ArticlePDF AvailableLiterature Review

Abstract and Figures

Background: Although it is well recognized that autism is associated with altered patterns of over- and under-connectivity, specifics are still a matter of debate. Little has been done so far to synthesize available literature using whole-brain electroencephalography (EEG) and magnetoencephalography (MEG) recordings. Objectives: 1) To systematically review the literature on EEG/MEG functional and effective connectivity in autism spectrum disorder (ASD), 2) to synthesize and critically appraise findings related with the hypothesis that ASD is characterized by long-range underconnectivity and local overconnectivity, and 3) to provide, based on the literature, an analysis of tentative factors that are likely to mediate association between ASD and atypical connectivity (e.g., development, topography, lateralization). Methods: Literature reviews were done using PubMed and PsychInfo databases. Abstracts were screened, and only relevant articles were analyzed based on the objectives of this paper. Special attention was paid to the methodological characteristics that could have created variability in outcomes reported between studies. Results: Our synthesis provides relatively strong support for long-range underconnectivity in ASD, whereas the status of local connectivity remains unclear. This observation was also mirrored by a similar relationship with lower frequencies being often associated with underconnectivity and higher frequencies being associated with both under- and over-connectivity. Putting together these observations, we propose that ASD is characterized by a general trend toward an under-expression of lower-band wide-spread integrative processes compensated by more focal, higher-frequency, locally specialized, and segregated processes. Further investigation is, however, needed to corroborate the conclusion and its generalizability across different tasks. Of note, abnormal lateralization in ASD, specifically an elevated left-over-right EEG and MEG functional connectivity ratio, has been also reported consistently across studies. Conclusions: The large variability in study samples and methodology makes a systematic quantitative analysis (i.e. meta-analysis) of this body of research impossible. Nevertheless, a general trend supporting the hypothesis of long-range functional underconnectivity can be observed. Further research is necessary to more confidently determine the status of the hypothesis of short-range overconnectivity. Frequency-band specific patterns and their relationships with known symptoms of autism also need to be further clarified.
Content may be subject to copyright.
RESEARCH ARTICLE
Is functional brain connectivity atypical in
autism? A systematic review of EEG and MEG
studies
Christian O’Reilly
1,2¤
*, John D. Lewis
3
, Mayada Elsabbagh
1,2
1Douglas Mental Health University Institute, 6875 Boulevard Lasalle, Verdun, Canada, 2Department of
Psychiatry, McGill University, 1033 Pine Avenue West, Montreal, QC, Canada, 3McGill Center for Integrative
Neuroscience, Montreal Neurological Institute, McGill University, 3801 University Street, Montre
´al, QC,
Canada
¤Current address: Blue Brain Projet, E
´cole Polytechnique Fe
´de
´rale de Lausanne, Campus Biotech, Chemin
des Mines 9, Geneva, Switzerland
*christian.oreilly@epfl.ch
Abstract
Background
Although it is well recognized that autism is associated with altered patterns of over- and
under-connectivity, specifics are still a matter of debate. Little has been done so far to syn-
thesize available literature using whole-brain electroencephalography (EEG) and magneto-
encephalography (MEG) recordings.
Objectives
1) To systematically review the literature on EEG/MEG functional and effective connectivity
in autism spectrum disorder (ASD), 2) to synthesize and critically appraise findings related
with the hypothesis that ASD is characterized by long-range underconnectivity and local
overconnectivity, and 3) to provide, based on the literature, an analysis of tentative factors
that are likely to mediate association between ASD and atypical connectivity (e.g., develop-
ment, topography, lateralization).
Methods
Literature reviews were done using PubMed and PsychInfo databases. Abstracts were
screened, and only relevant articles were analyzed based on the objectives of this paper.
Special attention was paid to the methodological characteristics that could have created
variability in outcomes reported between studies.
Results
Our synthesis provides relatively strong support for long-range underconnectivity in ASD,
whereas the status of local connectivity remains unclear. This observation was also mirrored
by a similar relationship with lower frequencies being often associated with underconnectiv-
ity and higher frequencies being associated with both under- and over-connectivity. Putting
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 1 / 28
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: O’Reilly C, Lewis JD, Elsabbagh M (2017)
Is functional brain connectivity atypical in autism?
A systematic review of EEG and MEG studies.
PLoS ONE 12(5): e0175870. https://doi.org/
10.1371/journal.pone.0175870
Editor: Alessandro Gozzi, Istituto Italiano di
Tecnologia, ITALY
Received: August 31, 2016
Accepted: March 31, 2017
Published: May 3, 2017
Copyright: ©2017 O’Reilly et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the paper.
Funding: This work is supported by grants from
the Fonds de recherche du Que
´bec – Sante
´(FRQS;
http://www.frqs.gouv.qc.ca/en/), Bourgeois
Foundation, Brain Canada (http://www.
braincanada.ca/), and the Azrieli Foundation (grant
number BC_Azrieli_MIRI_3388; http://www.
azrielifoundation.org/) attributed to ME. The
funders had no role in study design, data collection
and analysis, decision to publish, or preparation of
the manuscript.
together these observations, we propose that ASD is characterized by a general trend
toward an under-expression of lower-band wide-spread integrative processes compensated
by more focal, higher-frequency, locally specialized, and segregated processes. Further
investigation is, however, needed to corroborate the conclusion and its generalizability
across different tasks. Of note, abnormal lateralization in ASD, specifically an elevated left-
over-right EEG and MEG functional connectivity ratio, has been also reported consistently
across studies.
Conclusions
The large variability in study samples and methodology makes a systematic quantitative
analysis (i.e. meta-analysis) of this body of research impossible. Nevertheless, a general
trend supporting the hypothesis of long-range functional underconnectivity can be observed.
Further research is necessary to more confidently determine the status of the hypothesis of
short-range overconnectivity. Frequency-band specific patterns and their relationships with
known symptoms of autism also need to be further clarified.
Introduction
It is well recognized that autism and autism spectrum disorder (ASD)–hereafter used inter-
changeably–is associated with altered patterns of connectivity, compared to neurotypical (NT)
controls. Increased interest in connectivity reflects a shift from understanding the biological
basis of autism as focal brain abnormalities affecting specific systems towards an overall pat-
tern of brain reorganization. Moreover, new evidence on early development of white matter
tracks suggests that connectivity could be among the earliest markers of autism, with initial
signs emerging within the first year of life [13].
Although the autistic brain was initially hypothesized as exhibiting a pattern of overall
underconnectivity [4,5] or by long-range underconnectivity and local overconnectivity [6], a
more subtle mixture of hypo- and hyper-connectivity is now emphasized [7]. Often, findings
remain unreplicated and conclusions divergent regarding the nature of altered connectivity in
autism. Several reasons may explain the differences in findings and conclusions including con-
ceptual (e.g., definitions, theoretical models), methodological (e.g., measurement modalities
and paradigms, participant characteristics), or analytical (e.g., quality control and processing
pipelines).
Previous literature reviews have partially addressed questions about connectivity in autism.
These reviews have predominantly focused on structural connectivity using diffusion imaging
[810] and correlated activity using functional magnetic resonance imaging (fMRI) [1012].
In contrast, other aspects of connectivity including functional and effective connectivity of
electrophysiological activity reported in electroencephalography (EEG) and magnetoencepha-
lography (MEG) have not been the focus of systematic synthesis (however, see [13] for a narra-
tive synthesis of coherence in EEG resting state). Yet, such syntheses are paramount in getting
a clear view of the relationship between brain connectivity and autism considering that 1) dif-
ferent recording modalities can provide contrasting points of view on mechanisms altering
brain connectivity (e.g., synaptic functions, degree of myelination, inhibitory/excitatory bal-
ance, network properties) and 2) their results are not necessarily in good agreement [14].
Because of their high temporal resolution and their direct relationship with neuronal activity
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 2 / 28
Competing interests: The authors have declared
that no competing interests exist.
(as opposed to a proxy such as hemodynamic), EEG/MEG connectivity analyses can provide
valuable information about dynamic activation and deactivation of functional networks. Fur-
ther, these can be observed for different oscillatory frequencies depending on the role of each
network in integration versus segregation of information, on top-down versus bottom-up
propagation of signals, and the tasks or functions they support. Synthesizing evidence about
altered functional network connectivity in autism is essential for establishing a coherent theo-
retical account of the pathophysiology of the condition. Thus, our goal is to fill knowledge
gaps by comprehensively reviewing literature on EEG/MEG functional and effective brain
connectivity in autism, with a focus on factors influencing over versus under connectivity. We
begin with an overview of relevant connectivity concepts and measurement approaches.
Connectivity: A multi-faceted concept
Brain connectivity is a broad, multi-faceted concept [15]. In human neuroscience, connectivity
can refer to physical interconnection of brain regions through bundles of axons (structural/
anatomical connectivity), to statistical dependencies (e.g., correlation, coherence, consistency
in phase-lag) between time series of cerebral activity in different brain regions (functional con-
nectivity), or to causal interactions between brain regions (directed/effective connectivity) [16].
Structural/anatomical connectivity is generally assessed with deterministic or probabilistic
tractography of diffusion weighted images recorded using magnetic resonance imaging (MRI)
scanners. The two other types (i.e., functional and effective) are assessed mainly using electro-
magnetic (e.g., electroencephalography (EEG), magnetoencephalography (MEG), local field
potentials (LFP), spike trains), hemodynamic (e.g., functional MRI (fMRI) or near infra-red
spectroscopy (NIRS)) and, to some extent, nuclear recordings (e.g., positron emission tomog-
raphy (PET), single-photon emission computed tomography (SPECT)).
Although the term connectivity is often used interchangeably in the literature to denote any
or all of these variants, different measures can show surprisingly little agreement [14,17]. Aside
from measurement issues and biases, a few reasons may explain the discrepancy among mea-
sures. One of these reasons is the complexity of the relationship between structural and func-
tional connectivity. For example, neural networks have an intricate structure of excitatory and
inhibitory neurons forming local microcircuits (i.e., “a minimal number of interacting neu-
rons that can collectively produce a functional output” [18]), which synaptic connectivity can
either amplify or attenuate measures of functional connectivity. How this micro-connectivity
impacts on macroscopic structural and functional connectivity is unclear. Functional connec-
tivity can also be modulated by factors independent from structural connectivity (e.g., synaptic
depression, properties of sensory afferent signals) and can appear through indirect paths, not
structurally connected by a direct track but functionally coordinated by an intermediate struc-
ture (e.g., cortico-thalamo-cortical pathways) not considered by direct structural connectivity.
Similarly, different modalities of functional connectivity show complex interdependencies.
For example, the relationship between electrical and hemodynamic activity has been shown to
depend on frequency and spatial scale, with particular EEG rhythms generating region-depen-
dent variations in blood oxygen levels and glucose metabolism [19,20]. Cross-frequency
dependencies can even be observed between recording modalities. For example, slow hemody-
namic rhythms are known to be correlated with the amplitude of fast gamma-band EEG/MEG
activity [21,22].
Furthermore, it is worth remembering that functional connectivity is generally based on
similarity of signals observed between pair of regions and, in most cases, no control of the
potential contribution of a third sources is made. This situation may be improved by multivar-
iate approaches [23], but it seems unlikely to be completely controlled, particularly for cortical
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 3 / 28
activity initiated by subcortical structures (e.g., thalamo-cortical pathways) since such third
sources are generally hidden (i.e., not measured) and are therefore difficult to estimate and
mitigate [24]. The situation is much different when assessing structural connectivity by mea-
suring the area of a fiber bundle section linking two regions. In the latter case, there is no
potential hidden third source confound. Therefore, different forms of connectivity need to be
treated as different, yet related, constructs. Integrating knowledge from EEG/MEG with other
functional (e.g., fMRI, PET) or structural (e.g., tractography) modalities is expected to be a
fruitful avenue because of possible complementarity between modalities, but it must be per-
formed very cautiously in view of many potential pitfalls.
Connectivity hypotheses in ASD
The large amount of literature on connectivity in autism yields multiple distinct hypotheses
about the nature of over- and/or under-connectivity. We considered three classes of inter-
related hypotheses regarding connection length, topological specificity, and developmental
effects.
Over- and under-connectivity in relation with connection length. A popular hypothesis
considers autism to be characterized by long-range underconnectivity potentially combined
with local overconnectivity [6,2527]. This hypothesis is predominantly supported by struc-
tural and functional MRI and post-mortem immunocytochemistry investigations. For the
structural part, this pattern can be explained by ASD-related abnormalities at the cellular level:
long-range structural underconnectivity can result from a degradation of fiber bundles
[1,2840], with many studies reporting decreased fractional anisotropy and/or higher mean
diffusivity in ASD for the superior longitudinal fasciculus, occipitofrontal fasciculus, unci-
nate fasciculus, inferior longitudinal fasciculus, cingulum, and corpus callosum [41];
local structural overconnectivity can result from a decrease of apoptosis, axonal pruning,
and dendritic degradation, and from an increase of neurogenesis [42].
However, local connectivity may often be obfuscated by long-distance connections (e.g., in
presence of long-range crossing fibers in diffusion imaging). In such cases, long-range under-
connectivity makes local connectivity more clearly observable without meaning that there is a
true increase in local connectivity (i.e., there is an augmentation of the relative number of local
connections because of a loss of long-distance connections, but there is no augmentation in
their absolute number).
For the functional part, many fMRI studies reported long-range underconnectivity in ASD,
whereas local overconnectivity has been reported less consistently [26,43,44]. In a recent
review of the literature, 26 out of 33 fMRI studies were shown to report reduction or loss of–
sometime local but most often long-range–connectivity in ASD [41]. The prefrontal cortex
and the posterior cingulate cortex were most often shown to exhibit long-range underconnec-
tivity. Other regions (e.g., precuneus, anterior cingulate cortex, superior temporal gyrus, poste-
rior superior temporal sulcus, anterior insula, parietal lobule) showed primarily long-range
underconnectivity, but were also associated in some studies with long-range overconnectivity.
Various experimental observations have been proposed as potential correlates or causes for
local functional overconnectivity, such as smaller but more numerous cortical neurons and
mini-columns, which might indicate a bias toward local processing [45]. It may also be caused
by a higher excitatory/inhibitory ratio favoring local interactions [46,47], for example through
a deficient GABAergic signaling [48]. Appropriate GABAergic activity is also important for
normal operation of local circuitry such as appropriate functional segregation of mini-column
through lateral inhibition provided by GABAergic basket cells [49]. It is also involved in
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 4 / 28
generation of gamma-band activity through parvalbumin-expressing fast-spiking interneurons
[50]. Activity in the gamma-band is associated mostly with local computation [51,52], it is
involved in many processes (e.g., perceptual binding and selective attention [53]) showing
alteration in ASD, and abnormalities in this frequency band have been reported consistently
enough to be proposed as being a marker of ASD [54].
Finally, related to the hypothesis concerning long-range versus short-range connectivity,
the operational definition in the reviewed literature remains elusive. This is particularly prob-
lematic in EEG/MEG, since these modalities do not have a good spatial accuracy and compari-
son of nearby pairs of sensors (i.e., local connections) is confounded with volume conduction
(see the discussion for more on this topic). One possible definition–and the one we use in this
review–for EEG/MEG long-range connectivity is that it reflects inter-lobar or inter-hemi-
spheric connections. However, we also interpret evidence showing a graded response of con-
nectivity with respect to distance, for example, when the connectivity is correlated with the
inter-sensor distance.
Topological specificity. Evidence for altered connectivity in various brain regions comes
from investigations of brain structure (MRI, diffusion imaging, post-mortem immunocytochem-
istry, etc.). ASD-related abnormalities have been documented in the frontal lobe [25,55,56],
including abnormal organization of neurons and microglial cells [57]. Evidence from fMRI sup-
ports a pattern of underconnectivity from frontal to other brain regions [58]; i.e., lower frontal to
parietal connectivity [59] and reduced antero-posterior connectivity [58,60]. However, these stud-
ies were conducted in adults and may therefore be characterizing a cascade effect that appears
over development. For example, underconnectivity in 24-month-olds with ASD have been shown
to be predominantly in occipital regions, with important abnormalities in temporal lobes, but
almost no abnormalities in frontal areas [61].
Abnormal connectivity between the occipital lobe and the other regions is also often
reported; an observation that might be related with structural and functional abnormalities in
processing of visual input in ASD [6266]. In a large database of resting-state fMRI recordings,
underconnectivity was also found in all lobes, but particularly for the temporal, whereas over-
connectivity was mainly affecting connectivity with subcortical structures, particularly for con-
nections linking the thalamus and the globus pallidus to the primary parietal sensorimotor
regions [67].
Contrary to hypotheses about deficiency in specific regions, an alternative hypothesis sug-
gests an overall non-topographically-specific alteration of brain connectivity in ASD. Support
for this hypothesis is based on evidence of a more randomly connected brain in ASD [68],
which results in a cross-interaction between the degree of connectivity of brain regions and
the diagnosis of ASD; i.e., ASD shows an increase (respectively, a decrease) in connectivity for
pairs of regions which display low (respectively, high) connectivity in controls [69].
