Review

Getting answers from babies

about autism
Mayada Elsabbagh and Mark H. Johnson
Centre for Brain and Cognitive Development, Birkbeck, University of London, Henry Wellcome Building, London, WC1E 7HX, UK

Because autism is rarely diagnosed before two years of

age, little is known about its early symptoms and causes.
In order to determine the earliest manifestations of the
condition, recent interest has focused on infants at
genetic risk for autism. Current evidence indicates that
overt behavioural symptoms emerge around the end of
the first year. However, studies using laboratory brain
function measures have reported differences in groups
of infants at-risk compared with low-risk controls during
their first year. Some infants displaying such early differences, however, do not subsequently receive a diagnosis. As the search for early markers continues, infants atrisk present a persuasive model for gene by environment
interactions leading to variable developmental pathways.
What can babies tell us about autism?
Autism Spectrum Disorders (ASD) affect 1 in 100 to150
children [1]. The annual societal cost of ASD in the UK
exceeds 27 billion [2]. Our current knowledge of the early
neural, behavioural and cognitive profile is very poor, and
little is known about the underlying causes of ASD or the
process through which symptoms emerge. Because a confirmed diagnosis of ASD can only be made from around two
or three years of age (Box 1) researchers, until recently,
have relied on limited retrospective data on infants
younger than two years of age prior to diagnosis [3,4]. It
is against this background of scientific and clinical challenges that investigators have recently turned to the prospective study of younger siblings of children already
diagnosed with ASD (often referred to as infant siblings).
Although community-based studies are needed to determine recurrence risk in these infants, some studies report
that around 20 per cent of younger siblings go on to
a diagnosis. Studies with this risk group have only
recently begun in earnest, in part due to the large number
of infants that are required to be studied over a period of
several years before revealing the minority who receive a
diagnosis.
Another motivation for research in this area is that
infants at-risk, by virtue of being genetic relatives of
children with autism, might share some characteristics
with affected individuals, even if they do not themselves go
on to receive a diagnosis. In adults the Broader Autism
Phenotype (BAP) refers to behavioural and brain characteristics associated with ASD found not only in affected
individuals, but also in their relatives [57]. As such,
Corresponding author: Elsabbagh, M. (m.elsabbagh@bbk.ac.uk).

several characteristics of ASD might not be atypical, but

their co-occurrence and severity within an individual
determine whether they manifest as normative differences, as opposed to diagnosable symptoms. The BAP
includes overlapping clinical characteristics [7] as well
as differences in face processing [6,8], theory of mind [9],
executive function [10,11] and central coherence [12].
Further interest in studying infant siblings comes from
the field of developmental cognitive neuroscience, where
studying atypical developmental trajectories is thought to
illuminate basic mechanisms that underlie the emergence
of typical social and cognitive skills and their associated
brain functions. This would in turn help address challenging questions regarding the development of the social
brain [13]. Recently, several new methods have been
developed for studying behaviour, cognition and brain
function in very young infants (Figure 1; Box 2). Our goal
in this article is to present an overview of current infant
siblings research within a broader context of developmental cognitive neuroscience, and to explore the potential
implications that studies of infant siblings have on our
basic understanding of gene and environment interactions
in early functional brain development.
Does ASD emerge over time?
Early signs
Despite recent advances in our understanding of the
genetic and neurobiological basis of ASD, the condition
is currently diagnosed on the basis of behavioural characteristics that can take qualitatively different forms in
infancy. Tools currently used for screening and diagnosis,
relating to the three areas affected in ASD (social abilities,
communication and repetitive behaviours), have been validated for children of 18 months and older but not for
younger infants (Box 1). Thus, most approaches to studying infant siblings aim to identify risk markers through
precursors of later developing symptoms. Because ASD is a
complex condition that encompasses characteristics outside the social domain (e.g. motor coordination and visual
attention), some investigators view these as equally
important candidate risk markers, despite the fact that
such characteristics are not specific to, or universal in ASD
[4,14,15].
Emerging findings indicate that during their first year
many infants who go on to a later diagnosis show surprisingly few overt behavioural signs of atypicality. As a group,
these infants might interact well with their caregivers and
show the expected level of social reciprocity for their age
[1517]. However, during their second year they begin to

