M., Goode, S., Heemsbergen, J., Jordan, H., Mawhood, L., Schopler,E., 1989. Autism diagnostic observation schedule: a standardized observation ofcommunicative and social behavior. J. Autism Dev. Disord. 19, 185–212.

ord, C., Rutter, M., Le Couteur, A., 1994. Autism Diagnostic Interview – Revised:a revised version of a diagnostic interview for caregivers of individuals withpossible pervasive developmental disorders. J. Autism Dev. Disord. 24, 659–685.

ason, R.A., Williams, D.L., Kana, R.K., Minshew, N., Just, M.A., 2008. Theory of Minddisruption and recruitment of the right hemisphere during narrative compre-hension in autism. Neuropsychologia 46, 269–280.

cAlonan, G.M., Cheung, V., Cheung, C., Suckling, J., Lam, G.Y., Tai, K.S., Yip, L., Mur-phy, D.G., Chua, S.E., 2005. Mapping the brain in autism. A voxel-based MRI studyof volumetric differences and intercorrelations in autism. Brain 128, 268–276.

icheloyannis, S., Vourkas, M., Tsirka, V., Karakonstantaki, E., Kanatsouli, K., Stam,C.J., 2009. The influence of ageing on complex brain networks: a graph theoret-ical analysis. Hum. Brain Mapp. 30, 200–208.

izuno, A., Villalobos, M.E., Davies, M.M., Dahl, B.C., Muller, R.A., 2006. Partiallyenhanced thalamocortical functional connectivity in autism. Brain Res. 1104,160–174.

onk, C.S., Peltier, S.J., Wiggins, J.L., Weng, S.J., Carrasco, M., Risi, S., Lord, C., 2009.Abnormalities of intrinsic functional connectivity in autism spectrum disorders.NeuroImage 47, 764–772.

ostofsky, S.H., Powell, S.K., Simmonds, D.J., Goldberg, M.C., Caffo, B., Pekar, J.J.,2009. Decreased connectivity and cerebellar activity in autism during motortask performance. Brain 132, 2413–2425.

urias, M., Webb, S.J., Greenson, J., Dawson, G., 2007. Resting state cortical connec-tivity reflected in EEG coherence in individuals with autism. Biol. Psychiatry 62,270–273.

oonan, S.K., Haist, F., Muller, R.A., 2009. Aberrant functional connectivity in autism:evidence from low-frequency BOLD signal fluctuations. Brain Res. 1262, 48–63.

oriuchi, M., Kikuchi, Y., Yoshiura, T., Kira, R., Shigeto, H., Hara, T., Tobimatsu, S.,Kamio, Y., 2010. Altered white matter fractional anisotropy and social impair-ment in children with autism spectrum disorder. Brain Res. 1362, 141–149.

aakki, J.J., Rahko, J., Long, X., Moilanen, I., Tervonen, O., Nikkinen, J., Starck, T., Remes,J., Hurtig, T., Haapsamo, H., Jussila, K., Kuusikko-Gauffin, S., Mattila, M.L., Zang, Y.,Kiviniemi, V., 2010. Alterations in regional homogeneity of resting-state brainactivity in autism spectrum disorders. Brain Res. 1321, 169–179.

erez Velazquez, J.L., Barcelo, F., Hung, Y., Leshchenko, Y., Nenadovic, V., Belkas, J.,Raghavan, V., Brian, J., Garcia Dominguez, L., 2009. Decreased brain coordinatedactivity in autism spectrum disorders during executive tasks: reduced long-range synchronization in the fronto-parietal networks. Int. J. Psychophysiol. 73,341–349.

erkins, T., Stokes, M., McGillivray, J., Bittar, R., 2010. Mirror neuron dysfunction inautism spectrum disorders. J. Clin. Neurosci. 17, 1239–1243.

ierce, K., Haist, F., Sedaghat, F., Courchesne, E., 2004. The brain response to person-ally familiar faces in autism: findings of fusiform activity and beyond. Brain 127,2703–2716.

ugliese, L., Catani, M., Ameis, S., Dell’Acqua, F., Thiebaut de Schotten, M., Murphy, C.,Robertson, D., Deeley, Q., Daly, E., Murphy, D.G., 2009. The anatomy of extendedlimbic pathways in Asperger syndrome: a preliminary diffusion tensor imagingtractography study. NeuroImage 47, 427–434.

ubinov, M., Sporns, O., 2010. Complex network measures of brain connectivity:uses and interpretations. NeuroImage 52, 1059–1069.