Developmental hypotheses. The typical developmental course of connectivity remains
poorly understood but its determinants (e.g., pruning, myelination) and its indexes (e.g., frac-
tional anisotropy, magnetization transfer ratio) appear to follow an inverted U-shape trajec-
tory during maturation, which progress from posterior to anterior regions and from primary
to association cortical areas. These maturational patterns appear to differ in autism from very
early in development [1,2]. Diffusion imaging shows early elevated fractional anisotropy (1.8–
3.3 year-old range in [70]; 1.5–5.8 year-old range in [71]; elevated at 6 month-old followed by
reduction to below controls at 24 month-old in [1]) followed by a reduction below neurotypi-
cal values later in life (7–33 year-old in [28]; 14.6 ±3.4 year-old in [72]). This observation sug-
gests a possible age-related inversion of trends in measured connectivity (i.e., from initial
overconnectivity to later underconnectivity) potentially due to slight differences in develop-
mental trajectories. Interesting relationships might be hypothesized linking this potential
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 5 / 28
initial overconnectivity with the early overgrowth of the brain in ASD [55,73] and the early
maturation of white matter tracts previously reported in toddlers and young children with
ASD [1,70,71,74]. Because early stages of typical development involve initial structural over-
connectivity followed by a pruning of connections in the maturing brain [75], differences in
maturation rates can be expected to modulate the degree of connectivity. For example, a delay
in the onset of axonal remodeling could drive brain overgrowth, yielding an abnormally large
brain and therefore abnormal long-distance connections. Such connections are more likely to
be pruned given that this process is driven by competition for neurotrophins [76], and the
increased conduction delays and cellular costs associated with longer fibers puts them at a dis-
advantage in this competition. Thus, a delay in the onset of axonal remodeling could drive
brain overgrowth and thereby favor short-distance connections over long-distance connec-
tions via typical developmental mechanisms [77,78].
Differences in developmental trajectories of brain connectivity observed with EEG/MEG
might also be exacerbated if excitatory and inhibitory neurons mature at different speeds (e.g.,
a slower rate of reduction of intra-cellular chloride ionic concentration during development
would result in delayed transition of GABAergic neurons from excitatory to inhibitory [48]).
Such an altered development would shift the normal evolution of the excitatory/inhibitory
ratio (already known to be abnormal in ASD [47]). In turn, it is likely that this imbalance
would influence the degree of functional connectivity in cortical pyramidal neurons, the cell
population generally considered as the main source for observable EEG/MEG activity. Simi-
larly, differences in maturation of feedforward versus feedback pathways might be responsible
for some transient developmental aberration in overall connectivity. Since feedback connec-
tions develop later than feedforward [79] and ASD is a developmental disorder, feedback pro-
jections could be preferentially altered.
Regarding which brain region might show the largest developmental differences, fMRI
studies have reported the default mode network as being particularly vulnerable to ASD
[80,81]. The associated brain regions would show a slower maturation than in neurotypical
controls. Among the structures involved in this network, the impact of slower maturation is
likely to be more prominent for the frontal lobe since it is one of the last structures to mature
in the brain, a process that continues up to the mid-twenties [82].
Scope of the review
Taken together, these different hypotheses demonstrate the complexity of the impact of ASD
on connectivity, which is most probably characterized by both generalized atypicality (e.g.,
long-range underconnectivity, increased randomness in connection patterns) and more local-
ized abnormalities (i.e., specific deficits in some nuclei or brain regions) that develops during
the first years of life.
Although the current body of evidence related to connectivity comes mostly from fMRI
[1012] or from structural approaches using diffusion imaging [810], there has been an
increase in complementary methods. Specifically, EEG/MEG functional and effective connec-
tivity can reveal qualitatively different results which are important to our understanding of
how the brain wiring is altered in autism. Evidence synthesis on that topic is currently missing.
Therefore, in this paper, we comprehensively review the literature on EEG/MEG functional
and effective brain connectivity in autism, against the different hypotheses described above,
particularly focusing on factors influencing overconnectivity versus underconnectivity.
Although it is currently unclear if local connectivity in EEG/MEG should be assessed by com-
paring the activity between pairs/groups of sensors or alternatively using single point measure-
ments across a whole group of sensors (e.g., through spectral power), this review is limited to
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 6 / 28
studies measuring connectivity using the former approach. Complementary review of other
novel methods to characterize brain activity in ASD such as quantitative EEG can be consulted
elsewhere (e.g., see [83,84]).
Methods
Our review focuses on original research assessing functional and effective connectivity as mea-
sured by EEG or MEG. Searches on PubMed and PsychInfo executed on March 6
th
2016 using
the research string “(EEG OR MEG OR electroencephaloOR magnetoencephalo) AND con-
nectivity AND autis” returned 102 and 67 hits respectively. Only journal and conference
papers published in English were reviewed.
Studies were retained if they presented original experimental findings on participants with
an ASD phenotype, including individuals diagnosed at any age as well as studies comparing
low versus high-risk infants defined by the presence or absence of an older sibling diagnosed
with ASD. One study correlating ASD traits in a neurotypical population with differences in
connectivity features was also included.
We excluded articles discussing functional or effective brain connectivity without reporting
new experimental results on this topic. We also excluded articles reviewing neurofeedback
approaches using connectivity measurements, methodological articles proposing new
approaches for assessing brain connectivity, and studies inferring on connectivity only
through indirect measurements, e.g., EEG/MEG complexity, ERP timing, spectral power, or
local inter-trial synchronization.
Appraisal of methodological quality of included studies was conducted by one of the
authors (COR). This led to exclusion of one study due to probable methodological issues
related to multiple comparisons, confounders, and mismatching statistics [85], and of a second
study because of a too small sample (i.e., two subjects per group [86]). A third study [87] was
retained but interpreted with caution because of inappropriate control of multiple statistical
comparisons. In that study, the authors addressed the problem of multiple comparisons by
computing, for each node of their graphs, an average t-value from 48 leave-one-subject-out t-
tests. However, if standard statistical hypotheses are valid (normality, absence of outliers, etc.),
this “bootstrapped” evaluation of the mean t-value will converge toward the t-value for the
whole sample (i.e., the uncorrected t-value). However, t-tests are still being computed at every
node resulting in N independent statistical tests with an alpha-threshold at 0.05, which will
provide up to 5% of false positives under the null hypothesis. The “leave-one-out” approach
does nothing to correct for these multiple comparisons.
A total of 52 papers were retained, 31 using EEG [88118] and 21 using MEG [87,119
138]. The selection process is depicted in Fig 1. For each study, one of the authors (COR)
extracted methodological dimensions that potentially impacted results and conclusions,
such as:
connectivity metrics;
recording parameters (sampling frequency, reference electrode, electrode grid density);
experimental paradigms (e.g., resting-state, event-related experiments, sleep studies);
sample characteristics (age, gender, IQ, diagnostic confirmation method, sample size).
The wide variability in study characteristics along these methodological dimensions pre-
cluded a meta-analysis to address the main objectives of the review. Instead, we synthesized
and critically appraised findings of the over-arching hypothesis that ASD is characterized by
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 7 / 28
long-range underconnectivity and local overconnectivity (as a valid generalization of a more
complex phenomenon), and systematically reported results related to factors that can impact
this hypothesis (exceptions to the generalization; development, topography, lateralization).
Fig 1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart describing the paper selection process.
https://doi.org/10.1371/journal.pone.0175870.g001
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 8 / 28
Results
Synthesis per methodological dimensions
We first synthesized included studies across several methodological dimensions that may
account for some inconsistencies in including connectivity metrics, recording parameters,
experimental paradigm, and sample characteristics. Outcome of this analysis can be found as
supplementary documents S2 File along with tables (see S1 and S2 Tables) summarizing the
information extracted for included studies (split by EEG vs. MEG).
ASD is characterized by long-range underconnectivity; Local
overconnectivity remains uncertain
As reviewed in the introduction, a prominent hypothesis is that autism is characterized by
local overconnectivity and long-range underconnectivity [25,26]. EEG/MEG studies provide
partial support for this hypothesis with some studies showing both phenomena in ASD popu-
lation [90,114]. This distribution of under/overconnectivity is also correlated with the presence
of autistic traits in the neurotypical population [89].
Overall, the case for long-range underconnectivity is well supported. This is particularly
clear for the case of inter-hemispheric connections, for which a decrease in EEG/MEG func-
tional connectivity has been reported in many studies [93,94,96,100,107,114,134]. A smaller
number of studies report mixed long-range overconnectivity and underconnectivity [97] or
increased long-range connectivity in infants [113] and adolescents [116,117], in adult rapid-
eye movement (REM) sleep [108], and during a picture-naming task in adults [120]. Also,
some indication for long-range overconnectivity has been found during slow wave sleep [92],
although the effect was marginal (p-values = [0.1, 0.05]) when corrected for multiple compari-
sons. In sum, while ASD is characterized by a general long-range underconnectivity, it is likely
that this pattern is modulated by task requirements or developmental processes.
Local overconnectivity in ASD is less robustly established. A few studies report local over-
connectivity [89,90,114,117,139], while others found local underconnectivity in ASD
[96,97,125,130,140] or a mix of both patterns [112]. Similarly, indirect support for local over-
connectivity is provided by reports of an enhanced local synchronization [141], which can be
associated with higher local functional connectivity (i.e., functional connectivity is correla-
tional in nature and assesses how the activity between two regions is similar, which often
means synchronized). There are, however, many theoretical and practical issues related to the
definition and the reliable measurement of local connectivity in EEG/MEG (see the Discussion
section) which make it difficult to reach definitive conclusions.
The hypothesized pattern of long-range underconnectivity and short-range overconnectiv-
ity is paralleled by the general trends observed as a function of frequency (see Fig 2). Overall,
evidence suggests underconnectivity in ASD at lower frequencies (delta to beta bands) with
potential overconnectivity in higher bands. This tendency is particularly well illustrated in rest-
ing-state by the work of Ye and collaborators [138]. These general trends seem rather robust
despite the substantial methodological variability across studies, e.g., paradigms, samples, and
connectivity measures (S1 and S2 Tables). This pattern of results is different from reported
EEG power abnormalities in autism, which show a U-shape curve with increased low-fre-
quency (delta, theta) and high-frequency (beta, gamma) activity and reduced middle-ranged
frequencies (alpha) in ASD [13].
General underconnectivity in lower frequency bands has been observed as a function of the
number of autistic traits in a neurotypical sample [89] as well as for all age groups of ASD par-
ticipants (e.g., in infants and young children [91], in adolescents [93], and in adults [94]) and
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 9 / 28
for a wide range of paradigms (e.g., resting-state [96], event-related [100], and sleep [105]),
suggesting that it is associated with neurodevelopmental abnormalities that are not limited to a
specific brain region or state [93]. Evidence of overconnectivity in high-frequency bands is
more scarce but accumulating, particularly in MEG studies (see Fig 2).
Tentative relationships can be proposed between frequency bands, the scale of connectivity,
and the degree of connectivity. The modulation of over vs. under connectivity depending on the
frequency band is consistent with slower oscillators involving more neurons in larger volumes
[52]. Such a relationship is to be expected from “wiring economy”. Indeed, the coordination of
higher frequency activity across regions requires faster communication. Therefore, for any given
distance between two regions, larger (and costlier) axons are required to coordinate faster fre-
quencies. In this context, it is more efficient to bias connectivity such that high-frequency band
Fig 2. Summary of band-specific increases versus decreases in EEG/MEG functional connectivity in ASD (compared to NT
controls). The “N” column list the total size of the sample (i.e., sum of participants in all groups). Frequency is varying along the x-
axis, from 1 Hz to 120 Hz.
https://doi.org/10.1371/journal.pone.0175870.g002
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 10 / 28
activity is shared locally, whereas slower frequency bands are used for long-range interaction.
Thus, we expect high-frequencies to be preferentially associated with more localized processes
and activity in lower bands to be preferentially associated with more wide-spread integrative pro-
cesses. Indeed, integrative top-down processes (i.e., processes integrating a priori knowledge
about the world with incoming signals from the senses to generate a percept) involving long-
range connections are often associated with slower rhythms (delta, theta, alpha), whereas syn-
chronization of local cortical networks through bottom-up processes (i.e., processes modifying
the internal representation of the world to minimize its mismatch with information from the
senses) tends to be associated with faster frequencies (beta, gamma) [142,143]. Such a generaliza-
tion, however, will need further investigation in view of several exceptions (e.g., the existence of
large-axon long-range fast-spiking interneurons synchronizing high-frequency rhythms between
distant regions [144]). Further, both reports of MEG underconnectivity in high-frequency bands
are associated with long-range connections [87,134], with one of these studies explicitly limiting
its investigation to long-range inter-hemispheric connections [134].
An alternative hypothesis from graph theory: A more uniform altered
connectivity in ASD
EEG and MEG connectivity studies using graph analysis generally report autism to be associ-
ated with sub-optimal network properties, such as less clustering, larger characteristic path,
reduced eigenvector centrality (a measure of the importance of nodes as communication
hubs), and an architecture less typical of small-world networks [8991,114,133,135,136].
Small-world networks are thought to be striking an optimal balance between integration and
segregation, making them particularly efficient. This topography is present in a wide range of
contexts, such as in social networks or flight networks, but also in neural networks [145]. It is
characterized by each node having a relatively small number of neighbours but being able to
reach any other node by only a small number of steps (i.e., although they tend to form small
cliques, each node is separated from every other node by only a few levels of separation, thanks
to a relatively small number of “hub” nodes providing between-clique connectivity).
Neural networks in ASD have been shown to be more resilient because of their greater
homogeneity and their more uniform architecture [114]. This, in turn, results in a less optimal
balance between local specialization (segregation) and global integration [145]. Findings from
network analysis seems consistent with the hypothesis of a more randomly connected brain. It
has been supported as supported by EEG/MEG studies showing smaller patches (more local-
ized, less blended) of increased MEG signal complexity in controls compared to ASD [122], a
much more disorganized pattern of connectivity in ASD [119], and more redundant networks
in ASD (i.e., more randomness in a network increases redundancy and resilience but decreases
efficiency and specialization) [114].
However, these analyses have been performed on graph topologies, which preserve very little
relationship with spatial properties of brain activity. Specifically, edges of these graphs are not
weighted by the length of the connections and the angles between the edges are unspecified.
Therefore, the spatial arrangement of the nodes is lost and the results cannot be reconciled with
the findings reviewed above on the impact of connection length on over/underconnectivity.
Factors modulating connectivity patterns
Although the hypothesized pattern of over and under connectivity in autism is a valuable con-
ceptual generalization, many factors may modulate connectivity at a finer scale reflecting con-
text-dependent modulations in connectivity. Below we consider the factors that were prominent
in the reviewed of the literature.
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 11 / 28
Lateralization. Many cognitive processes display brain lateralization, language being a
classical example. Since symptoms associated with autism are associated with lateralized cogni-
tive functions (e.g., executive functions and language), many studies have looked for and
found atypicalities in lateralization of brain functions in autism [146]. In the reviewed litera-
ture, many studies reported function- or structure-dependent EEG/MEG connectivity abnor-
malities related to ASD. The most evident are reports of abnormal lateralization of functional
connectivity, with an elevated left-over-right EEG and MEG functional connectivity ratio in
ASD.
For example, studying gamma-band connectivity in relation with face processing, a left-
ward (instead of the normal rightward) lateralization emerging around one year of life was
found [103]. The authors hypothesized that this abnormal lateralization might be related with
potential differences in face recognition in autism [147,148] because it suggests a more fea-
tural-based (processing and recognition of local features in a visual scene; typical of left hemi-
sphere) than configural-based (integration of all the parts of a stimulus in a coherent percept;
typical of right hemisphere) face recognition processes [149]. Also, in one-year-old infants at
high-risk for autism (HRA) that have later been diagnosed with ASD, an increase in the alpha-
band connectivity predominant in the left fronto-centro-parietal region during video watching
was found in one study [113], although a second study did not report any hemispheric differ-
ences between HRA infants and low-risk controls [115].
Similarly, an elevated long-range fronto-posterior functional connectivity in children with
ASD was observed prominently in the left-hemisphere [116]. Such increased left-hemisphere
coherence has also been observed in low-functioning ASD children when compared with
high-functioning ASD children [117].
A rightward reduction of beta-band connectivity (phase lag index) between the occipital
lobe and frontal, temporal, and parietal areas in a numerosity task (i.e., participants had to esti-
mate the number of dots either distributed to form an animal shape or randomly positioned)
was also reported in ASD [119]. The authors suggested impairment in the capacity of using a
global interpretation style (Gestalt perception; right hemisphere) in ASD and potentially a
shape recognition strategy relying too heavily on a local processing of visual information (left
hemisphere) [150]. This hypothesis seems to be coherent with the observed decrease in long-
range connectivity in ASD given that such connectivity is thought to be important for global
integration of widespread brain activity. Also consistent with an elevated left-over-right func-
tional connectivity ratio in ASD is the report of increased connectivity (Granger causality) 1)
from the left inferior frontal gyrus to the left fusiform area in high-beta and low-gamma fre-
quencies and 2) from the left superior temporal gyrus to the left occipital lobe in the beta and
gamma bands during a picture naming task [120]. Another study showed reduced right hemi-
sphere temporal-central alpha-band coherence in ASD adolescents compared to neurotypical
controls [101].