1364-6613/$ see front matter 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.tics.2009.12.005

81

Review

Trends in Cognitive Sciences Vol.14 No.2

Box 1. Diagnosis and characteristics of ASD in young

children
The process of identifying and diagnosing ASD varies among
clinicians and communities. Three elements are commonly regarded as essential. These are surveillance, screening and diagnosis
[56]. Diagnostic evaluation requires specialist and interdisciplinary
clinical judgement relying on a variety of sources including direct
observation and assessment as well as parent report. Increasingly,
validated instruments including the Autism Diagnostic Observation
Schedule (ADOS) and the Autism Diagnostic Interview (ADI) are
used in the clinical setting. These instruments can lead to
provisional diagnoses from as early as 18 months of age but tend
to be more reliable with increasing age of the child.
Despite the lack of information about ASD in the very early years,
some research has focused on young children diagnosed with ASD
at 2 to 3 years of age (see [4] for a detailed review). The evidence
indicates that ASD in early childhood is associated with impairments in several social-communicative skills including social
orienting and scanning, face and eye contact processing, imitation,
and communication [57,58]. Impairments in joint attention are
specifically characteristic of young children with ASD, relative to
children with developmental delay or other conditions. Moreover,
individual differences in joint attention skills among children with
autism have been associated with difficulties in other social and
communication skills. It has therefore been proposed that early
impairment in joint attention might lead to difficulties in other areas
of development [57,58].
After the onset of these symptoms in the early years, different sets
of abilities show varying trajectories of development. For example,
some abilities, such as face processing, can begin as significantly
impaired but over time compensatory strategies and atypical neural
systems can restore behavioural performance to within the typical
range. By contrast, difficulties in joint attention remain characteristic
of young children with ASD.

show differences across a range of measures, summarised

in Table 1. These early signs of ASD, that can first appear
around 12 months, include several precursors to later
developing symptoms in social and communicative behaviours. Other studies indicate that early signs might include behaviours that fall outside the range used in
diagnosing ASD in children, for example motor and temperamental characteristics. The nature of early behavioural signs seems to vary between infants and can
also change over time. Even by 18 months, diagnosis is
at best tentative and its stability is poor. However, differences in the affected group do become increasingly clear
after this age.
Different pathways towards an ASD diagnosis?
The lack of uniformity in the nature and timing of early
markers among infants, and their probabilistic nature in
terms of predicting diagnosis, has generated interest in the
Table 1. Characteristics of ASD emerging between 12 and 24
months. Early signs are variable and initially have low
predictive value that then increases with age
Deficits and delays in emerging joint attention [17,18]
Decreased response to name [19]
Decreased imitation [15]
Delays in verbal and non-verbal communication [20]
Motor delay [17]
Elevated frequency of repetitive behaviours, e.g. hand waving [21]
Atypical visuo-motor exploration of objects [22]
Extremes of temperament [23]
Decreased flexibility in disengaging visual attention [15]
82

Figure 1. Converging clinical and experimental research methods for studying

autism in infancy. (a) The Autism Observation Schedule for Infants (AOSI) is a
semi-structured tool for the assessment of early behavioural expression of risk for
autism in infants. (b) A baby with her parents having an EEG sensor net fitted in
preparation for a study. (c) Group-level differences between infants at-risk for
autism and a control group in treatment of direct gaze relative to averted gaze
revealed by ERP [32].