ahyoun, C.P., Belliveau, J.W., Soulieres, I., Schwartz, S., Mody, M., 2010. Neu-roimaging of the functional and structural networks underlying visuospatial vs.linguistic reasoning in high-functioning autism. Neuropsychologia 48, 86–95.

chmitz, N., Rubia, K., Daly, E., Smith, A., Williams, S., Murphy, D.G., 2006. Neuralcorrelates of executive function in autistic spectrum disorders. Biol. Psychiatry59, 7–16.

chopler, E., Reichler, R.J., DeVellis, R.F., Daly, K., 1988. Toward objective classifica-tion of childhood autism: Childhood Autism Rating Scale (CARS). J. Autism Dev.Disord. 10, 91–103.

chultz, R.T., Gauthier, I., Klin, A., Fulbright, R.K., Anderson, A.W., Volkmar, F., Skud-larski, P., Lacadie, C., Cohen, D.J., Gore, J.C., 2000. Abnormal ventral temporalcortical activity during face discrimination among individuals with autism andAsperger syndrome. Arch. Gen. Psychiatry 57, 331–340.

heikhani, A., Behnam, H., Mohammadi, M.R., Noroozian, M., Mohammadi, M.,2010. Detection of abnormalities for diagnosing of children with autism dis-orders using of quantitative electroencephalography analysis. J. Med. Syst.,doi:10.1007/s10916-010-9560-6 (first published online August 14, 2010).

hih, P., Keehn, B., Oram, J.K., Leyden, K.M., Keown, C.L., Muller, R.A., 2011. Functional
differentiation of posterior superior temporal sulcus in autism: a functionalconnectivity magnetic resonance imaging study. Biol. Psychiatry 70, 270–277.
hih, P., Shen, M., Ottl, B., Keehn, B., Gaffrey, M.S., Muller, R.A., 2010. Atypical networkconnectivity for imitation in autism spectrum disorder. Neuropsychologia 48,2931–2939.

avioral Reviews 36 (2012) 604–625 625

Shukla, D.K., Keehn, B., Lincoln, A.J., Muller, R.A., 2010a. White matter compromiseof callosal and subcortical fiber tracts in children with autism spectrum disor-der: a diffusion tensor imaging study. J. Am. Acad. Child Adolesc. Psychiatry 49,1269–127800.

Shukla, D.K., Keehn, B., Muller, R.A., 2010b. Regional homogeneity of fMRI time seriesin autism spectrum disorders. Neurosci. Lett. 476, 46–51.

Shukla, D.K., Keehn, B., Muller, R.A., 2011a. Tract-specific analyses of diffusion ten-sor imaging show widespread white matter compromise in autism spectrumdisorder. J. Child Psychol. Psychiatry 52, 286–295.

Shukla, D.K., Keehn, B., Smylie, D.M., Muller, R.A., 2011b. Microstructural abnor-malities of short-distance white matter tracts in autism spectrum disorder.Neuropsychologia 49, 1378–1382.

Smith, S.M., Jenkinson, M., Johansen-Berg, H., Rueckert, D., Nichols, T.E., Mackay,C.E., Watkins, K.E., Ciccarelli, O., Cader, M.Z., Matthews, P.M., Behrens, T.E., 2006.Tract-based spatial statistics: voxelwise analysis of multi-subject diffusion data.NeuroImage 31, 1487–1505.

Solomon, M., Ozonoff, S.J., Ursu, S., Ravizza, S., Cummings, N., Ly, S., Carter, C.S., 2009.The neural substrates of cognitive control deficits in autism spectrum disorders.Neuropsychologia 47, 2515–2526.

Sporns, O., Chialvo, D.R., Kaiser, M., Hilgetag, C.C., 2004. Organization, developmentand function of complex brain networks. Trends Cogn. Sci. 8, 418–425.

Sporns, O., Tononi, G., Edelman, G.M., 2000. Connectivity and complexity: therelationship between neuroanatomy and brain dynamics. Neural Netw. 13,909–922.

Stanfield, A.C., McIntosh, A.M., Spencer, M.D., Philip, R., Gaur, S., Lawrie, S.M.,2008. Towards a neuroanatomy of autism: a systematic review and meta-analysis of structural magnetic resonance imaging studies. Eur. Psychiatry 23,289–299.

Sundaram, S.K., Kumar, A., Makki, M.I., Behen, M.E., Chugani, H.T., Chugani, D.C.,2008. Diffusion tensor imaging of frontal lobe in autism spectrum disorder.Cereb. Cortex 18, 2659–2665.