In an event-related paradigm in which emotional stimuli evoke bilateral activation of the
insula in neurotypical adults [151], a less bilateral pattern (i.e., more lateralized) was found in
ASD with significantly lower connectivity (phase lag index) of the right (but not the left) insula
with areas including the right fusiform, right inferior temporal gyrus, and superior frontal
regions in ASD [133].
Using photic stimulation, a leftward predominance in connectivity increased at stimulated
frequencies has been observed in ASD, with a larger group difference at higher frequencies
[106,107]. Although these studies did not report such asymmetry in their resting-state record-
ings, Coben et al. [96] found an increase in relative theta power but a reduced absolute beta
power over the right hemisphere for the autistic group during resting-state. These authors
argue that such an excess in theta power might be related to abnormal brain functioning in
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 12 / 28
ASD, based on similar observations made in children with attention deficit/hyperactivity dis-
order [152], learning disabilities [153], and mental retardation [154]. Leftward asymmetry and
altered EEG power lateralization was also reported [155]. Still in resting-state, Murias et al.
[112] reported theta-band overconnectivity in ASD within left hemisphere frontal and tempo-
ral cortex, whereas Barttfeld et al. [90] observed an increase in delta-band local connectivity in
lateral frontal electrodes, which was particularly salient in left hemisphere.
During the REM phase of sleep, more important increases in coherence were also noted in
the left hemisphere, particularly for the delta band [108]. Kikuchi et al. [127] reported the
more contradictory results with an increased rightward connectivity lateralization in gamma
band in a sample of 3–7 years old infants with ASD, but not on a 5–7 years old sample [128]
during passive video watching. Although they found no between-group differences in laterality
in that latter sample, within the ASD sample, they showed that more rightward laterality in
connectivity was correlated with better performances in reading and in pattern reasoning.
This associate a leftward lateralization with more severe symptoms, as shown in the other stud-
ies, and is consistent with the idea that such spatial skills typically relies more on the right
hemisphere [156].
In sum, a consistent pattern of increased left-over-right EEG and MEG connectivity ratio
has been observed in ASD. This pattern may well reflect specific perceptual characteristics of
this population in which local components are more strongly processed (segregation) relative
to global relationships between components (integration; Gestalt style). This altered processing
style can be hypothesized to be a consequence of the observed drop in long-range connectivity.
A lack of proper long-range connectivity may not provide an adequate substrate for the nor-
mal development of the rightward more global and integrative style of processing. Compensa-
tory development of more leftward local-processing style may, however, still be possible if
shorter-distance connectivity is not as severely affected by the disease. It is also relevant to note
that an enhanced EEG power has been consistently reported in the left hemisphere across all
frequencies (reviewed in [13]) which may be a confounding factors in studies not controlling
for potential impact of an increase in power on the resulting increase in connectivity (e.g., as
can be seen in the case of coherence computed using a common reference [157]).
Topography of connectivity differences in ASD. Cognitive processes that are known to
be atypical in ASD are, to some extent, localized to specific brain regions. Thus, there is an
interest in investigating the topography of connectivity abnormalities to further associate
altered patterns of connectivity with observable phenotypes and symptoms. Among the
reviewed studies, some discussed general topological differences at sensor-level, whereas oth-
ers used more precise analyses of cortical sources to localize abnormalities. For example, in
MEG, a reduced coherence between the frontal eye field (a region involved in voluntary eye
movements) and dorsal anterior cingulate cortex in an ASD population in a saccade/anti-sac-
cade task has been shown [124]. Abnormal underconnectivity of the fusiform face area with
various other structures (left precuneus, right inferior temporal gyrus, and superior frontal
regions) in tasks involving face stimuli has also been reported using cross-frequency coupling
[125] and phase lag index [133].
These results are examples showing that more subtle patterns of connectivity are most cer-
tainly region- and function-dependent. Consistent with the topological hypotheses reviewed
in the introduction, a large number of studies included in S1 and S2 Tables report involvement
of frontal [90,93,108,112,113,121,137,138,141] or occipital regions
[90,96,100,108,112,121,122,138,158], although reports of significant differences in connectivity
can be found between virtually all pairs of brain regions. The variability in experimental para-
digms, analyses, and sample characteristics prevents establishing clearer generalizations of
these findings. It is worth noting though that frontal and occipital regions have the longest
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 13 / 28
inter-hemispheric connections, which may account for the prevalence of observed undercon-
nectivity in these regions if long-range connections are preferentially affected by ASD. Also,
studies in infants are under-represented in this literature, which may cause abnormalities
emerging at this age to be under-represented. Further, some regions of the brain may also be
poorly covered by small-grid EEG studies, such as inferior regions of temporal lobes.
Development. Differences in EEG/MEG functional connectivity are emergent rather than
static over development [103,115]. However, a consistent portrait of how these connectivity
differences emerge is yet to be established. Righi et al. [115] reported a decrease of connectivity
(coherence) in the gamma band for HRA infants, a trend that was more pronounced for the
portion of HRA infants which later developed ASD, whereas Orekhova et al. [113] reported
increased alpha-band connectivity (debiased weighted phase lag index) in HRA infants who
later developed ASD. Relying on reported correlation between EEG alpha-band coherence and
structural integrity of white matter [159], these authors relate this alpha-band hyper-connec-
tivity in toddlers with the frequently reported alpha-band hypo-connectivity in adolescence/
adulthood by highlighting the abnormal trajectory of white matter maturation in ASD: early
maturation of white matter tracks in toddlers and young children [1,70,71,74] followed by
slowing of white matter increase in toddlerhood [1] and later childhood [160,161] ending in
predominant hypo-connectivity in adulthood [26]. Thus, patterns of connectivity differences
between ASD and NT participants should not be thought of as being static over time. Although
more corroboration is still needed, it might well be captured by a general early hyper-connec-
tivity followed by a regression toward hypo-connectivity later in development.
Developmental change continues to impact connectivity patterns later in development.
Using resting-state MEG recordings from 6–21 years old individuals, Kitzbichler et al. [130]
found in their NT control group an initially strong beta, theta, and delta-band connectivity
involving frontal regions. This connectivity decreases later with maturation, presumably evolv-
ing toward more specific interconnections. This developmental change was not observed in
ASD participants who initially started with a low frontal connectivity and stayed at this level.
Discussion
Summary of the main observations reported in the reviewed literature
Our systematic review of a large body of evidence suggests that ASD is characterized by a pat-
tern of EEG/MEG functional connectivity that is in general more randomly organized, with
abnormal connectivity often involving frontal or occipital regions–at least in adult samples–
although abnormal connectivity patterns have been reported in almost every region. Abnormal
lateralization of activity in resting-state and in specific tasks seems to be typical, with a very
systematic report of elevated left-over-right EEG and MEG functional connectivity ratio in
ASD. Both abnormal intra- and inter-hemispheric connectivity can be observed with a general
trend toward underconnectivity, but with probable local or condition-specific (task, spectral
band, brain region) overconnectivity. Underconnectivity can most reliably be observed in
lower frequency bands (hypothesized to be preferentially involved in long-range integrative
networks) whereas a stronger tendency for overconnectivity can be observed (see Fig 2) in
high-frequency ranges (hypothesized to be generally associated with more localized processes).
These findings appear to hold despite substantial variability across several methodological
dimensions including recording characteristics and analytic approaches. More investigation
on the relationship between over/underconnectivity and EEG/MEG frequencies is neverthe-
less needed to corroborate these conclusions. This study might further benefit from looking at
potential modulation across regions of high-frequency activity by the phase of slower rhythms
(i.e., between-region phase-amplitude coupling) [162], how these different frequencies relate
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 14 / 28
to feed-forward and feed-back processes at play in information integration, and how these dif-
ferent observations may or may not be integrated in a theory-driven accounts, such as the pre-
dictive-coding framework [163172].
Heterogeneity in ASD
Despite these general patterns related to categorical diagnosis of ASD, several findings high-
light the importance of considering variability in the ASD phenotype related to connectivity.
For example, individuals with an Asperger’s diagnosis can be differentiated from those with
other ASDs relatively accurately (92.3% in [98]) based on their patterns of connectivity.
Although the authors did not provide detailed information on how groups compared with
respect to ADOS scores, they used a large sample of ASD children (N = 430), which is likely to
overlap with the Asperger’s sample (N = 26) regarding the severity of the condition. Similarly,
ASD with Fragile X Syndrome or with a de novo chromosomal mutation causing agenesis of
corpus callosum is another example of a subgroup of individuals with ASD that might be
clearly defined from a clinical point of view. These persons reported a normal level of attention
to details, with the preserved ability to appreciate the whole rather than a preoccupation with
patterns or parts [132]. Thus, connectivity features shown to correlate with greater attention to
details in ASD (e.g., eye-open resting-state alpha power and coherence in posterior regions
[110]) are probably under-represented in this subgroup. Thus, it might become increasingly
important to subgroup or control for the different conditions that are grouped under the ASD
umbrella to better understand the impact of this heterogeneity on observed connectivity pat-
terns and, hopefully, to reconcile some contradictions in the literature.
Further, abnormalities in EEG/MEG functional connectivity increase with increasing
symptoms severity in autism. Measures of connectivity abnormalities reported in surveyed
papers has been correlated with the presence of autistic traits in the NT population [89] and
with the severity of autism symptoms in ASD samples: ADOS scores were shown to correlate
with connectivity strength in gamma and beta bands [130], with theta-band coherence in left-
anterior and right-posterior regions [129], with phase-amplitude coupling in the fusiform face
area (for face-processing task) [125], and local functional connectivity based on phase-locking
[126]; SRS scores correlated with regional average complexity and connectivity node strength
[87]; visual reasoning and reading abilities correlated with lateralization of coherence in the
parieto-temporal regions in gamma band; imagination measure on the parent-report adult
AQ correlated with global connectivity in the right superior temporal gyrus [132].
Limitations
The conclusions derived from our systematic review are limited by a few potential methodo-
logical confounders. These are discussed hereafter.
Interpretation of EEG/MEG functional connectivity. The hypothesis underlying every
approach used to measure functional connectivity in EEG/MEG is that the estimated connec-
tivity (i.e., the similarity between two or more time series, computed through correlation,
coherence, or similar measures) is proportional to physiological connectivity between the
brain areas generating these time series. There are nevertheless pitfalls in interpretation of con-
nectivity measures due to our limited understanding of their underlying physiology, e.g., hid-
den sources, differential sensor sensitivity for different kind of synaptic activity, etc.
Statistical power. The low statistical power in the reviewed studies may be misleading.
With a sample size typical of these studies (we take N = 15 per group for this example), the
power of a two-tail two-sample t-test with a significance level at 0.05 is only 0.26 for a size-
effect considered as medium (Cohen’s d at 0.5). That is, three studies out of four will report
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 15 / 28
negative results for medium-sized differences in brain connectivity. Controlling for multiple
comparisons further reduces the power of these tests such that only very strong effects may be
reliably detected in most studies. When applied to the Fig 2, this implies that potential
counter-evidence regarding underconnectivity can only be provided by studies reporting over-
connectivity (and vice-versa) for comparable frequency bands, and, importantly, not by stud-
ies reporting inconclusive results.
Publication bias: As mentioned previously, the context of individual studies may impact
on obtained results. When results are grouped, these contexts are partly lost and this can pro-
duce a bias in the overall conclusions. That is, studies looking for specific information (e.g.,
gamma activity over long-distance connections) may report specific findings (e.g., under-con-
nectivity) which would not be the main outcome if they had investigated a broader context
(e.g., connectivity of gamma activity in general). This creates a potential publication bias
where topic of high interest is generally more scrutinized, resulting in the publication of more
positive findings.
Head and brain size. The significant bias in head and brain size between ASD and NT
participants may result in biases impacting the distance between EEG sensors and the proper-
ties of electrical signal propagation from the cortex to the scalp. Differences in brain size (not
head size) may also have some effect on the signal-to-noise ratio in MEG since the cortical
surface is in average closer to the sensors for larger brains. This effect is expected to be more
important in children since brain size differences between ASD and NT are larger at this age.
Entering these factors as co-variates in statistical analyses is advisable to control their impact
on group comparisons of brain connectivity. Although brain size might not be readily accessi-
ble in studies not including MRI scans, standard metrics for the head size are easily measured
(e.g., standard head measuring procedures are defined in any 10/20 EEG electrode placement
manual).
Volume conduction and assessment of short-range connectivity in EEG/MEG. We
noted that, compared to long-range functional connectivity, studies of local connectivity are
more scarce. This pattern might be related to intrinsic limitations of EEG and MEG recording
modalities for evaluating short-range connectivity, particularly in coherence analyses performed
at sensor level. To record an EEG/MEG potential, a large pool (around 50,000) of synchronously
activated cells with parallel apical dendrites (i.e., pyramidal cells) spanning a cortical patch of
40–200 mm
2
is required [173]. This has two implications. First, it means that simple power anal-
ysis could be used to assess local connectivity–with higher EEG/MEG power indicating larger
local connectivity–if between-subject comparisons were not biased by various factors such as the
conductivities of the skull and other tissues that can cause a general offset of power measure-
ments. Although EEG/MEG power studies were out of the scope of the current review, it is rele-
vant to note that a recent review of this literature has shown no consistent pattern of altered
EEG/MEG power in ASD. It suggest that the literature is more supportive of an increased vari-
ability of EEG/MEG responses in ASD [174] (see, however, [175] for a recent study that chal-
lenges this theory).
Second, the localization on the scalp of such cortical sources is intrinsically limited by the
spatial extent of the cell assembly necessary to generate a recordable potential. More impor-
tantly, this theoretical spatial resolution is affected by a smearing of the electrical activity as it
travels from the cortical (or sub-cortical) sources to the sensors. For EEG, potentials are reach-
ing sensors through omnidirectional volume conduction. For that reason, they spread over
larger territories than the initial source area and create severe biases when estimating coher-
ence between close sensors (<10 cm) [176]. This phenomenon can induce false between-
group differences if autism is correlated with differences in global electromagnetic properties
of the head tissues, such as impedances of the different layer of tissues (scalp, skull, dura mater,
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 16 / 28
etc.). Observed differences in extra-axial fluid in infants who develop ASD [177] is one such
factor that may be confounding coherence results. The effect of field spread due to volume
conduction is less severe in MEG, but is still present. At sensor-level, different approaches
(e.g., phase lag index, imaginary part of coherency) have been devised to remove zero-lag
activity between different sensors on account that such an activity can be associated with vol-
ume conduction. Such approaches, however, discard any potential physiological connectivity
with zero-lag that can emerge in systems with feedback-loops such as neural networks [178].
Further, a zero-lag synchronization can be expected in neural oscillators generating large-scale
EEG/MEG oscillations [179]. Such zero-lag functional connectivity can be observed experi-
mentally for example in area 17 of the cat visual cortex where initial zero-lag interhemispheric
synchronization of neuronal activity can be disrupted by sectioning the corpus callosum [180].
Further, the long-standing hypothesis that volume conduction propagates instantaneously for
the frequency range of interest in EEG analysis–an hypothesis depending on the validity of the
quasistatic approximation of the Maxwell equations for volume conduction–is challenged by
recent experimental work showing propagation delays of volume conducted EEG waves [181].
One way to partly mitigate the problem of volume conduction would be to compute con-
nectivity on EEG sources estimated using more accurate electromagnetic model of the propa-
gation of electrical activity. Although the effect of field spread cannot be completely resolved
by the current level of sophistication of source-reconstruction algorithms, such an approach
should definitely be used to supplement sensor-level analyses as it reduces the impact of vol-
ume conduction and helps better link electrical activity with brain regions [182]. Further,
using cortical sources would allow to compute the distance of connections along the cortical
sheet, which would provide a much better estimated of distances for correlational analyses
than the bird fly distance between sensors (i.e., two sensors above two nearby gyri may be
close-by but connections must be significantly longer to follow the cortical sheet forming a sul-
cus between these two gyri).
Discriminant validity. Our review is limited to EEG/MEG functional connectivity in
ASD, but future work should also help clarify cross-cutting issue common to multiple neuro-
developmental conditions. It remains to be established if the pattern of connectivity we
observed in ASD is specific to the condition or alternatively reflects a more general pattern of
changes common to a broad group of neurodevelopmental disorders [183].
Conclusion and future directions
This review illustrates the large heterogeneity of both the methods and the results of studies
investigating brain electrophysiological connectivity in ASD. Some of this variability might be
reduced by further improving the methods adopted (e.g. using source reconstruction, better
controlling for ASD phenotypes).