notion that there are multiple pathways to ASD. Traditionally, two subgroups have been hypothesised based on their
trajectory of symptoms [24]. The first trajectory characterises infants whose symptoms appear early in development
and become clearer with age. The second trajectory
involves typical initial stages of development followed by
a phase of regression. Prospective studies of infant siblings have lent support to the first of these trajectories in
which early subtle signs first appear around the end of the
first year then develop into clearer symptoms by two years.
Notwithstanding this pattern, the evidence also indicates
significant variation in the rate of change over time among
infant siblings [2527]. In some cases of ASD, delays or
atypicality can begin later and appear more gradually,
leading to a plateau in typical skills. The regression
pathway hypothesised to occur in a minority of children

Review
Box 2. Methods for studying emerging ASD
Several complementary approaches for investigating ASD symptoms or precursors in the early years have been pursued. One of the
first innovative behavioural approaches, the Autism Observation
Scale for Infants (AOSI), utilises a battery of risk markers [14,37,59].
The semi-structured method (Figure 1a) focuses on precursors of
symptoms found in older children, including response to name, eye
contact, social reciprocity and imitation. The battery also includes
items assessing atypical visual and motor skills. Some infants later
diagnosed with ASD begin to be differentiated from the control
group of infants with no family history of autism at around 12
months. At that age, siblings who do not go on to a diagnosis can
also exhibit some markers of risk falling between affected siblings
and control groups.
Clinical research methods have been complemented by experimental methods such as direct measurement of naturally occurring
activity in the brain (EEG; Figure 1b). For example, converging
analytic methods have revealed group-level differences in response
to pictures displaying direct compared to averted gaze within the
first year of life in infant siblings (Figure 1c), as well as baseline
differences in resting EEG response. These findings might reflect
early manifestations of genetic risk for ASD across multiple
developing neural systems, including eye gaze processing, a
developmental precursor to joint attention.
Due to the technical challenges involved in using structural brain
imaging methods with infants, technologies such as Magnetic
Resonance Imaging have only recently started to be used in studies
with infant siblings. One hypothesis currently being tested is
whether an atypically rapid rate of brain growth can be used as an
early marker of ASD. This builds on previous findings of differences
in brain size in young children [60] and underconnectivity in adults
with ASD [61]. The proposal is that variations in head growth
trajectory within the infancy period lead to both functional and
anatomical consequences, which could explain the variability in
trajectories of emerging symptoms [62]. A promising future
development will be the incorporation of multiple behavioural,
brain imaging, and genetic measures within individual infants.

diagnosed with ASD has yet to be established in infant

siblings studies. These hypothetical trajectories (Figure 2)
are based on findings from behavioural studies, and it
remains possible that clearer or more consistent trajectories will become apparent when underlying functional
brain development is understood.

Trends in Cognitive Sciences

Vol.14 No.2

Getting beneath behaviour

Whereas the majority of infant siblings studies to date
have searched for atypicalities in social behaviour, a few
laboratories have recently used methods from developmental cognitive neuroscience with this risk group. The motivation is that more direct measurements of brain function or
cognition might reveal indicators of atypical development
before these become evident in the overt social behaviour of
the infant. Because work with such methods is relatively
new, there is little current data on the predictive value of
the measures in relation to diagnostic outcome at three
years or beyond. However, in contrast to the behavioural
studies described above, such methods have already documented differences during the first year between groups of
infant siblings and low-risk control groups. For example,
measurement of saccadic reaction time in tasks targeting
developing brain attention networks have shown reduced
flexibility in disengaging from a central stimulus to orient
towards a peripheral one [28], a finding parallel to that
observed in older children diagnosed with ASD [29].
Furthermore, the study of contrast thresholds in infants
at-risk highlighted differences in neural systems mediating early visual processing [30]. More direct electrophysiological measurement of brain activity using Event Related
Potentials (ERP) has found early group differences in
response to face stimuli [31] and in sensitivity to the
direction of eye gaze, a developmental precursor to joint
attention [32]. Other emerging findings indicate differences in the risk group in resting state spontaneous
EEG activity at the gamma frequency [32].
There are at least three reasons for differences in the
overall pattern of results from measures of overt social
behaviour and measures of underlying brain functions.
First, as mentioned above, it might be that methods from
developmental cognitive neuroscience are simply more
sensitive at indicating risk in individual infants earlier
in development. Until studies using these methods have
followed up large cohorts of infants at-risk to the point of
diagnosis this issue will remain unresolved. A second

Figure 2. Siblings studies have highlighted variation in the nature and rate of change in behaviour in infants at risk. This has led to proposals of hypothetical variable
trajectories as illustrated. Available findings support variability in onset of clear behavioural symptoms whereas the regression trajectory has not been established in infant
siblings.