Teipel, S.J., Pogarell, O., Meindl, T., Dietrich, O., Sydykova, D., Hunklinger, U.,Georgii, B., Mulert, C., Reiser, M.F., Moller, H.J., Hampel, H., 2009. Regionalnetworks underlying interhemispheric connectivity: an EEG and DTI study inhealthy ageing and amnestic mild cognitive impairment. Hum. Brain Mapp. 30,2098–2119.

Thakkar, K.N., Polli, F.E., Joseph, R.M., Tuch, D.S., Hadjikhani, N., Barton, J.J., Manoach,D.S., 2008. Response monitoring, repetitive behaviour and anterior cingulateabnormalities in autism spectrum disorders (ASD). Brain 131, 2464–2478.

Thomas, C., Humphreys, K., Jung, K.J., Minshew, N., Behrmann, M., 2011. The anatomyof the callosal and visual-association pathways in high-functioning autism: a DTItractography study. Cortex 47, 863–873.

Turner, K.C., Frost, L., Linsenbardt, D., McIlroy, J.R., Muller, R.A., 2006. Atypicallydiffuse functional connectivity between caudate nuclei and cerebral cortex inautism. Behav. Brain Funct. 2, 34.

Vargas, D.L., Nascimbene, C., Krishnan, C., Zimmerman, A.W., Pardo, C.A., 2005. Neu-roglial activation and neuroinflammation in the brain of patients with autism.Ann. Neurol. 57, 67–81.

Villalobos, M.E., Mizuno, A., Dahl, B.C., Kemmotsu, N., Muller, R.A., 2005. Reducedfunctional connectivity between V1 and inferior frontal cortex associated withvisuomotor performance in autism. NeuroImage 25, 916–925.

Von Stein, A., Sarnthein, J., 2000. Different frequencies for different scales of corticalintegration: from local gamma to long range alpha/theta synchronization. Int. J.Psychophysiol. 38, 301–313.

Wallace, G.L., Dankner, N., Kenworthy, L., Giedd, J.N., Martin, A., 2010. Age-relatedtemporal and parietal cortical thinning in autism spectrum disorders. Brain 133,3745–3754.

Wechsler, D., 1997. Wechsler Adult Intelligence Scale—Third Edition (WAIS-III). ThePsychological Corporation, San Antonio, TX.

Weinstein, M., Ben-Sira, L., Levy, Y., Zachor, D.A., Ben Itzhak, E., Artzi, M., Tarrasch, R.,Eksteine, P.M., Hendler, T., Ben Bashat, D., 2011. Abnormal white matter integrityin young children with autism. Hum. Brain Mapp. 32, 534–543.

Welchew, D.E., Ashwin, C., Berkouk, K., Salvador, R., Suckling, J., Baron-Cohen, S.,Bullmore, E., 2005. Functional disconnectivity of the medial temporal lobe inAsperger’s syndrome. Biol. Psychiatry 57, 991–998.

Weng, S.J., Wiggins, J.L., Peltier, S.J., Carrasco, M., Risi, S., Lord, C., Monk, C.S.,2010. Alterations of resting state functional connectivity in the defaultnetwork in adolescents with autism spectrum disorders. Brain Res. 1313,202–214.

Wiggins, J.L., Peltier, S.J., Ashinoff, S., Weng, S.J., Carrasco, M., Welsh, R.C., Lord, C.,Monk, C.S., 2011. Using a self-organizing map algorithm to detect age-relatedchanges in functional connectivity during rest in autism spectrum disorders.Brain Res. 1380, 187–197.

Williams, D.L., Goldstein, G., Minshew, N.J., 2006. Neuropsychologic functioning inchildren with autism: further evidence for disordered complex information-processing. Child Neuropsychol. 12, 279–298.

Williams, D.L., Minshew, N.J., 2007. Understanding autism and related disorders:what has imaging taught us? Neuroimaging Clin. N. Am. 17, 495–509, ix.

Wilson, T.W., Rojas, D.C., Reite, M.L., Teale, P.D., Rogers, S.J., 2007. Children and ado-lescents with autism exhibit reduced MEG steady-state gamma responses. Biol.Psychiatry 62, 192–197.

Yu, C., Li, J., Liu, Y., Qin, W., Li, Y., Shu, N., Jiang, T., Li, K., 2008. White matter tractintegrity and intelligence in patients with mental retardation and healthy adults.NeuroImage 40, 1533–1541.

Zang, Y., Jiang, T., Lu, Y., He, Y., Tian, L., 2004. Regional homogeneity approach tofMRI data analysis. NeuroImage 22, 394–400.

http://dx.doi.org/10.1007/s10916-010-9560-6

Brain connectivity and high functioning autism: A promising path of research that needs refined models, methodological convergence, and stronger behavioral links

Documents