Research on electrophysiological functional connectivity in autism has been pioneered by
EEG coherence analyses performed on small sensor grids, which provided crude connectivity
assessment between very large brain regions (e.g., lobes). Currently, high-density grids are
available in EEG and are included by default in every MEG system. Furthermore, algorithms
for the estimation of cortical (and even sub-cortical [184]) sources from recorded activity have
been developed and are now routinely used in a large body of studies presenting both time-
and spatially-resolved brain activations and connectivity. Only one EEG study has benefited
from this potential, whereas half of MEG studies did. Results from this latter subset of
researches are generally more convincing, not only because source reconstruction helps in
mitigating confounders such as volume conduction [182], but also because they associate
observed activity with specific brain structures. Because they are non-invasive, cheap, and can
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 17 / 28
be made widely available for clinical applications, small-grid sensor-level analyses have some
potential for the development of biomarkers for diagnostic purposes and potentially to close a
biofeedback loop in experimental therapeutic approaches. However, high-density grids com-
bined with sensor-level and source-level analyses provide a much more fertile ground for
building and testing hypotheses and theories on functional connectivity in autism.
Studies examining individual variability within ASD and across neurodevelopmental con-
ditions remain very sparse and future studies need to pay more attention to mapping connec-
tivity onto phenotypic differences. The use of functional connectivity features for diagnostic
application is also relevant since high accuracy (85–95%) has been reported by many indepen-
dent research teams [88,97,102,125,126,135]. However, how such biomarkers would perform
in infants for an early diagnosis is still an open question and fraught with several pragmatic
and ethical complications [185].
A clearer theoretical foundation is necessary to efficiently establishes the role of connectiv-
ity length with respect to over/under-connectivity using EEG/MEG functional connectivity.
Strong theoretical grounding can help address outstanding questions in Table 1.
Further work is also needed to better understand the complex interactions between fre-
quency bands, brain regions, physiology of EEG/MEG oscillators (e.g., roles of specific cell
types, neurotransmitters, ion channels; see [186]), and how they relate to different cognitive
processes. Specific task-/event-related protocol will need to be devised to disentangle these dif-
ferent dimensions in a principled way. This knowledge is instrumental in designing future
studies that could link together ASD symptoms, brain processes, and connectivity
abnormalities.
Finally, given the nonlinear evolution of brain properties (e.g., size, white matter integrity,
connectivity, etc.), the developmental evolution of these properties strikes us as a very impor-
tant area of investigation since these changes are generating confusion in interpretation of the
current literature. Aside from helping to reconcile apparent contradictions in the literature, a
better understanding of how developmental factors induce ASD-related brain changes early in
development would provide invaluable insights on the pathogenesis of ASD. Performing such
analyses in a multi-modal framework may also further our understanding of the dynamics of
ASD-related abnormalities in brain connectivity and help resolve some of the apparent contra-
dictions arising when comparing results across modalities.
Table 1. Research questions needing clear answers to provide a solid foundation for linking connec-
tion length versus EEG/MEG functional connectivity in autism.
1. How could short and long-range connectivity be clearly defined based on unambiguous biological
substrates, e.g., using anatomical concepts which can be directly measured/imaged such as cortical
columns, gyri, cerebral lobes, etc.?
2. Are short and long-range connectivity distinct concepts (i.e., physiologically different) or is connectivity
better captured as a dimension? Accordingly, should long versus short-range connectivity be assessed as a
categorical problem (e.g., using ANOVA) or as a continuous one (e.g., using correlational analysis)?
3. What are the most appropriate methods to measure short and long-range connections in EEG/MEG?
What confounds need to be more systematically controlled for (e.g., head circumference, brain volume)?
4. Should EEG/MEG power (a point measurement) or a connectivity metric (i.e., a two-point measurement)
be used for the assessment of local activity?
5. Considering that volume conduction might have a non-zero-lag component and genuine connectivity is
likely to have a zero-lag component, how should volume conduction be controlled for when measuring local
connectivity?
6. How are frequency bands associated with EEG/MEG functional connection length and over/under-
connectivity in autism given the current knowledge about the role of these different frequency bands in top-
down/bottom-up integration/segregation and given the pathophysiology models accounting for autism
symptomatology?
https://doi.org/10.1371/journal.pone.0175870.t001
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 18 / 28
Supporting information
S1 File. PRISMA 2009 checklist.
(DOC)
S2 File. Detailed synthesis per methodological dimensions.
(DOCX)
S1 Table. Included papers investigating EEG functional and/or effective connectivity in
ASD.
(DOCX)
S2 Table. Included papers investigating MEG functional and/or effective connectivity in
ASD.
(DOCX)
Acknowledgments
The authors want to acknowledge the contribution of Mai Khalil who has provided guidance
regarding the methodology of the systematic review.
Author Contributions
Conceptualization: CO JL ME.
Funding acquisition: ME.
Investigation: CO.
Methodology: CO ME.
Project administration: ME.
Supervision: ME.
Visualization: CO.
Writing – original draft: CO.
Writing – review & editing: CO JL ME.
References
1. Wolff JJ, Gu H, Gerig G, Elison JT, Styner M, Gouttard S, et al. Differences in white matter fiber tract
development present from 6 to 24 months in infants with autism. Am J Psychiatry. 2012; 169: 589–
600. https://doi.org/10.1176/appi.ajp.2011.11091447 PMID: 22362397
2. Wolff JJ, Gerig G, Lewis JD, Soda T, Styner MA, Vachet C, et al. Altered corpus callosum morphology
associated with autism over the first 2 years of life. Brain. England; 2015; 138: 2046–2058. https://doi.
org/10.1093/brain/awv118 PMID: 25937563
3. Jin Y, Wee C-Y, Shi F, Thung K-H, Yap P-T, Shen D. Identification of Infants at Risk for Autism Using
Multi-parameter Hierarchical White Matter Connectomes. Machine learning in medical imaging. MLMI
(Workshop). 2015. pp. 170–177.
4. Hughes JR. Autism: the first firm finding = underconnectivity? Epilepsy Behav. 2007; 11: 20–24.
https://doi.org/10.1016/j.yebeh.2007.03.010 PMID: 17531541
5. Just MA, Cherkassky VL, Keller TA, Minshew NJ. Cortical activation and synchronization during sen-
tence comprehension in high-functioning autism: evidence of underconnectivity. Brain. 2004; 127:
1811–1821. https://doi.org/10.1093/brain/awh199 PMID: 15215213
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 19 / 28
6. Belmonte MK, Allen G, Beckel-Mitchener A, Boulanger LM, Carper RA, Webb SJ. Autism and abnor-
mal development of brain connectivity. J Neurosci. United States; 2004; 24: 9228–9231. https://doi.
org/10.1523/JNEUROSCI.3340-04.2004 PMID: 15496656
7. Kana RK, Uddin LQ, Kenet T, Chugani D, Mu¨ller R-A. Brain connectivity in autism. Front Hum Neu-
rosci. Frontiers Media S.A.; 2014; 8: 349. https://doi.org/10.3389/fnhum.2014.00349 PMID: 24917800
8. Anagnostou E, Taylor MJ. Review of neuroimaging in autism spectrum disorders: what have we
learned and where we go from here. Mol Autism. 2011; 2: 4. https://doi.org/10.1186/2040-2392-2-4
PMID: 21501488
9. McFadden K, Minshew N. Evidence for Dysregulation of Axonal Growth and Guidance in the Etiology
of ASD. Front Hum Neurosci. 2013; 7.
10. Mu¨ller R-A. Anatomical and Functional Connectivity in Autism Spectrum Disorders. In: Patel VB, Pre-
edy VR, Martin CR, editors. Comprehensive Guide to Autism. Springer New York; 2014. pp. 49–75.
11. Thai NJ, Longe O, Rippon G. Disconnected brains: what is the role of fMRI in connectivity research?
Int J Psychophysiol. 2009; 73: 27–32. Available: http://www.ncbi.nlm.nih.gov/pubmed/19530277
PMID: 19530277
12. Uddin LQ, Supekar K, Lynch CJ, Khouzam A, Phillips J, Feinstein C, et al. Salience network-based
classification and prediction of symptom severity in children with autism. JAMA Psychiatry. 2013; 70:
869–879. https://doi.org/10.1001/jamapsychiatry.2013.104 PMID: 23803651
13. Wang J, Barstein J, Ethridge LE, Mosconi MW, Takarae Y, Sweeney JA. Resting state EEG abnor-
malities in autism spectrum disorders. J Neurodev Disord. 2013; 5: 24. https://doi.org/10.1186/1866-
1955-5-24 PMID: 24040879
14. Reid AT, Lewis J, Bezgin G, Khundrakpam B, Eickhoff SB, McIntosh AR, et al. A cross-modal, cross-
species comparison of connectivity measures in the primate brain. Neuroimage. 2015; 125: 311–331.
https://doi.org/10.1016/j.neuroimage.2015.10.057 PMID: 26515902
15. Horwitz B. The elusive concept of brain connectivity. Neuroimage. 2003; 19: 466–470. Available:
http://www.ncbi.nlm.nih.gov/pubmed/12814595 PMID: 12814595
16. Friston KJ. Functional and effective connectivity in neuroimaging: A synthesis. Hum Brain Mapp.
Wiley; 1994; 2: 56–78.
17. Jones SE, Beall EB, Najm I, Sakaie KE, Phillips MD, Zhang M, et al. Low consistency of four brain con-
nectivity measures derived from intracranial electrode measurements. Front Neurol. 2014; 5: 272.
https://doi.org/10.3389/fneur.2014.00272 PMID: 25566178
18. Silberberg G, Grillner S, LeBeau FEN, Maex R, Markram H. Synaptic pathways in neural microcircuits.
Trends Neurosci. 2005; 28: 541–551. https://doi.org/10.1016/j.tins.2005.08.004 PMID: 16122815
19. Nunez PL, Silberstein RB. On the relationship of synaptic activity to macroscopic measurements:
does co-registration of EEG with fMRI make sense? Brain Topogr. 2000; 13: 79–96. PMID: 11154104
20. Pineda JA. The functional significance of mu rhythms: translating “seeing” and “hearing” into “doing.”
Brain Res Brain Res Rev. 2005; 50: 57–68. https://doi.org/10.1016/j.brainresrev.2005.04.005 PMID:
15925412
21. Lachaux JP, Fonlupt P, Kahane P, Minotti L, Hoffmann D, Bertrand O, et al. Relationship between
task-related gamma oscillations and BOLD signal: new insights from combined fMRI and intracranial
EEG. Hum Brain Mapp. 2007; 28: 1368–1375. https://doi.org/10.1002/hbm.20352 PMID: 17274021
22. Nir Y, Fisch L, Mukamel R, Gelbard-Sagiv H, Arieli A, Fried I, et al. Coupling between neuronal firing
rate, gamma LFP, and BOLD fMRI is related to interneuronal correlations. Curr Biol. 2007; 17: 1275–
1285. https://doi.org/10.1016/j.cub.2007.06.066 PMID: 17686438
23. Blinowska KJ. Review of the methods of determination of directed connectivity from multichannel
data. Med Biol Eng Comput. 2011; 49: 521–529. https://doi.org/10.1007/s11517-011-0739-x PMID:
21298355
24. Bastos AM, Schoffelen J-M. A Tutorial Review of Functional Connectivity Analysis Methods and Their
Interpretational Pitfalls [Internet]. Frontiers in Systems Neuroscience. 2016. p. 175. Available: http://
journal.frontiersin.org/article/10.3389/fnsys.2015.00175 https://doi.org/10.3389/fnsys.2015.00175
PMID: 26778976
25. Courchesne E, Pierce K. Why the frontal cortex in autism might be talking only to itself: local over-con-
nectivity but long-distance disconnection. Curr Opin Neurobiol. 2005; 15: 225–230. https://doi.org/10.
1016/j.conb.2005.03.001 PMID: 15831407
26. Vissers ME, X Cohen M, Geurts HM. Brain connectivity and high functioning autism: A promising path
of research that needs refined models, methodological convergence, and stronger behavioral links.
Neurosci Biobehav Rev. 2012; 36: 604–625. https://doi.org/10.1016/j.neubiorev.2011.09.003 PMID:
21963441
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 20 / 28
27. Picci G, Gotts SJ, Scherf KS. A theoretical rut: revisiting and critically evaluating the generalized
under/over-connectivity hypothesis of autism. Dev Sci. England; 2016; 19: 524–549. https://doi.org/
10.1111/desc.12467 PMID: 27412228
28. Alexander AL, Lee JE, Lazar M, Boudos R, DuBray MB, Oakes TR, et al. Diffusion tensor imaging of
the corpus callosum in Autism. Neuroimage. 2007; 34: 61–73. https://doi.org/10.1016/j.neuroimage.
2006.08.032 PMID: 17023185
29. Brito AR, Vasconcelos MM, Domingues RC, Hygino da Cruz LC Jr., Rodrigues Lde S, Gasparetto EL,
et al. Diffusion tensor imaging findings in school-aged autistic children. J Neuroimaging. 2009; 19:
337–343. https://doi.org/10.1111/j.1552-6569.2009.00366.x PMID: 19490374
30. Casanova MF, El-Baz A, Elnakib A, Switala AE, Williams EL, Williams DL, et al. Quantitative analysis
of the shape of the corpus callosum in patients with autism and comparison individuals. Autism. 2011;
15: 223–238. https://doi.org/10.1177/1362361310386506 PMID: 21363871
31. Egaas B, Courchesne E, Saitoh O. Reduced size of corpus callosum in autism. Arch Neurol. 1995; 52:
794–801. PMID: 7639631
32. Hardan AY, Minshew NJ, Keshavan MS. Corpus callosum size in autism. Neurology. 2000; 55: 1033–
1036. PMID: 11061265
33. Hong S, Ke X, Tang T, Hang Y, Chu K, Huang H, et al. Detecting abnormalities of corpus callosum
connectivity in autism using magnetic resonance imaging and diffusion tensor tractography. Psychia-
try Res. 2011; 194: 333–339. https://doi.org/10.1016/j.pscychresns.2011.03.009 PMID: 22047729
34. Keary CJ, Minshew NJ, Bansal R, Goradia D, Fedorov S, Keshavan MS, et al. Corpus callosum vol-
ume and neurocognition in autism. J Autism Dev Disord. 2009; 39: 834–841. https://doi.org/10.1007/
s10803-009-0689-4 PMID: 19165587
35. Keller TA, Kana RK, Just MA. A developmental study of the structural integrity of white matter in
autism. Neuroreport. 2007; 18: 23–27. https://doi.org/10.1097/01.wnr.0000239965.21685.99 PMID:
17259855
36. Manes F, Piven J, Vrancic D, Nanclares V, Plebst C, Starkstein SE. An MRI study of the corpus callo-
sum and cerebellum in mentally retarded autistic individuals. J Neuropsychiatry Clin Neurosci. 1999;
11: 470–474. https://doi.org/10.1176/jnp.11.4.470 PMID: 10570760
37. Piven J, Bailey J, Ranson BJ, Arndt S. An MRI study of the corpus callosum in autism. Am J Psychia-
try. 1997; 154: 1051–1056. Available: http://www.ncbi.nlm.nih.gov/pubmed/9247388 https://doi.org/
10.1176/ajp.154.8.1051 PMID: 9247388
38. Stanfield AC, McIntosh AM, Spencer MD, Philip R, Gaur S, Lawrie SM. Towards a neuroanatomy of
autism: a systematic review and meta-analysis of structural magnetic resonance imaging studies. Eur
Psychiatry. 2008; 23: 289–299. https://doi.org/10.1016/j.eurpsy.2007.05.006 PMID: 17765485
39. Travers BG, Adluru N, Ennis C, Tromp do PM, Destiche D, Doran S, et al. Diffusion tensor imaging in
autism spectrum disorder: a review. Autism Res. 2012; 5: 289–313. https://doi.org/10.1002/aur.1243
PMID: 22786754
40. Vidal CN, Nicolson R, DeVito TJ, Hayashi KM, Geaga JA, Drost DJ, et al. Mapping corpus callosum
deficits in autism: an index of aberrant cortical connectivity. Biol Psychiatry. 2006; 60: 218–225.
https://doi.org/10.1016/j.biopsych.2005.11.011 PMID: 16460701
41. Rane P, Cochran D, Hodge SM, Haselgrove C, Kennedy DN, Frazier JA. Connectivity in Autism: A
Review of MRI Connectivity Studies. Harv Rev Psychiatry. United States; 2015; 23: 223–244. https://
doi.org/10.1097/HRP.0000000000000072 PMID: 26146755
42. Piven J, Arndt S, Bailey J, Andreasen N. Regional brain enlargement in autism: a magnetic resonance
imaging study. J Am Acad Child Adolesc Psychiatry. 1996; 35: 530–536. https://doi.org/10.1097/
00004583-199604000-00020 PMID: 8919716
43. Muller RA, Shih P, Keehn B, Deyoe JR, Leyden KM, Shukla DK. Underconnected, but how? A survey
of functional connectivity MRI studies in autism spectrum disorders. Cereb Cortex. 2011; 21: 2233–
2243. https://doi.org/10.1093/cercor/bhq296 PMID: 21378114
44. Wass S. Distortions and disconnections: disrupted brain connectivity in autism. Brain Cogn. 2011; 75:
18–28. https://doi.org/10.1016/j.bandc.2010.10.005 PMID: 21055864
45. Casanova MF, El-Baz A, Vanbogaert E, Narahari P, Trippe J. Minicolumnar width: Comparison
between supragranular and infragranular layers. J Neurosci Methods. 2009; 184: 19–24. https://doi.