83

Review
account builds on the view of some authors that their
findings constitute group-level differences that are not
attributable to the small number of infants going on to
receive an ASD diagnosis (Box 2), and thus that they
represent an early form of the BAP. This account might
help explain some of the current inconsistencies in describing the behavioural phenotype in unaffected siblings
(those who do not go on to receive a later diagnosis). When
considering behaviour, unaffected siblings appear to
muddy the waters in terms of finding early markers for
ASD. Behavioural differences between this group, affected
siblings and low-risk control infants are difficult to detect
at six months of age [16,33,34]. Between 12 and 24 months,
groups inconsistently overlap in some early signs [3537].
At least in some studies, affected and unaffected siblings
continue to be indistinguishable early in this developmental period but diverge in their trajectories over time [37].
Interestingly, by school age, unaffected siblings reach
typical levels of functioning with very few residual delays
or deficits [38]. In summary, according to this view variability in the expression of behavioural risk within the first
years might reflect an infant BAP revealed more clearly
through measurement of underlying brain and cognitive
functions.
A third account of the diverging results between
measures of overt social behaviour and those of underlying
brain and cognitive functions is more speculative in view of
limited data on infant siblings. It is possible that in most
infants, expression of risk is measurable in the form of
subtle aspects of brain and cognitive function. It is only in a
subset of these infants that initially subtle differences
become compounded, leading into a developmental trajectory that results in an ASD outcome. In the majority of
infants, however, well-described processes of brain adaptation and plasticity (discussed next) may restore the
developmental trajectory to a typical course.
Babies at-risk for ASD and geneenvironment
interaction
Researchers in areas of genetics and developmental psychopathology have invoked gene by environment interaction
to help explain why an early genetic or environmental

Trends in Cognitive Sciences Vol.14 No.2

perturbation affects some individuals and not others [39].

In Figure 3 we illustrate some of the underlying assumptions about how genetic or environmental risk factors operating within the infancy period give rise to later behavioural
outcomes. Although our focus here are those factors leading
to ASD, it is likely that similar factors might also contribute
to other behavioural phenotypes such as those observed in
language or attention disorders [28].
Since pioneering studies have estimated heritability for
ASD to exceed 90% [40,41] much research has focused on
genetic risk as contributing to the emergence of the condition. Once viewed as a realistic endeavour, the search for
candidate genes as causal factors has been hampered by
heterogeneity in DNA loci and variable penetrance and
expression. Most geneticists have now abandoned the
notion of susceptibility genes of major effect as accounting
for a large number of ASD cases. Emerging consensus
focuses on genetic heterogeneity, where at least some cases
appear to result from de novo mutations in the form of rare
copy number variants, leading to genomic imbalance and
change in gene expression [42]. In such cases, genetic
mutations might be the overriding risk factor giving rise
to ASD (Figure 3a).
However, it is important to emphasise that such
mutations, along with all of the other genetic factors discovered so far, account for no more than 20% of ASD cases,
with no individual factor explaining more than 1 to 2%.
Therefore, a gene-dosage model is increasingly popular, in
which susceptibility to ASD might be determined by cumulative genetic and non-genetic effects reaching a threshold
[4245]. According to the gene-dosage model, infant siblings outcome is determined by the nature, proportion, or
combination of genetic and non-genetic hits (Figure 3b).
In addition to considering genetic factors, some models
also emphasise the role of the early environment in determining outcomes. For example, in some cases typical allelic
variation determines childrens response to different early
social or physical environments resulting in the emergence
of developmental disorders [39]. In this context, environmental risk is viewed as additive to the gene dose, and
thus influences the infants overall susceptibility to
adverse outcome (Figure 3b). Thus, specific combinations