org/10.1016/j.jneumeth.2009.07.011 PMID: 19616026
46. Rubenstein JL, Merzenich MM. Model of autism: increased ratio of excitation/inhibition in key neural
systems. Genes Brain Behav. 2003; 2: 255–267. PMID: 14606691
47. Nelson SB, Valakh V. Excitatory/Inhibitory Balance and Circuit Homeostasis in Autism Spectrum Dis-
orders. Neuron. United States; 2015; 87: 684–698. https://doi.org/10.1016/j.neuron.2015.07.033
PMID: 26291155
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 21 / 28
48. Cellot G, Cherubini E. GABAergic signaling as therapeutic target for autism spectrum disorders. Front
Pediatr. 2014; 2: 70. https://doi.org/10.3389/fped.2014.00070 PMID: 25072038
49. Lund JS, Angelucci A, Bressloff PC. Anatomical Substrates for Functional Columns in Macaque Mon-
key Primary Visual Cortex. Cereb Cortex. 2003; 13: 15–24. PMID: 12466211
50. Sohal VS, Zhang F, Yizhar O, Deisseroth K. Parvalbumin neurons and gamma rhythms enhance corti-
cal circuit performance. Nature. 2009; 459: 698–702. https://doi.org/10.1038/nature07991 PMID:
19396159
51. Kopell N, Ermentrout GB, Whittington MA, Traub RD. Gamma rhythms and beta rhythms have differ-
ent synchronization properties. Proc Natl Acad Sci U S A. 2000; 97: 1867–1872. PMID: 10677548
52. von Stein A, Sarnthein J. Different frequencies for different scales of cortical integration: from local
gamma to long range alpha/theta synchronization. Int J Psychophysiol. 2000; 38: 301–313. PMID:
11102669
53. Fries P. Neuronal gamma-band synchronization as a fundamental process in cortical computation.
Annu Rev Neurosci. 2009; 32: 209–224. https://doi.org/10.1146/annurev.neuro.051508.135603
PMID: 19400723
54. Rojas DC, Wilson LB. Gamma-band abnormalities as markers of autism spectrum disorders. Biomark
Med. 2014; 8: 353–368. https://doi.org/10.2217/bmm.14.15 PMID: 24712425
55. Courchesne E, Campbell K, Solso S. Brain growth across the life span in autism: age-specific changes
in anatomical pathology. Brain Res. 2011; 1380: 138–145. https://doi.org/10.1016/j.brainres.2010.09.
101 PMID: 20920490
56. Zikopoulos B, Barbas H. Changes in prefrontal axons may disrupt the network in autism. J Neurosci.
2010; 30: 14595–14609. https://doi.org/10.1523/JNEUROSCI.2257-10.2010 PMID: 21048117
57. Morgan JT, Chana G, Abramson I, Semendeferi K, Courchesne E, Everall IP. Abnormal microglial-
neuronal spatial organization in the dorsolateral prefrontal cortex in autism. Brain Res. 2012; 1456:
72–81. https://doi.org/10.1016/j.brainres.2012.03.036 PMID: 22516109
58. Just MA, Keller TA, Malave VL, Kana RK, Varma S. Autism as a neural systems disorder: a theory of
frontal-posterior underconnectivity. Neurosci Biobehav Rev. 2012; 36: 1292–1313. https://doi.org/10.
1016/j.neubiorev.2012.02.007 PMID: 22353426
59. Just MA, Cherkassky VL, Keller TA, Kana RK, Minshew NJ. Functional and anatomical cortical under-
connectivity in autism: evidence from an FMRI study of an executive function task and corpus callo-
sum morphometry. Cereb Cortex. 2007; 17: 951–961. https://doi.org/10.1093/cercor/bhl006 PMID:
16772313
60. Cherkassky VL, Kana RK, Keller TA, Just MA. Functional connectivity in a baseline resting-state net-
work in autism. Neuroreport. 2006; 17: 1687–1690. https://doi.org/10.1097/01.wnr.0000239956.
45448.4c PMID: 17047454
61. Lewis JD, Evans AC, Pruett JR, Botteron K, Zwaigenbaum L, Estes A, et al. Network inefficiencies in
autism spectrum disorder at 24 months. Transl Psychiatry. 2014; 4: e388. https://doi.org/10.1038/tp.
2014.24 PMID: 24802306
62. Amaral DG, Schumann CM, Nordahl CW. Neuroanatomy of autism. Trends Neurosci. 2008; 31: 137–
145. https://doi.org/10.1016/j.tins.2007.12.005 PMID: 18258309
63. Batty M, Meaux E, Wittemeyer K, Roge B, Taylor MJ. Early processing of emotional faces in children
with autism: An event-related potential study. J Exp Child Psychol. 2011; 109: 430–444. https://doi.
org/10.1016/j.jecp.2011.02.001 PMID: 21458825
64. Hyde KL, Samson F, Evans AC, Mottron L. Neuroanatomical differences in brain areas implicated in
perceptual and other core features of autism revealed by cortical thickness analysis and voxel-based
morphometry. Hum Brain Mapp. 2010; 31: 556–566. https://doi.org/10.1002/hbm.20887 PMID:
19790171
65. Jemel B, Mimeault D, Saint-Amour D, Hosein A, Mottron L. VEP contrast sensitivity responses reveal
reduced functional segregation of mid and high filters of visual channels in autism. J Vis. 2010; 10: 13.
66. Vandenbroucke MW, Scholte HS, van Engeland H, Lamme VA, Kemner C. A neural substrate for
atypical low-level visual processing in autism spectrum disorder. Brain. 2008; 131: 1013–1024. https://
doi.org/10.1093/brain/awm321 PMID: 18192288
67. Di Martino A, Yan C-G, Li Q, Denio E, Castellanos FX, Alaerts K, et al. The autism brain imaging data
exchange: towards a large-scale evaluation of the intrinsic brain architecture in autism. Mol Psychiatry.
2014; 19: 659–667. https://doi.org/10.1038/mp.2013.78 PMID: 23774715
68. Lewis JD, Theilmann RJ, Townsend J, Evans AC. Network efficiency in autism spectrum disorder and
its relation to brain overgrowth. Front Hum Neurosci. 2013; 7: 845. https://doi.org/10.3389/fnhum.
2013.00845 PMID: 24368901
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 22 / 28
69. Hahamy A, Behrmann M, Malach R. The idiosyncratic brain: distortion of spontaneous connectivity
patterns in autism spectrum disorder. Nat Neurosci. 2015; 18: 302–309. https://doi.org/10.1038/nn.
3919 PMID: 25599222
70. Ben Bashat D, Kronfeld-Duenias V, Zachor DA, Ekstein PM, Hendler T, Tarrasch R, et al. Accelerated
maturation of white matter in young children with autism: a high b value DWI study. Neuroimage.
2007; 37: 40–47. https://doi.org/10.1016/j.neuroimage.2007.04.060 PMID: 17566764
71. Weinstein M, Ben-Sira L, Levy Y, Zachor DA, Ben Itzhak E, Artzi M, et al. Abnormal white matter integ-
rity in young children with autism. Hum Brain Mapp. 2011; 32: 534–543. https://doi.org/10.1002/hbm.
21042 PMID: 21391246
72. Barnea-Goraly N, Kwon H, Menon V, Eliez S, Lotspeich L, Reiss AL. White matter structure in autism:
preliminary evidence from diffusion tensor imaging. Biol Psychiatry. 2004; 55: 323–326. PMID:
14744477
73. Courchesne E, Pierce K, Schumann CM, Redcay E, Buckwalter JA, Kennedy DP, et al. Mapping early
brain development in autism. Neuron. 2007; 56: 399–413. https://doi.org/10.1016/j.neuron.2007.10.
016 PMID: 17964254
74. Billeci L, Calderoni S, Tosetti M, Catani M, Muratori F. White matter connectivity in children with autism
spectrum disorders: a tract-based spatial statistics study. BMC Neurol. 2012; 12: 148. https://doi.org/
10.1186/1471-2377-12-148 PMID: 23194030
75. LaMantia AS, Rakic P. Axon overproduction and elimination in the corpus callosum of the developing
rhesus monkey. J Neurosci. 1990; 10: 2156–2175. PMID: 2376772
76. van Ooyen A, Willshaw DJ. Competition for neurotrophic factor in the development of nerve connec-
tions. Proc Biol Sci. 1999; 266: 883–892. https://doi.org/10.1098/rspb.1999.0719 PMID: 10380678
77. Lewis JD, Theilmann RJ, Fonov V, Bellec P, Lincoln A, Evans AC, et al. Callosal fiber length and inter-
hemispheric connectivity in adults with autism: brain overgrowth and underconnectivity. Hum Brain
Mapp. 2013; 34: 1685–1695. https://doi.org/10.1002/hbm.22018 PMID: 22359385
78. Lewis JD, Elman JL. Growth-related neural reorganization and the autism phenotype: a test of the
hypothesis that altered brain growth leads to altered connectivity. Dev Sci. 2008; 11: 135–155. https://
doi.org/10.1111/j.1467-7687.2007.00634.x PMID: 18171375
79. Batardiere A, Barone P, Knoblauch K, Giroud P, Berland M, Dumas AM, et al. Early specification of
the hierarchical organization of visual cortical areas in the macaque monkey. Cereb Cortex. 2002; 12:
453–465. PMID: 11950763
80. Washington SD, Gordon EM, Brar J, Warburton S, Sawyer AT, Wolfe A, et al. Dysmaturation of the
default mode network in autism. Hum Brain Mapp. 2014; 35: 1284–1296. https://doi.org/10.1002/hbm.
22252 PMID: 23334984
81. Wiggins JL, Peltier SJ, Ashinoff S, Weng SJ, Carrasco M, Welsh RC, et al. Using a self-organizing
map algorithm to detect age-related changes in functional connectivity during rest in autism spectrum
disorders. Brain Res. 2011; 1380: 187–197. https://doi.org/10.1016/j.brainres.2010.10.102 PMID:
21047495
82. Sowell ER, Thompson PM, Holmes CJ, Jernigan TL, Toga AW. In vivo evidence for post-adolescent
brain maturation in frontal and striatal regions. Nat Neurosci. 1999; 2: 859–861. https://doi.org/10.
1038/13154 PMID: 10491602
83. Billeci L, Sicca F, Maharatna K, Apicella F, Narzisi A, Campatelli G, et al. On the application of quanti-
tative EEG for characterizing autistic brain: a systematic review. Front Hum Neurosci. 2013; 7: 442.
https://doi.org/10.3389/fnhum.2013.00442 PMID: 23935579
84. Coben R, Mohammad-Rezazadeh I, Cannon RL. Using quantitative and analytic EEG methods in the
understanding of connectivity in autism spectrum disorders: a theory of mixed over- and under-con-
nectivity. Front Hum Neurosci. 2014; 8: 45. https://doi.org/10.3389/fnhum.2014.00045 PMID:
24616679
85. Sheikhani A, Behnam H, Mohammadi MR, Noroozian M, Mohammadi M. Detection of abnormalities
for diagnosing of children with autism disorders using of quantitative electroencephalography analysis.
J Med Syst. 2012; 36: 957–963. https://doi.org/10.1007/s10916-010-9560-6 PMID: 20711644
86. Rezaie R, Zouridakis G, Choudhri AF, Wheless JW, Vo
¨lgyi E, Van Poppel K, et al. Functional connec-
tivity in tuberous sclerosis complex with autistic spectrum disorder preliminary findings. J Pediatr Neu-
rol. 2013; 11: 79–82.
87. Ghanbari Y, Bloy L, Christopher Edgar J, Blaskey L, Verma R, Roberts TP. Joint analysis of band-spe-
cific functional connectivity and signal complexity in autism. J Autism Dev Disord. 2015; 45: 444–460.
https://doi.org/10.1007/s10803-013-1915-7 PMID: 23963593
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 23 / 28
88. Ahmadlou M, Adeli H, Adeli A. Fuzzy Synchronization Likelihood-wavelet methodology for diagnosis
of autism spectrum disorder. J Neurosci Methods. 2012; 211: 203–209. https://doi.org/10.1016/j.
jneumeth.2012.08.020 PMID: 22968137
89. Barttfeld P, Amoruso L, Ais J, Cukier S, Bavassi L, Tomio A, et al. Organization of brain networks gov-
erned by long-range connections index autistic traits in the general population. J Neurodev Disord.
2013; 5: 16. https://doi.org/10.1186/1866-1955-5-16 PMID: 23806204
90. Barttfeld P, Wicker B, Cukier S, Navarta S, Lew S, Sigman M. A big-world network in ASD: dynamical
connectivity analysis reflects a deficit in long-range connections and an excess of short-range connec-
tions. Neuropsychologia. 2011; 49: 254–263. https://doi.org/10.1016/j.neuropsychologia.2010.11.024
PMID: 21110988
91. Boersma M, Kemner C, de Reus MA, Collin G, Snijders TM, Hofman D, et al. Disrupted functional
brain networks in autistic toddlers. Brain Connect. 2013; 3: 41–49. https://doi.org/10.1089/brain.2012.
0127 PMID: 23259692
92. Buckley AW, Scott R, Tyler A, Mahoney JM, Thurm A, Farmer C, et al. State-Dependent Differences
in Functional Connectivity in Young Children With Autism Spectrum Disorder. EBioMedicine. 2015; 2:
1905–1915. https://doi.org/10.1016/j.ebiom.2015.11.004 PMID: 26844269
93. Carson AM, Salowitz NM, Scheidt RA, Dolan BK, Van Hecke A V. Electroencephalogram coherence
in children with and without autism spectrum disorders: decreased interhemispheric connectivity in
autism. Autism Res. 2014; 7: 334–343. https://doi.org/10.1002/aur.1367 PMID: 24623657
94. Catarino A, Andrade A, Churches O, Wagner AP, Baron-Cohen S, Ring H. Task-related functional
connectivity in autism spectrum conditions: an EEG study using wavelet transform coherence. Mol
Autism. 2013; 4: 1. https://doi.org/10.1186/2040-2392-4-1 PMID: 23311570
95. Chan AS, Han YMY, Sze SL, Lau EM. Neuroenhancement of Memory for Children with Autism by a
Mind-Body Exercise. Front Psychol. 2015; 6: 1893. https://doi.org/10.3389/fpsyg.2015.01893 PMID:
26696946
96. Coben R, Clarke AR, Hudspeth W, Barry RJ. EEG power and coherence in autistic spectrum disorder.
Clin Neurophysiol. 2008; 119: 1002–1009. https://doi.org/10.1016/j.clinph.2008.01.013 PMID:
18331812
97. Duffy FH, Als H. A stable pattern of EEG spectral coherence distinguishes children with autism from
neuro-typical controls—a large case control study. BMC Med. 2012; 10: 64. https://doi.org/10.1186/
1741-7015-10-64 PMID: 22730909
98. Duffy FH, Shankardass A, McAnulty GB, Als H. The relationship of Asperger’s syndrome to autism: a
preliminary EEG coherence study. BMC Med. 2013; 11: 175. https://doi.org/10.1186/1741-7015-11-
175 PMID: 23902729
99. Garcia Dominguez L, Stieben J, Perez Velazquez JL, Shanker S. The imaginary part of coherency in
autism: differences in cortical functional connectivity in preschool children. PLoS One. 2013; 8:
e75941. https://doi.org/10.1371/journal.pone.0075941 PMID: 24098409
100. Isler JR, Martien KM, Grieve PG, Stark RI, Herbert MR. Reduced functional connectivity in visual
evoked potentials in children with autism spectrum disorder. Clin Neurophysiol. 2010; 121: 2035–
2043. https://doi.org/10.1016/j.clinph.2010.05.004 PMID: 20605520
101. Jaime M, McMahon CM, Davidson BC, Newell LC, Mundy PC, Henderson HA. Brief Report: Reduced
Temporal-Central EEG Alpha Coherence During Joint Attention Perception in Adolescents with
Autism Spectrum Disorder. J Autism Dev Disord. 2015;
102. Jamal W, Das S, Oprescu IA, Maharatna K, Apicella F, Sicca F. Classification of autism spectrum dis-
order using supervised learning of brain connectivity measures extracted from synchrostates. J Neural
Eng. 2014; 11: 46019.
103. Keehn B, Vogel-Farley V, Tager-Flusberg H, Nelson CA. Atypical hemispheric specialization for faces
in infants at risk for autism spectrum disorder. Autism Res. 2015; 8: 187–198. https://doi.org/10.1002/
aur.1438 PMID: 25808162
104. Khadem A, Hossein-Zadeh G-A, Khorrami A. Long-Range Reduced Predictive Information Transfers
of Autistic Youths in EEG Sensor-Space During Face Processing. Brain Topogr. 2016; 29: 283–295.
https://doi.org/10.1007/s10548-015-0452-4 PMID: 26433373
105. Kulisek R, Hrncir Z, Hrdlicka M, Faladova L, Sterbova K, Krsek P, et al. Nonlinear analysis of the sleep
EEG in children with pervasive developmental disorder. Neuro Endocrinol Lett. 2008; 29: 512–517.