Figure 3. Different models for how single or multiple genetic and/or environmental risk factors in infancy lead to ASD behavioural outcomes in childhood. In model c, risk
factors predict outcomes through dynamic changes over the course of development. Alternatively, the developmental process might buffer initial vulnerability, canalising
the impact of risk.

84

Review
of genetic and environmental risk factors can determine
the nature of the resulting phenotype, that is ASD or
another developmental condition, as well as variability
in expression ranging from severe diagnosable outcomes
to milder normative characteristics. Consistent with this
model, findings from infant siblings research indicate
variability in the early expression of risk, but much work
will be required in future studies to specify which combinations lead to which outcomes (also Box 3).
Although relatively simple Gene by Environment
models might help explain the onset of some developmental disorders [39], in other cases the effects of their interaction seem to be more complex. For example, evidence
from research on language, and literacy impairment
[46,47], as well as conditions such as Williams Syndrome,
where the genetic underpinnings are clearly understood
[48], supports probabilistic and indirect mapping between
genetic and/or environmental factors and developmental
outcomes [49] (Figure 3c). In these cases the favoured view
is that dynamic gene by environment interactions during
the period of maximal brain plasticity lead to variable
developmental trajectories, not readily predicted by a
simple model of additive risk.
Computational models (neural networks) designed to
illuminate non-linear interactions between genetic and
environmental factors [50] have reaffirmed older ideas in
the literature about the different effects of early perturbations on developmental trajectories [51]. In some cases, an
early perturbation can reset a developmental trajectory
along a different route, as a result of the compounding of
atypical brain computations or behaviour eliciting or
recruiting an atypical environment. In other cases, the
typical developmental trajectory can be resilient in the face
of genetic or early environmental hits through canalisation,
a process through which brain adaptation and plasticity
maintains or restores the typical trajectory. Such trajectory
differences, emerging from the dynamic interaction of
multiple factors over time, are not easily decomposable into
separate genetic and environmental contributions.

Trends in Cognitive Sciences

Vol.14 No.2

Figure 4. Hypothetical trajectories for expression of risk in infant siblings at both

neural and behavioural levels. The ASD trajectory is characterised by a high-dose
of mediating risk factors, the impact of which becomes compounded and amplified
over time. The BAP characterises infants with a low-dose of risk factors leading to
sub-clinical expression of the condition. Canalisation involves similar initial
expression of risk but eventual restoration of the typical developmental trajectory.

Examples of hypothetical developmental trajectories in

infant siblings based on the latter account are illustrated in
Figure 4. In the affected group, early and widespread,
albeit subtle differences, in multiple brain systems might
become compounded over time as a result of atypical
interactions between brain systems and the external
environment. A second group show less initial sub-clinical
characteristics and continue within the BAP profile over
time. A third possibility is that although early manifestations of risk can be observed in infant siblings as a group,
canalisation restores the trajectory of development back to
its typical path. Such dynamic pathways within the early
developmental period might arise as a consequence of
compensatory plasticity in the developing brain. However,
it has been recently proposed that even some older children, who already exhibit a clear profile of autistic symptoms, may reach optimal outcomes [52]. The precise factors
involved in this adaptation in infancy or in childhood have
yet to be verified.

Box 3. Taking individual differences seriously: infant intermediate phenotypes?