PMID: 18766147
106. Lazarev V V, Pontes A, Mitrofanov AA, deAzevedo LC. Interhemispheric asymmetry in EEG photic
driving coherence in childhood autism. Clin Neurophysiol. 2010; 121: 145–152. https://doi.org/10.
1016/j.clinph.2009.10.010 PMID: 19951847
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 24 / 28
107. Lazarev V V, Pontes A, Mitrofanov AA, deAzevedo LC. Reduced interhemispheric connectivity in
childhood autism detected by electroencephalographic photic driving coherence. J Autism Dev Disord.
2015; 45: 537–547. https://doi.org/10.1007/s10803-013-1959-8 PMID: 24097142
108. Leveille C, Barbeau EB, Bolduc C, Limoges E, Berthiaume C, Chevrier E, et al. Enhanced connectivity
between visual cortex and other regions of the brain in autism: a REM sleep EEG coherence study.
Autism Res. 2010; 3: 280–285. https://doi.org/10.1002/aur.155 PMID: 20717953
109. Machado C, Estevez M, Rodriguez R, Leisman G, Melillo R, Chinchilla M, et al. Anatomic and Func-
tional Connectivity Relationship in Autistic Children during Three Different Experimental Conditions.
Brain Connect. 2015;
110. Mathewson KJ, Jetha MK, Drmic IE, Bryson SE, Goldberg JO, Schmidt LA. Regional EEG alpha
power, coherence, and behavioral symptomatology in autism spectrum disorder. Clin Neurophysiol.
2012; 123: 1798–1809. https://doi.org/10.1016/j.clinph.2012.02.061 PMID: 22405935
111. Matlis S, Boric K, Chu CJ, Kramer MA. Robust disruptions in electroencephalogram cortical oscilla-
tions and large-scale functional networks in autism. BMC Neurol. 2015; 15: 97. https://doi.org/10.
1186/s12883-015-0355-8 PMID: 26111798
112. Murias M, Webb SJ, Greenson J, Dawson G. Resting state cortical connectivity reflected in EEG
coherence in individuals with autism. Biol Psychiatry. 2007; 62: 270–273. https://doi.org/10.1016/j.
biopsych.2006.11.012 PMID: 17336944
113. Orekhova E V, Elsabbagh M, Jones EJ, Dawson G, Charman T, Johnson MH, et al. EEG hyper-con-
nectivity in high-risk infants is associated with later autism. J Neurodev Disord. 2014; 6: 40. https://doi.
org/10.1186/1866-1955-6-40 PMID: 25400705
114. Peters JM, Taquet M, Vega C, Jeste SS, Fernandez IS, Tan J, et al. Brain functionalnetworks in syn-
dromic and non-syndromic autism: a graph theoretical study of EEG connectivity. BMC Med. 2013;
11: 54. https://doi.org/10.1186/1741-7015-11-54 PMID: 23445896
115. Righi G, Tierney AL, Tager-Flusberg H, Nelson CA. Functional connectivity in the first year of life in
infants at risk for autism spectrum disorder: an EEG study. PLoS One. 2014; 9: e105176. https://doi.
org/10.1371/journal.pone.0105176 PMID: 25140874
116. Chan AS, Han YMY, Sze SL, Cheung M, Leung WW, Chan RCK, et al. Disordered Connectivity Asso-
ciated with Memory Deficits in Children with Autism Spectrum Disorders. Res Autism Spectr Disord.
2011; 5: 237–245.
117. Han Y, Chan A, Sze S, Cheung M, Wong C, Lam J, et al. Altered immune function associated with dis-
ordered neural connectivity and executive dysfunctions: a neurophysiological study on children with
autism spectrum disorders. 2013; 7: 662–674.
118. La
´za
´r AS, La
´za
´r ZI, Bı
´ro
´A, Győri M, Ta
´rnok Z, Prekop C, et al. Reduced fronto-cortical brain connec-
tivity during NREM sleep in Asperger syndrome: An EEG spectral and phase coherence study. Clin
Neurophysiol. 2010; 121: 1844–1854. https://doi.org/10.1016/j.clinph.2010.03.054 PMID: 20434395
119. Bangel KA, Batty M, Ye AX, Meaux E, Taylor MJ, Doesburg SM. Reduced beta band connectivity dur-
ing number estimation in autism. Neuroimage Clin. 2014; 6: 202–213. https://doi.org/10.1016/j.nicl.
2014.08.020 PMID: 25379432
120. Buard I, Rogers SJ, Hepburn S, Kronberg E, Rojas DC. Altered oscillation patterns and connectivity
during picture naming in autism. Front Hum Neurosci. 2013; 7: 742. https://doi.org/10.3389/fnhum.
2013.00742 PMID: 24265611
121. Doesburg SM, Vidal J, Taylor MJ. Reduced Theta Connectivity during Set-Shifting in Children with
Autism. Front Hum Neurosci. 2013; 7: 785. https://doi.org/10.3389/fnhum.2013.00785 PMID:
24294201
122. Garcia Dominguez L, Velazquez JL, Galan RF. A model of functional brain connectivity and back-
ground noise as a biomarker for cognitive phenotypes: application to autism. PLoS One. 2013; 8:
e61493. https://doi.org/10.1371/journal.pone.0061493 PMID: 23613864
123. Perez Velazquez JL, Galan RF. Information gain in the brain’s resting state: A new perspective on
autism. Front Neuroinform. 2013; 7: 37. https://doi.org/10.3389/fninf.2013.00037 PMID: 24399963
124. Kenet T, Orekhova E V, Bharadwaj H, Shetty NR, Israeli E, Lee AK, et al. Disconnectivity of the corti-
cal ocular motor control network in autism spectrum disorders. Neuroimage. 2012; 61: 1226–1234.
https://doi.org/10.1016/j.neuroimage.2012.03.010 PMID: 22433660
125. Khan S, Gramfort A, Shetty NR, Kitzbichler MG, Ganesan S, Moran JM, et al. Local and long-range
functional connectivity is reduced in concert in autism spectrum disorders. Proc Natl Acad Sci U S A.
2013; 110: 3107–3112. https://doi.org/10.1073/pnas.1214533110 PMID: 23319621
126. Khan S, Michmizos K, Tommerdahl M, Ganesan S, Kitzbichler MG, Zetino M, et al. Somatosensory
cortex functional connectivity abnormalities in autism show opposite trends, depending on direction
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 25 / 28
and spatial scale. Brain. 2015; 138: 1394–1409. https://doi.org/10.1093/brain/awv043 PMID:
25765326
127. Kikuchi M, Shitamichi K, Yoshimura Y, Ueno S, Hiraishi H, Hirosawa T, et al. Altered brain connectivity
in 3-to 7-year-old children with autism spectrum disorder. Neuroimage Clin. 2013; 2: 394–401. https://
doi.org/10.1016/j.nicl.2013.03.003 PMID: 24179793
128. Kikuchi M, Yoshimura Y, Shitamichi K, Ueno S, Hirosawa T, Munesue T, et al. A custom magnetoen-
cephalography device reveals brain connectivity and high reading/decoding ability in children with
autism. Sci Rep. 2013; 3: 1139. https://doi.org/10.1038/srep01139 PMID: 23355952
129. Kikuchi M, Yoshimura Y, Hiraishi H, Munesue T, Hashimoto T, Tsubokawa T, et al. Reducedlong-
range functional connectivity in young children with autism spectrum disorder. Soc Cogn Affect Neu-
rosci. 2015; 10: 248–254. https://doi.org/10.1093/scan/nsu049 PMID: 24652855
130. Kitzbichler MG, Khan S, Ganesan S, Vangel MG, Herbert MR, Hamalainen MS, et al. Altered develop-
ment and multifaceted band-specific abnormalities of resting state networks in autism. Biol Psychiatry.
2015; 77: 794–804. https://doi.org/10.1016/j.biopsych.2014.05.012 PMID: 25064418
131. Lajiness-O’Neill R, Richard AE, Moran JE, Olszewski A, Pawluk L, Jacobson D, et al. Neural syn-
chrony examined with magnetoencephalography (MEG) during eye gaze processing in autism spec-
trum disorders: preliminary findings. J Neurodev Disord. 2014; 6: 15. https://doi.org/10.1186/1866-
1955-6-15 PMID: 24976870
132. Lau YC, Hinkley LB, Bukshpun P, Strominger ZA, Wakahiro ML, Baron-Cohen S, et al. Autism traits in
individuals with agenesis of the corpus callosum. J Autism Dev Disord. 2013; 43: 1106–1118. https://
doi.org/10.1007/s10803-012-1653-2 PMID: 23054201
133. Leung RC, Ye AX, Wong SM, Taylor MJ, Doesburg SM. Reduced beta connectivity during emotional
face processing in adolescents with autism. Mol Autism. 2014; 5: 51. https://doi.org/10.1186/2040-
2392-5-51 PMID: 25371811
134. Peiker I, David N, Schneider TR, Nolte G, Schottle D, Engel AK. Perceptual Integration Deficits in
Autism Spectrum Disorders Are Associated with Reduced Interhemispheric Gamma-Band Coher-
ence. J Neurosci. 2015; 35: 16352–16361. https://doi.org/10.1523/JNEUROSCI.1442-15.2015 PMID:
26674862
135. Pollonini L, Patidar U, Situ N, Rezaie R, Papanicolaou AC, Zouridakis G. Functional connectivity net-
works in the autistic and healthy brain assessed using Granger causality. Conf Proc IEEE Eng Med
Biol Soc. 2010; 2010: 1730–1733. https://doi.org/10.1109/IEMBS.2010.5626702 PMID: 21096408
136. Tsiaras V, Simos PG, Rezaie R, Sheth BR, Garyfallidis E, Castillo EM, et al. Extracting biomarkers of
autism from MEG resting-state functional connectivity networks. Comput Biol Med. 2011; 41: 1166–
1177. https://doi.org/10.1016/j.compbiomed.2011.04.004 PMID: 21592470
137. Urbain C, Vogan VM, Ye AX, Pang EW, Doesburg SM, Taylor MJ. Desynchronization of fronto-tempo-
ral networks during working memory processing in autism. Hum Brain Mapp. 2016; 37: 153–164.
https://doi.org/10.1002/hbm.23021 PMID: 26485059
138. Ye AX, Leung RC, Schafer CB, Taylor MJ, Doesburg SM. Atypical resting synchrony in autism spec-
trum disorder. Hum Brain Mapp. 2014; 35: 6049–6066. https://doi.org/10.1002/hbm.22604 PMID:
25116896
139. Ghanbari Y, Bloy L, Shankar V, Edgar JC, Roberts TP, Schultz RT, et al. Functionally driven brain net-
works using multi-layer graph clustering. Med Image Comput Comput Assist Interv. 2014; 17: 113–
120.
140. Coskun MA, Loveland KA, Pearson DA, Papanicolaou AC, Sheth BR. Functional assays of local con-
nectivity in the somatosensory cortex of individuals with autism. Autism Res. 2013; 6: 190–200.
https://doi.org/10.1002/aur.1276 PMID: 23427110
141. Perez Velazquez JL, Barcelo F, Hung Y, Leshchenko Y, Nenadovic V, Belkas J, et al. Decreased
brain coordinated activity in autism spectrum disorders during executive tasks: reduced long-range
synchronization in the fronto-parietal networks. Int J Psychophysiol. 2009; 73: 341–349. https://doi.
org/10.1016/j.ijpsycho.2009.05.009 PMID: 19465065
142. Uhlhaas PJ, Singer W. Abnormal neural oscillations and synchrony in schizophrenia. Nat Rev Neu-
rosci. 2010; 11: 100–113. https://doi.org/10.1038/nrn2774 PMID: 20087360
143. von Stein A, Chiang C, Konig P. Top-down processing mediated by interareal synchronization. Proc
Natl Acad Sci U S A. 2000; 97: 14748–14753. https://doi.org/10.1073/pnas.97.26.14748 PMID:
11121074
144. Buzsa
´ki G, Wang X-J. Mechanisms of Gamma Oscillations. Annu Rev Neurosci. 2012; 35: 203–225.
https://doi.org/10.1146/annurev-neuro-062111-150444 PMID: 22443509
145. Sporns O, Zwi JD. The small world of the cerebral cortex. Neuroinformatics. 2004; 2: 145–162. https://
doi.org/10.1385/NI:2:2:145 PMID: 15319512
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 26 / 28
146. Lindell AK, Hudry K. Atypicalities in cortical structure, handedness, and functional lateralization for lan-
guage in autism spectrum disorders. Neuropsychol Rev. United States; 2013; 23: 257–270. https://
doi.org/10.1007/s11065-013-9234-5 PMID: 23649809
147. Golarai G, Grill-Spector K, Reiss AL. Autism and the development of face processing. Clin Neurosci
Res. 2006; 6: 145–160. https://doi.org/10.1016/j.cnr.2006.08.001 PMID: 18176635
148. Sasson NJ. The development of face processing in autism. J Autism Dev Disord. 2006; 36: 381–394.
https://doi.org/10.1007/s10803-006-0076-3 PMID: 16572261
149. Scott LS. Featural and configural face processing in adults and infants: a behavioral and electrophysi-
ological investigation. Perception. 2006; 35: 1107–1128. https://doi.org/10.1068/p5493 PMID:
17076069
150. Fiebelkorn IC, Foxe JJ, McCourt ME, Dumas KN, Molholm S. Atypical category processing and hemi-
spheric asymmetries in high-functioning children with autism: revealed through high-density EEG
mapping. Cortex. 2013; 49: 1259–1267. https://doi.org/10.1016/j.cortex.2012.04.007 PMID:
22652240
151. Duerden EG, Arsalidou M, Lee M, Taylor MJ. Lateralization of affective processing in the insula. Neu-
roimage. 2013; 78: 159–175. https://doi.org/10.1016/j.neuroimage.2013.04.014 PMID: 23587690
152. Clarke AR, Barry RJ, McCarthy R, Selikowitz M. EEG analysis in Attention-Deficit/Hyperactivity Disor-
der: a comparative study of two subtypes. Psychiatry Res. 1998; 81: 19–29. PMID: 9829647
153. Dykman RA, Holcomb PJ, Oglesby DM, Ackerman PT. Electrocortical frequencies in hyperactive,
learning-disabled, mixed, and normal children. Biol Psychiatry. 1982; 17: 675–685. PMID: 7104419
154. Katada A, Ozaki H, Suzuki H, Suhara K. Developmental characteristics of normal and mentally
retarded children’s EEGs. Electroencephalogr Clin Neurophysiol. 1981; 52: 192–201. PMID: 6167428
155. Stroganova TA, Nygren G, Tsetlin MM, Posikera IN, Gillberg C, Elam M, et al. Abnormal EEG laterali-
zation in boys with autism. Clin Neurophysiol. 2007; 118: 1842–1854. https://doi.org/10.1016/j.clinph.
2007.05.005 PMID: 17581774
156. Prabhakaran V, Smith JA, Desmond JE, Glover GH, Gabrieli JD. Neural substrates of fluid reasoning:
an fMRI study of neocortical activation during performance of the Raven’s Progressive Matrices Test.
Cogn Psychol. 1997; 33: 43–63. https://doi.org/10.1006/cogp.1997.0659 PMID: 9212721
157. Fein G, Raz J, Brown FF, Merrin EL. Common reference coherence data are confounded by power
and phase effects. Electroencephalogr Clin Neurophysiol. 1988; 69: 581–584. Available: http://www.
ncbi.nlm.nih.gov/pubmed/2453336 PMID: 2453336
158. Cantor DS, Thatcher RW, Hrybyk M, Kaye H. Computerized EEG analyses of autistic children. J
Autism Dev Disord. 1986; 16: 169–187. PMID: 3722118
159. Teipel SJ, Pogarell O, Meindl T, Dietrich O, Sydykova D, Hunklinger U, et al. Regional networks under-
lying interhemispheric connectivity: an EEG and DTI study in healthy ageing and amnestic mild cogni-
tive impairment. Hum Brain Mapp. 2009; 30: 2098–2119. https://doi.org/10.1002/hbm.20652 PMID:
18781594
160. Walker L, Gozzi M, Lenroot R, Thurm A, Behseta B, Swedo S, et al. Diffusion tensor imaging in young
children with autism: biological effects and potential confounds. Biol Psychiatry. 2012; 72: 1043–1051.
https://doi.org/10.1016/j.biopsych.2012.08.001 PMID: 22906515
161. Mak-Fan KM, Morris D, Vidal J, Anagnostou E, Roberts W, Taylor MJ. White matter and development
in children with an autism spectrum disorder. Autism. 2013; 17: 541–557. https://doi.org/10.1177/
1362361312442596 PMID: 22700988
162. Jochaut D, Lehongre K, Saitovitch A, Devauchelle AD, Olasagasti I, Chabane N, et al. Atypical coordi-
nation of cortical oscillations in response to speech in autism. Front Hum Neurosci. 2015; 9: 171.
https://doi.org/10.3389/fnhum.2015.00171 PMID: 25870556
163. Brock J. Alternative Bayesian accounts of autistic perception: comment on Pellicano and Burr. Trends
Cogn Sci. 2012; 16: 573–575. https://doi.org/10.1016/j.tics.2012.10.005 PMID: 23123383
164. Lawson RP, Rees G, Friston KJ. An aberrant precision account of autism. Front Hum Neurosci. 2014;
8: 302. https://doi.org/10.3389/fnhum.2014.00302 PMID: 24860482
165. Pellicano E, Burr D. When the world becomes “too real”: a Bayesian explanation of autistic perception.
Trends Cogn Sci. 2012; 16: 504–510. https://doi.org/10.1016/j.tics.2012.08.009 PMID: 22959875
166. Van de Cruys S, Evers K, Van der Hallen R, Van Eylen L, Boets B, de-Wit L, et al. Precise minds in
uncertain worlds: predictive coding in autism. Psychol Rev. 2014; 121: 649–675. https://doi.org/10.