There is currently growing interest in using laboratory methods as
biomarkers of risk for a diagnosis of ASD later in life. Biomarkers are
defined as measurable factors specifically associated with a particular
condition, which can be used to ascertain an individuals susceptibility for that condition [63]. Although developing such biomarkers is
a long-term aim of research on infant siblings, findings have already
highlighted that individual differences need to be considered
seriously as these can reflect different developmental pathways to
outcome.
A standard approach in infant siblings research focuses on
identifying group differences in early behaviours as defined by
clinical outcomes, for example ASD vs. unaffected. These category
groups are defined through cut-off scores across a number of
instruments such as the ADOS and the ADI and in combination with
expert clinical judgment. One disadvantage of this approach is that
potentially meaningful variability in the data might be lost. Increasingly, researchers in areas of genetics have advocated the use of
dimensional intermediate or endo-phenotypes, viewed as closer to
the genotype than complex clinical characterisation [64,65]. Specifically, measures of quantitative traits associated with ASD and
overlapping with other disorders are viewed as better candidates
for gene mapping than diagnostic classification [45,46]. The assump-

tion here is that diagnosed forms of ASD, which are themselves

highly variable, are extremes of what is otherwise typical variation.
Even in clinical circles, once viewed as essential for screening
specificity, access to services, treatment design, and reducing
heterogeneity of participants in research, categorical approaches are
gradually being complemented by dimensional ones in ASD [66] and
in other developmental disorders [46,67]. It is probable that the
development of such infant intermediate phenotypes based on
experimental methods will significantly advance research across
these various disciplines. A further implication of recent findings from
infant siblings is that such intermediate phenotypes need not be static
over development. In other words, understanding the impact of
dynamic gene by environment interactions in how these characteristics change over time will be essential.
Infant siblings research has introduced new challenges for infancy
science, which traditionally tests group performance on single
measures. Group studies on infant siblings have demonstrated
success in improving our understanding of the early ASD phenotype
and have allowed comparisons with existing research. However,
future success rests on our ability to complement group data with
validation of dimensional and longitudinal measures for assessing
both risk and outcome.

85

Review
Box 4. Outstanding questions
 Will a combination of risk markers early in infancy be more
effective than individual markers of risk in predicting diagnostic
outcomes for ASD or other overlapping conditions?
 Are brain measures more sensitive than behavioural methods in
identifying and quantifying risk and outcomes?
 Will laboratory methods provide viable future biomarkers of risk?
 Will infant siblings research provide more objective methods for
diagnosing ASD than the current behavioural criteria?
 Can we develop early intervention or prevention techniques that
target those infants who are likely to go on to serious symptoms?
 Can putative canalisation in some infant siblings provide a model
for intervention in those who are subsequently affected?

Different models for gene by environment interactions

provide exciting opportunities for future research on infant
siblings (Box 4). Genotyping of participating families has
already begun in some projects. By contrast, there is
currently less focus on environmental factors. In this
regard, most researchers are rightly cautious in view of
the recent past, where an influential theory attributed the
cause of ASD to parenting style [53]. This idea has been
dismissed by converging evidence including some of the
findings reviewed here. However, models that incorporate
environmental influence raise the possibility that modulating the early social environment might help infants
overcome the adverse impact of genetic vulnerability. It
is against this background that some scientists are beginning to pilot behavioural interventions based on modifying
the early social environment for the prodromal period of
the condition, modelled after those designed for very young
affected children. The success or otherwise of such interventions will help determine the extent to which early
environmental factors can influence outcome in infants
at-risk [54,55].
Concluding remarks
Attempting to understand the ways in which genes, brains
and the environment operate and interact is an enormous
challenge. Owing to the fruitful integration of theories from
different disciplines, innovative methodologies, and a successful alliance of scientists, babies and their families, we
are at the cusp of very exciting discoveries. Studies of infant
siblings of children affected by ASD are beginning to reveal
how genetic risk for developing the condition takes them
through diverging routes in their development. In addition
to advancing basic and clinical science, research with infant
siblings provides promising opportunities for exploring
whether early interventions are successful in reducing
the impact of the adverse symptoms in at least some cases.
Acknowledgements
We wish to sincerely thank all the families taking part in the British
Autism Study of Infant Siblings (BASIS, www.basisnetwork.org). We
wish to specifically thank those families and researchers who have given
consent to the use of their photographs. BASIS is approved by the London
Multicentre Research Ethics Committee, UK. Although the specific views
expressed are our own, we are grateful to our collaborators Patrick
Bolton, Tony Charman, Andrew Pickles, Atsushi Senju, Teodora Gliga,
Janice Fernandes and Kim Davies for helpful discussion. This research is
supported by the UK Medical Research Council (G0701484) and Autism
Speaks, UK. ME is supported by a Leverhulme Trust Early Career
Fellowship.
86

Trends in Cognitive Sciences Vol.14 No.2

References
1 Baird, G. et al. (2006) Prevalence of disorders of the autism spectrum in
a population cohort of children in South Thames: the Special Needs and
autism Project (SNAP). Lancet 15, 210215
2 Knapp, M. et al. (2007) Economic consequences of autism in the UK,
Foundation for People with Learning Disabilities, (UK)
3 Palomo, R. et al. (2006) Autism and family home movies: a
comprehensive review. J. Dev. Behav. Pediatr. 27, S5968
4 Elsabbagh, M. and Johnson, M.H. (2007) Infancy and autism: Progress,
prospects, and challenge. Prog. Brain Res. 164, 355383
5 Bolton, P. et al. (1994) A case-control family history study of autism.
J. Child Psychol. Psychiatry 35, 877900
6 Dawson, G. et al. (2002) Defining the broader phenotype of autism:
genetic, brain, and behavioral perspectives. Dev. Psychopathol. 14,
581611
7 Pickles, A. et al. (2000) Variable expression of the autism broader
phenotype: findings from extended pedigrees. J. Child Psychol.
Psychiatry 41, 491502
8 Dalton, K.M. et al. (2007) Gaze-fixation, brain activation, and amygdala
volume in unaffected siblings of individuals with autism. Biol.
Psychiatry 61, 5121512
9 Baron-Cohen, S. and Hammer, J. (1997) Parents of children with
Asperger Syndrome: what is the cognitive phenotype? J. Cogn.
Neurosci. 9, 548554
10 Belmonte, M. et al. (2009) Visual attention in autism families:
unaffected sibs share atypical frontal activation. J. Child Psychol.
Psychiatry DOI: 10.1111/j.1469-7610.2009.02153.x
11 Hughes, C. et al. (1999) Towards a cognitive phenotype for autism:
increased prevalence of executive dysfunction and superior spatial
span amongst siblings of children with autism. J. Child Psychol.
Psychiatry 40, 705718
12 Happe, F. et al. (2001) Exploring the cognitive phenotype of autism:
weak central coherence in parents and siblings of children with
autism: I Experimental tests. J. Child Psychol. Psychiatry 42, 299307
13 Johnson, M.H. et al. (2005) The emergence of the social brain network:
evidence from typical and atypical development. Dev. Psychopathol. 17,
599619
14 Zwaigenbaum, L. et al. (2005) Behavioral manifestations of autism in
the first year of life. Int. J. Dev. Neurosci. 23, 143152
15 Bryson, S.E. et al. (2007) A prospective case series of high-risk infants
who developed autism. J. Autism Dev. Disord. 37, 1224
16 Landa, R. and Garrett-Mayer, E. (2006) Development in infants with
autism spectrum disorders: a prospective study. J. Child Psychol.
Psychiatry 47, 629638
17 Yoder, P. et al. (2009) Predicting social impairment and ASD diagnosis
in younger siblings of children with autism spectrum disorder. J.
Autism Dev. Disord. DOI: 10.1007/s10803-009-0753-0
18 Sullivan, M. et al. (2007) Response to Joint Attention in Toddlers at
Risk for autism spectrum disorder: A Prospective Study. J. Autism Dev.
Disord. 37, 3748
19 Nadig, A.S. et al. (2007) A prospective study of response to name in
infants at risk for autism. Arch. Pediatr. Adolesc Med. 161, 378383
20 Mitchell, S. et al. (2006) Early language and communication
development of infants later diagnosed with autism spectrum
disorder. J. Dev. Behav. Pediatr. 27, S6978
21 Loh, A. et al. (2007) Stereotyped motor behaviors associated with
autism in high-risk infants: a pilot videotape analysis of a sibling
sample. J. Autism Dev. Disord. 37, 2536
22 Ozonoff, S. et al. (2008) Atypical object exploration at 12 months of age
is associated with autism in a prospective sample. Autism 12, 457
472
23 Garon, N. et al. (2009) Temperament and its relationship to autistic
symptoms in a high-risk infant sib cohort. J. Abnorm. Child Psychol.
37, 5978
24 Lord, C. et al. (2004) Regression and word loss in autistic spectrum
disorders. J. Child Psychol. Psychiatry 45, 936955
25 Landa, R.J. et al. (2007) Social and communication development in
toddlers with early and later diagnosis of autism spectrum disorders.
Arch. Gen. Psychiatry 64, 853864
26 Ozonoff, S. et al. (2008) The onset of autism: Patterns of symptom
emergence in the first years of life. Autism Res. 1, 320328
27 Landa, R.J. (2008) Diagnosis of autism spectrum disorders in the first 3
years of life. Nat. Clin. Prac. Neurol. 4, 138147