1037/a0037665 PMID: 25347312
167. van Boxtel JJ, Lu H. Impaired global, and compensatory local, biological motion processing in people
with high levels of autistic traits. Front Psychol. 2013; 4: 209. https://doi.org/10.3389/fpsyg.2013.
00209 PMID: 23630514
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 27 / 28
168. van Boxtel JJ, Lu H. A predictive coding perspective on autism spectrum disorders. Front Psychol.
2013; 4: 19. https://doi.org/10.3389/fpsyg.2013.00019 PMID: 23372559
169. Gomez C, Lizier JT, Schaum M, Wollstadt P, Grutzner C, Uhlhaas P, et al. Reduced predictable infor-
mation in brain signals in autism spectrum disorder. Front Neuroinform. 2014; 8: 9. https://doi.org/10.
3389/fninf.2014.00009 PMID: 24592235
170. Palmer CJ, Paton B, Kirkovski M, Enticott PG, Hohwy J. Context sensitivity in action decreases along
the autism spectrum: a predictive processing perspective. Proc Biol Sci. 2015;282.
171. Palmer CJ, Paton B, Hohwy J, Enticott PG. Movement under uncertainty: the effects of the rubber-
hand illusion vary along the nonclinical autism spectrum. Neuropsychologia. 2013; 51: 1942–1951.
https://doi.org/10.1016/j.neuropsychologia.2013.06.020 PMID: 23811293
172. Skewes JC, Jegindo EM, Gebauer L. Perceptual inference and autistic traits. Autism. 2015; 19: 301–
307. https://doi.org/10.1177/1362361313519872 PMID: 24523412
173. Brandeis D, Michel CM, Amzica F. From neuronal activity to scalp potential fields. In: Michel CM, Koe-
nig T, Brandeis D, Gianotti LRR, Wackermann J, editors. Electrical Neuroimaging. 1st ed. Cam-
bridge: Cambridge University Press; 2009. pp. 1–24.
174. David N, Schneider TR, Peiker I, Al-Jawahiri R, Engel AK, Milne E. Variability of cortical oscillation pat-
terns: A possible endophenotype in autism spectrum disorders? Neurosci Biobehav Rev. 2016; 71:
590–600. https://doi.org/10.1016/j.neubiorev.2016.09.031 PMID: 27746319
175. Butler JS, Molholm S, Andrade GN, Foxe JJ. An Examination of the Neural Unreliability Thesis of
Autism. Cereb Cortex. United States; 2016;
176. Srinivasan R, Nunez PL, Silberstein RB. Spatial filtering and neocortical dynamics: estimates of EEG
coherence. IEEE Trans Biomed Eng. 1998; 45: 814–826. https://doi.org/10.1109/10.686789 PMID:
9644890
177. Shen MD, Nordahl CW, Young GS, Wootton-Gorges SL, Lee A, Liston SE, et al. Early brain enlarge-
ment and elevated extra-axial fluid in infants who develop autism spectrum disorder. Brain. 2013; 136:
2825–2835. https://doi.org/10.1093/brain/awt166 PMID: 23838695
178. Vicente R, Gollo LL, Mirasso CR, Fischer I, Pipa G. Dynamical relaying can yield zero time lag neuro-
nal synchrony despite long conduction delays. Proc Natl Acad Sci. 2008; 105: 17157–17162. https://
doi.org/10.1073/pnas.0809353105 PMID: 18957544
179. Wang XJ. Neurophysiological and computational principles of cortical rhythms in cognition. Physiol
Rev. 2010; 90: 1195–1268. https://doi.org/10.1152/physrev.00035.2008 PMID: 20664082
180. Engel AK, Ko¨nig P, Kreiter AK, Singer W. Interhemispheric Synchronization of Oscillatory Neuronal
Responses in Cat Visual Cortex. Science (80-). American Association for the Advancement of Sci-
ence; 1991; 252: 1177–1179.
181. Rebollo B, Gonzalez Andino SL, Grave de Peralta Menendez R, Sanchez-Vives VM. Non-synaptic
wave propagation: modulation of slow oscillations by endogenous electric fields. Program No. 591.08.
2016 Neuroscience Meeting Planner. San Diego, CA: Society for Neuroscience, 2016. Online
182. Schoffelen JM, Gross J. Source connectivity analysis with MEG and EEG. Hum Brain Mapp. 2009; 30:
1857–1865. https://doi.org/10.1002/hbm.20745 PMID: 19235884
183. Vasa RA, Mostofsky SH, Ewen JB. The Disrupted Connectivity Hypothesis of Autism Spectrum Disor-
ders: Time for the Next Phase in Research. Biol Psychiatry Cogn Neurosci Neuroimaging. 2016; 1:
245–252. https://doi.org/10.1016/j.bpsc.2016.02.003 PMID: 28083565
184. Attal Y, Bhattacharjee M, Yelnik J, Cottereau B, Lefèvre J, Okada Y, et al. Modelling and detecting
deep brain activity with MEG and EEG. IRBM. 2009; 30: 133–138.
185. Walsh P, Elsabbagh M, Bolton P, Singh I. In search of biomarkers for autism: scientific, social and ethi-
cal challenges. Nat Rev Neurosci. 2011; 12: 603–612. https://doi.org/10.1038/nrn3113 PMID:
21931335
186. Kopell NJ, Gritton HJ, Whittington MA, Kramer MA. Beyond the connectome: the dynome. Neuron.
United States; 2014; 83: 1319–1328. https://doi.org/10.1016/j.neuron.2014.08.016 PMID: 25233314
Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies
PLOS ONE | https://doi.org/10.1371/journal.pone.0175870 May 3, 2017 28 / 28
... For example, for spectral power in the alpha band, there are various reports of increase [9][10][11], decrease [12][13][14] and no effects [15,16] in ASD compared to NT (neurotypicals). O'Reilly et al. performed a systematic review of the FC ASD literature [17]. Although a meta-analysis could not be performed due to large variability in methodology and samples across the 52 reports included, the authors concluded that there was a trend for a decrease of long-range FC in ASD-even as the frequency bands and brain/scalp regions of these effects were unclear, and despite several studies finding either no difference or even increased FC in ASD [18][19][20]. ...
... Therefore, significant group effects were confirmed by using F-tests on the individual parameters after fitting the models with a restricted maximum likelihood method [39]. The statistical significance of group differences in variance was determined through the log-likelihood ratio between models 3 and 2. The following transformations were applied before LME in order to convert the feature distributions to normality: x → log (x) (absolute power), x → x 4 (alpha reactivity), x → a tanh (x) (orthPowCorr) and x → x 0.11 Table 1 Overview of clinical and demographic characteristics of the participants included in the statistical analyses Child Children (age 6-11), Adol Adolescents (age [12][13][14][15][16][17], and adults are aged 18 years and above, IQ-full scale IQ, M male, F female, ADI social, ADI communication and ADI RRB refer to the Social, Communication and Restricted and Repetitive Behaviours total domain scores of the ADI-R (Autism Diagnostic Interview-Revised). ADOS Social Affect, ADOS RRB and ADOS Total refer to the Social Affect, Restricted and Repetitive Behaviours and Total calibrated severity scores in ADOS-2. ...
... A careful comparison with prior work is nonetheless critical to inform future work with EEG in ASD. Of particular relevance to this effort is O'Reilly et al. 's prior systematic analysis of EEG/MEG FC literature [17]. They evaluated the support for two popular hypotheses in the ASD field-long-range hypoconnectivity and local hyperconnectivity-and concluded that there was relatively strong support for long-range hypoconnectivity in ASD. ...
Article
Full-text available
Background: Understanding the development of the neuronal circuitry underlying autism spectrum disorder (ASD) is critical to shed light into its etiology and for the development of treatment options. Resting state EEG provides a window into spontaneous local and long-range neuronal synchronization and has been investigated in many ASD studies, but results are inconsistent. Unbiased investigation in large and comprehensive samples focusing on replicability is needed. Methods: We quantified resting state EEG alpha peak metrics, power spectrum (PS, 2-32 Hz) and functional connectivity (FC) in 411 children, adolescents and adults (n = 212 ASD, n = 199 neurotypicals [NT], all with IQ > 75). We performed analyses in source-space using individual head models derived from the participants' MRIs. We tested for differences in mean and variance between the ASD and NT groups for both PS and FC using linear mixed effects models accounting for age, sex, IQ and site effects. Then, we used machine learning to assess whether a multivariate combination of EEG features could better separate ASD and NT participants. All analyses were embedded within a train-validation approach (70%-30% split). Results: In the training dataset, we found an interaction between age and group for the reactivity to eye opening (p = .042 uncorrected), and a significant but weak multivariate ASD vs. NT classification performance for PS and FC (sensitivity 0.52-0.62, specificity 0.59-0.73). None of these findings replicated significantly in the validation dataset, although the effect size in the validation dataset overlapped with the prediction interval from the training dataset. Limitations: The statistical power to detect weak effects-of the magnitude of those found in the training dataset-in the validation dataset is small, and we cannot fully conclude on the reproducibility of the training dataset's effects. Conclusions: This suggests that PS and FC values in ASD and NT have a strong overlap, and that differences between both groups (in both mean and variance) have, at best, a small effect size. Larger studies would be needed to investigate and replicate such potential effects.
... 1.3.1 Use of EEG in diagnosing of ASD Diagnosis of ASD using EEG has been attempted from different directions but is not used clinically due to lack of agreement about theory and methods (Gurau, Bosl, & Newton, 2017). For example, some research has looked at epileptogenic comorbidity (Tuchman, Alessandri, & Cuccaro, 2010) and others have focused their attention on ASD abnormal neural connectivity (O'Reilly, Lewis, & Elsabbagh, 2017). In their systematic review of ASD subtypes Gurau, Bosl, and Newton (2017) report ASD differences in 40 studies were seen by applying functional connectivity analysis, spectral power analysis or information dynamics, but they acknowledged there is a lack of generalizability or diagnostic utility in any of these methods. ...
Thesis
Full-text available
Based on the current Diagnostic and Statistical Manual (DSM 5) (American Psychiatric Association, 2013), Autism spectrum disorder (ASD) is a behaviourally categorized syndrome, generally determined early in childhood, consisting of mild to severe social and communication deficits, perseverative and rigid ideation, often accompanied by various stereotypies (Whyatt, 2017). ASD is a life-long condition without any effective intervention to improve the core deficits (Juane Heflin & Simpson, 1998). Diagnosis of ASD is difficult as the diagnosis relies upon behavioural signs and symptoms and, to date, no biological distinctions identifiy ASD (Loth et al., 2016). Despite technological advances, currently approved medical treatment for ASD consists of behavioural training, drug-based symptom management and environmental accommodations (Cardon, 2016). Depending on the severity, people with ASD often find ways to cope and adapt to their deficits. The majority of people who are mildly impaired are able to live relatively normal lives, but the rest do not at great cost to both them and their communities. Over the past few decades, recognition of a need to improve the diagnosis and treatment of ASD has led to novel methods of evaluation, including the use of electroencephalography (EEG) applying various algorithms that employ fractal dimension (FD), entropy and other complex analytic methods. These more recent approaches to EEG analysis are facilitated by innovative mathematics that commenced over five decades ago as well as recent increases in calculation capacity of computers. We use the term ‘complexity analysis’ here to encompass fractal, entropy and other non-linear measures based on the model of the brain as a complex dynamic system (CDS) which depart from commonly used linear analysis. The purpose of this thesis is to employ complexity analysis of EEG signal data to attempt to distinguish higher risk ASD adults from lower risk controls and determine if a distinction can be made that is consistent with behavioural measures, related to the diagnostic criteria. Previous studies have reported that complexity measures can differentiate between resting states, tasks and conditions such as ASD better than linear measures. To avoid confounding our results by preprocessing the EEG data, as is common to remove artefacts and line noise that can interfere with the analysis, we ran both raw and two versions of preprocessed EEG as a control. Our research comprises three studies. Study 1 is a pilot investigation of EEG complexity features with three adult participants to distinguish a hierarchy between three conditions of a resting state (RS): eyes open (EO), eyes closed (EC) and during the transition between (EOEC), to compare six complexity measure algorithms: the correlation dimension (CD) that establishes a peak fractal embedding dimension, calculates the peak dimension and checks it with the false nearest neighbours method (CDPKFNN), the largest Lyapunov exponent (LLE), Higuchi’s fractal dimension (HFD), multi-scale entropy (MSE), multi-fractal detrended fluctuation analysis (MFDFA), and Kolmogorov complexity using Lempel-Ziv methodology (KC). Study 2 is a pilot study expanding on the first investigation to distinguish a hierarchy between three task conditions: a resting state—EO; a cognitive task—the Berg card sorting test (BCST) used for neuropsychological evaluation; and a social skills task—Reading the Mind in the Eyes Test (RMET) used to screen for ASD risk. These three task conditions were compared using the same six complexity measure values. Based on complexity values, Study 3 employs all six measures to distinguish between 39 adult participants who were scored on the autism-spectrum quotient (AQ), to compare a higher or lower risk for ASD. We also investigate some of the possible causes of inconsistent results in our studies and those found in the literature. For Study 1, we tested our hypothesis that the EO state was more complex and values would be higher in all measures by applying the Wilcoxon Signed Rank Test (WSRT) to compare probabilities between median same channel values. Our results for all three pilot participants showed CDPKFNN values were highest in the EO state in raw i.e. un-preprocessed EEG, but for HFD, EO values were lowest in preprocessed EEG. We found that no values for other measures were highest or lowest for any resting state and trends of higher or lower measures in two participants for one or more tests were mixed. This showed that, similar to some mixed results in the literature, EO raw EEG appeared to be a more complex as measured by CDPKFNN, while preprocessed EEG without low-pass band filtering had the lowest value when measured by HFD. In contradiction to the literature we found the other measures had no lowest or highest RS values. We attribute this result to a potential issue with HFD parameters that may not have been optimized for this data or a complexity feature difference between measures. We speculate that other measures failed to show a statistically significant difference between resting states due to a lack of parameter optimization for the data set and the small number of participants. In Study 2 we tested our hypothesis that EEG during the RMET was more complex and would be higher in all measures applying the WSRT to compare probabilities between median same channel values. We found several measures showed task differences with all three participants. CDPKFNN values showed that BCST had the lowest values in raw and preprocessed EEG without low-pass band filtering, with EO trending lower for raw EEG. Interestingly, LLE results for all tasks for all three participants showed a hierarchy of RMET>EO>BCST for preprocessed EEG, with BCST also lowest for all preprocessing setups and no trends. However, as in Study 1, HFD values went in the opposite direction, with RMET having the lowest values in raw EEG. Yet, MSE values, like LLE, set up the same hierarchy of RMET>EO>BCST for preprocessed EEG with trends for lower EO without low-pass band filtering. MFDFA had no highest values for any task but both BCST preprocessed and RMET without low-pass band filtering were lowest, with mixed trends. Finally, KC had RMET lowest for all three participants in the raw EEG setup with BCST trending higher. The overall results were mixed yet appeared to trend in favor of RMET as highest for both LLE and MSE; BCST lowest in CDPKFNN, LLE, and MSE; but with HFD and KC having RMET as lowest. We again speculate differences in complexity features, a lack of parameter optimisations and the small sample size contributed to these mixed results. For Study 3 we tested our hypotheses that overall or relative values would be lower for the higher AQ score group versus lower score group. First, we examined the frequency distributions for our behavioural measures and found some significant correlations between BCST perseverative scores and AQ scores, as well as a small correlation between RMET and AQ scores. For the EEG data, using WSRT to compare probabilities between median same channel values, we found no correlations or trends between AQ score and any of the measures in either of the two groups. Based on inspection of the boxplots and subsequent review of the pvalues, no pattern emerged that indicated a hierarchy of tasks overall based on AQ score. This was an interesting result given some indications from the pilot studies that a hierarchy may exist and the reporting in the literature of higher values for certain cognitive tasks. Due to the small number of participants in the pilot studies, the result is not inexplicable as there was a possibility of apparent significant hierarchy by chance (~ 3%). Despite its common practice in the literature, we did not select channels and frequencies that favored our hypothesis to avoid a type I error. We consider this early research in a relatively new area of EEG analysis that suggests further exploration, as many values were close to the significance threshold and optimization of parameters may improve results. American Psychiatric Association. (2013). Desk reference to the diagnostic criteria from DSM- 5. Washington, DC: American Psychiatric Publishing. Cardon, T. A. (Ed.) (2016). Technology and the treatment of children with autism spectrum disorder. Cham Heidelberg New York Dordrecht London: Springer International. Juane Heflin, L., & Simpson, R. L. (1998). Interventions for children and youth with autism: Prudent choices in a world of exaggerated claims and empty promises. Part I: Intervention and treatment option review. Focus on autism and other developmental disabilities, 13(4), 194-211. Loth, E., Spooren, W., Ham, L. M., Isaac, M. B., Auriche-Benichou, C., Banaschewski, T., . . . Bourgeron, T. (2016). Identification and validation of biomarkers for autism spectrum disorders. Nature Reviews Drug Discovery, 15(1), 70. Whyatt, C. (2017). The Autism Phenotype: Physiology vs. Psychology? In E. B. Torres & C. Whyatt (Eds.), Autism: The Movement Sensing Perspective Boca Raton: CRC Press. (pp. 23-41).
... This activity has been shown to control apoptosis and future connectivity in the developing cortex of mice [43], and in the larval zebrafish optic tectum it is involved in the processing of sensory information [28]. In addition, resting-state magnetic resonance studies have revealed an altered spontaneous functional connectivity in patients diagnosed with ASD or schizophrenia [44][45][46][47][48]. Here, ywhaz −/− larvae present decreased burst activity and synchronisation in the hindbrain, the region where ywhaz showed a higher expression in WT animals. ...
Article
Full-text available
Genetic variants in YWHAZ contribute to psychiatric disorders such as autism spectrum disorder and schizophrenia, and have been related to an impaired neurodevelopment in humans and mice. Here, we have used zebrafish to investigate the mechanisms by which YWHAZ contributes to neurodevelopmental disorders. We observed that ywhaz expression was pan-neuronal during developmental stages and restricted to Purkinje cells in the adult cerebellum, cells that are described to be reduced in number and size in autistic patients. We then performed whole-brain imaging in wild-type and ywhaz CRISPR/Cas9 knockout (KO) larvae and found altered neuronal activity and connectivity in the hindbrain. Adult ywhaz KO fish display decreased levels of monoamines in the hindbrain and freeze when exposed to novel stimuli, a phenotype that can be reversed with drugs that target monoamine neurotransmission. These findings suggest an important role for ywhaz in establishing neuronal connectivity during development and modulating both neurotransmission and behaviour in adults.
... 3 The connectivity in the brain can be anatomical (given by the physical and structural links between distinct areas) and functional (referring to the statistical dependencies among non-physically linked regions). 4 In the second case, the existence of functional networks entails a synchronized behavior between separate brain structures for the development of mental tasks. 3 Therefore, it is crucial to evaluate the regional interactions for a proper and complete description of the information processing dynamics of the brain. ...
Article
The identification of the emotional states corresponding to the four quadrants of the valence/arousal space has been widely analyzed in the scientific literature by means of multiple techniques. Nevertheless, most of these methods were based on the assessment of each brain region separately, without considering the possible interactions among different areas. In order to study these interconnections, this study computes for the first time the functional connectivity metric called cross-sample entropy for the analysis of the brain synchronization in four groups of emotions from electroencephalographic signals. Outcomes reported a strong synchronization in the interconnections among central, parietal and occipital areas, while the interactions between left frontal and temporal structures with the rest of brain regions presented the lowest coordination. These differences were statistically significant for the four groups of emotions. All emotions were simultaneously classified with a 95.43% of accuracy, overcoming the results reported in previous studies. Moreover, the differences between high and low levels of valence and arousal, taking into account the state of the counterpart dimension, also provided notable findings about the degree of synchronization in the brain within different emotional conditions and the possible implications of these outcomes from a psychophysiological point of view.
... fMRI studies also indicate disruption of default mode network (DMN) with reduced connectivity in ASD (Hull et al., 2017). However, several connectivity studies from MEG and fMRI reveal mixed connectivity pattern with global over and under connectivity in different stages of ASD (Hull et al., 2017;O'Reilly et al., 2017). fMRI studies also challenge local overconnectivity (Hull et al., 2017). ...
Article
Full-text available
Over the last two decades, there has been a tremendous increase in research activity on brain connectivity studies and its application in different neurological disorders. Studies have been focused on different connectivity patterns generated and potential biomarkers that could be derived to find the etymology of the disorder. In this review, the focus is on the utilization of wireless electroencephalogram monitoring system for functional connectivity analysis and its capacity for deciphering neurological disorders. The paper reviews different methods adopted to estimate connectivity and the possible convergence of connectivity patterns in four neurological disorders: epilepsy, autism spectrum disorder, Alzheimer and Parkinson's disease. The paper reviews the current status of connectivity research in the aforementioned neurological disorders and its potential in developing a smart e‐health service.
... Indeed, while an early detection of ASD is highly recommended to receive the services and supports the children need to reach their full potential (Dawson et al., 2010;Hyman et al., 2020), a final diagnosis is still obtained approximately during the third year of life (Salomone et al., 2016). However, neuroimaging and electrophysiological evidence consistently suggested that ASD-related atypical brain pattern in cortical regions crucial for socio-communicative skills (such as fronto-temporal areas) could be detected before the fullblown expression of symptoms (Conti et al., 2015;Ha et al., 2015;O'Reilly et al., 2017;Pagnozzi et al., 2018). ...
Article
Full-text available
Accumulating evidence suggests that functional Near-Infrared Spectroscopy (fNIRS) can provide an essential bridge between our current understanding of neural circuit organization and cortical activity in the developing brain. Indeed, fNIRS allows studying brain functions through the measurement of neurovascular coupling that links neural activity to subsequent changes in cerebral blood flow and hemoglobin oxygenation levels. While the literature offers a multitude of fNIRS applications to typical development, only recently this tool has been extended to the study of neurodevelopmental disorders (NDDs). The exponential rise of scientific publications on this topic during the last years reflects the interest to identify a “fNIRS signature” as a biomarker of high translational value to support both early clinical diagnosis and treatment outcome. The purpose of this systematic review is to describe the updating clinical applications of fNIRS in NDDs, with a specific focus on preschool population. Starting from this rationale, a systematic search was conducted for relevant studies in different scientific databases (Pubmed, Scopus, and Web of Science) resulting in 13 published articles. In these studies, fNIRS was applied in individuals with Autism Spectrum Disorder (ASD) or infants at high risk of developing ASD. Both functional connectivity in resting-state conditions and task-evoked brain activation using multiple experimental paradigms were used in the selected investigations, suggesting that fNIRS might be considered a promising method for identifying early quantitative biomarkers in the autism field.
... The above confirms the need to include the characteristics of hyperfocus among the diagnostic criteria for autism. Hyperfocus is a dimension also suggested by MEG studies (32) pointing to long-range underconnectivity in ASD patients. ...
Article
Full-text available
According to the scientific literature, 50 to 70% of individuals with autism spectrum disorder (ASD) also present with comorbid attention deficit hyperactivity disorder (ADHD). From a clinical perspective, this high rate of comorbidity is intriguing. What is the real significance of this dual diagnosis? Is ADHD in fact always present in such cases? Might the attentional impairment reported among our ASD patients actually be a distinct trait of their ASD—namely, impaired joint attention—rather than an ADHD attention deficit? Could their agitation be the consequence of this joint attention impairment or related to a physical restlessness etiologically very different from the agitation typical of ADHD? The neurobiological reality of ASD-ADHD comorbidity is a subject of debate, and amphetamine-based treatment can have paradoxical or undesirable effects in the ASD population. Consequently, does a dual diagnosis, notwithstanding its currency in the literature, prevent us from shedding sufficient light on major physiopathologic questions raised by the clinical picture of ASD?
Preprint
Full-text available
Sleep EEG reflects voltage differences relative to a reference, while its spectrum reflects its composition of various frequencies. In contrast, the envelope of the sleep EEG reflects the instantaneous amplitude of oscillations, while its spectrum reflects the rhythmicity of the occurrence of these oscillations. The sleep EEG spectrum is known to relate to demographic, psychological and clinical characteristics, but the envelope spectrum has been rarely studied. In study 1, we demonstrate in human invasive data from cortex-penetrating microelectrodes and subdural grids that the sleep EEG envelope spectrum reflects neuronal firing. In study 2, we demonstrate that the scalp EEG envelope spectrum is stable within individuals. A multivariate learning algorithm could predict age (r = 0.6) and sex (r = 0.5) from the EEG envelope spectrum. With age, oscillations shifted from a 4–5 second rhythm to faster rhythms. Our results demonstrate that the sleep envelope spectrum is a promising biomarker of demographic and disease-related phenotypes.
Article
Full-text available
Persons with and without autism process sensory information differently. Differences in sensory processing are directly relevant to social functioning and communicative abilities, which are known to be hampered in persons with autism. We collected functional magnetic resonance imaging (fMRI) data from 25 autistic individuals and 25 neurotypical individuals while they performed a silent gesture recognition task. We exploited brain network topology, a holistic quantification of how networks within the brain are organized to provide new insights into how visual communicative signals are processed in autistic and neurotypical individuals. Performing graph theoretical analysis, we calculated two network properties of the action observation network: local efficiency, as a measure of network segregation, and global efficiency, as a measure of network integration. We found that persons with autism and neurotypical persons differ in how the action observation network is organized. Persons with autism utilize a more clustered, local-processing-oriented network configuration (i.e., higher local efficiency), rather than the more integrative network organization seen in neurotypicals (i.e., higher global efficiency). These results shed new light on the complex interplay between social and sensory processing in autism.
Article
Radial progenitor development and function lay the foundation for the construction of the cerebral cortex. Radial glial scaffold, through its functions as a source of neurogenic progenitors and neuronal migration guide, is thought to provide a template for the formation of the cerebral cortex. Emerging evidence is challenging this limited view. Intriguingly, radial glial scaffold may also play a role in axonal growth, guidance, and neuronal connectivity. Radial glial cells not only facilitate the generation, placement, and allocation of neurons in the cortex but also regulate how they wire up. The organization and function of radial glial cells may thus be a unifying feature of the developing cortex that helps to precisely coordinate the right patterns of neurogenesis, neuronal placement, and connectivity necessary for the emergence of a functional cerebral cortex. This perspective critically explores this emerging view and its impact in the context of human brain development and disorders.
Article
Full-text available
Oscillatory neuronal activity may provide a mechanism for dynamic network coordination. Rhythmic neuronal interactions can be quantified using multiple metrics, each with their own advantages and disadvantages. This tutorial will review and summarize current analysis methods used in the field of invasive and non-invasive electrophysiology to study the dynamic connections between neuronal populations. First, we review metrics for functional connectivity, including coherence, phase synchronization, phase-slope index, and Granger causality, with the specific aim to provide an intuition for how these metrics work, as well as their quantitative definition. Next, we highlight a number of interpretational caveats and common pitfalls that can arise when performing functional connectivity analysis, including the common reference problem, the signal to noise ratio problem, the volume conduction problem, the common input problem, and the trial sample size bias problem. These pitfalls will be illustrated by presenting a set of MATLAB-scripts, which can be executed by the reader to simulate each of these potential problems. We discuss of how these issues can be addressed using current methods.
Conference Paper
Full-text available
Autism spectrum disorder (ASD) is a variety of developmental disorders that cause life-long communication and social deficits. However, ASD could only be diagnosed at children as early as 2 years of age, while early signs may emerge within the first year. White matter (WM) connectivity abnormalities have been documented in the first year of lives of ASD subjects. We introduce a novel multi-kernel support vector machine (SVM) framework to identify infants at high-risk for ASD at 6 months old, by utilizing the diffusion parameters derived from a hierarchical set of WM connectomes. Experiments show that the proposed method achieves an accuracy of 76%, in comparison to 70% with the best single connectome. The complementary information extracted from hierarchical networks enhances the classification performance, with the top discriminative connections consistent with other studies. Our framework provides essential imaging connectomic markers and contributes to the evaluation of ASD risks as early as 6 months.
Article
Full-text available
Significance statement: The weak central coherence account proposes a tendency of individuals with autism spectrum disorders (ASDs) to focus on details at the cost of an integrated coherent whole. Here, we provide evidence, at the behavioral and the neural level, that visual integration in object recognition is impaired in ASD, when details had to be integrated across both visual hemifields. We found enhanced interhemispheric gamma-band coherence in typically developed participants when communication between cortical hemispheres was required by the task. Importantly, participants with ASD failed to show this enhanced coherence between bilateral posterior superior temporal sulci. The findings suggest that visual integration is disturbed at the local and global synchronization scale, which might bear implications for object recognition in ASD.
Chapter
There is growing consensus that atypical cognitive and behavioral function in ASD is associated with abnormalities in brain network organization. This chapter reviews the anatomical and functional evidence on connectivity in ASD from diffusion tensor imaging (DTI) and functional connectivity MRI (FCMRI) studies. DTI studies overwhelmingly present evidence of white matter compromise across virtually all major fiber tracts of the brain in children and adults with ASD (although infants show a different pattern of early white matter development). While many FCMRI studies have found underconnectivity in ASD, others have reported diffuse overconnectivity. Both types of findings may reflect a reduced differentiation of networks in ASD. However, the field has been hampered by lack of methodological awareness, which is necessary to avoid misleading generalizations. Functional and anatomical connectivity have often been studied in isolation in the past, but a combination of imaging techniques will be necessary for a more comprehensive understanding of the impact of abnormal connectivity on cognitive processing in ASD.
Article
An emerging neuropathological theory of Autism, referred to here as "the neural unreliability thesis," proposes greater variability in moment-to-moment cortical representation of environmental events, such that the system shows general instability in its impulse response function. Leading evidence for this thesis derives from functional neuroimaging, a methodology ill-suited for detailed assessment of sensory transmission dynamics occurring at the millisecond scale. Electrophysiological assessments of this thesis, however, are sparse and unconvincing. We conducted detailed examination of visual and somatosensory evoked activity using high-density electrical mapping in individuals with autism (N = 20) and precisely matched neurotypical controls (N = 20), recording large numbers of trials that allowed for exhaustive time-frequency analyses at the single-trial level. Measures of intertrial coherence and event-related spectral perturbation revealed no convincing evidence for an unreliability account of sensory responsivity in autism. Indeed, results point to robust, highly reproducible response functions marked for their exceedingly close correspondence to those in neurotypical controls.
Article
In 2004, two papers proposed that pervasive functional under-connectivity (Just et al., 2004) or a trade-off between excessive local connectivity at the cost of distal under-connectivity (Belmonte et al., 2004) characterizes atypical brain organization in autism. Here, we take stock of the most recent and rigorous functional and structural connectivity findings with a careful eye toward evaluating the extent to which they support these original hypotheses. Indeed, the empirical data do not support them. From rsfMRI studies in adolescents and adults, there is an emerging consensus regarding long-range functional connections indicating cortico-cortical under-connectivity, specifically involving the temporal lobes, combined with subcortical-cortical over-connectivity. In contrast, there is little to no consensus regarding local functional connectivity or findings from task-based functional connectivity studies. The structural connectivity data suggest that white matter tracts are pervasively weak, particularly in the temporal lobe. Together, these findings are revealing how deeply complex the story is regarding atypical neural network organization in autism. In other words, distance and strength of connectivity as individual factors or as interacting factors do not consistently explain the patterns of atypical neural connectivity in autism. Therefore, we make several methodological recommendations and highlight developmental considerations that will help researchers in the field cultivate new hypotheses about the nature and mechanisms of potentially aberrant functional and structural connectivity in autism. Research highlights • Just et al. (2004) and Belmonte et al. (2004) proposed specific hypotheses of atypical connectivity patterns in autism. • We review the most recent and rigorous resting-state, task-based, and structural connectivity findings in autism.
Article
During the past decade, the disrupted connectivity theory has generated considerable interest as a pathophysiological model for autism spectrum disorders (ASD). This theory postulates that deficiencies in the way the brain coordinates and synchronizes activity amongst different regions may account for the clinical symptoms of ASD. This review critically examines the current structural and functional connectivity data in ASD and evaluates unresolved assumptions and gaps in knowledge that limit the interpretation of these data. Collectively, studies very often show group alterations in what are thought of as measures of cerebral connectivity, though the patterns of findings vary considerably. We argue that there are three principle needs in this research agenda. First, further basic research is needed to understand the links between measures commonly used (DTI, fMRI, EEG) and other (histological, computational) levels of analysis. Second, speculated causes of inconsistencies in the literature (age, clinical heterogeneity) demand studies that directly evaluate these interpretations. Finally, the field needs well-specified mechanistic models of altered cerebral communication in ASD whose predictions can be tested on multiple levels of analyses.