Review
28 Elsabbagh, M. et al. (2009) Visual orienting in the early broader autism
phenotype: disengagement and facilitation. J. Child Psychol.
Psychiatry 50, 637642
29 Landry, R. and Bryson, S.E. (2004) Impaired disengagement of
attention in young children with autism. J. Child Psychol.
Psychiatry 45, 11151122
30 McCleery, J.P. et al. (2007) Abnormal magnocellular pathway visual
processing in infants at risk for autism. Biol. Psychiatry 62, 10071014
31 McCleery, J.P. et al. (2009) Atypical faces vs. object processing and
hemispheric asymmetries in 10-month-old infants at risk for autism.
Biol. Psychiatry 66, 950957
32 Elsabbagh, M. et al. (2009) Neural correlates of eye gaze processing in
the infant broader autism phenotype. Biol. Psychiatry 65, 3138
33 Cassel, T.D. et al. (2007) Early Social and emotional communication in
the infant siblings of children with autism spectrum disorders: An
examination of the broad phenotype. J. Autism Dev. Disord. 37, 122
132
34 Yirmiya, N. et al. (2006) The development of siblings of children with
autism at 4 and 14 months: social engagement, communication, and
cognition. J. Child Psychol. Psychiatry 47, 511523
35 Presmanes, A.G. et al. (2007) Effects of different attentional cues on
responding to joint attention in younger siblings of children with
autism spectrum disorders. J. Autism Dev. Disord. 37, 133144
36 Toth, K. et al. (2007) Early social, imitation, play, and language
abilities of young non-autistic siblings of children with autism. J.
Autism Dev. Disord. 37, 145157
37 Bryson, S.E. et al. (2008) The Autism Observation Scale for Infants:
scale development and reliability data. J. Autism Dev. Disord. 38,
731738
38 Gamliel, I. et al. (2007) The development of young siblings of children
with autism from 4 to 54 months. J. Autism Dev. Disord. 37, 171183
39 Stevens, S.E. et al. (2009) Dopamine transporter gene polymorphism
moderates the effects of severe deprivation on ADHD symptoms:
Developmental continuities in gene-environment interplay. Am. J.
Med. Genet. Part B Neuropsychiatr. Genet. 150B, 753761
40 Bailey, A. et al. (1995) Autism as a strongly genetic disorder: evidence
from a British twin study. Psychol. Med. 25, 6377
41 Autism Genome Project (2007) Mapping autism risk loci using genetic
linkage and chromosomal rearrangements. Nat. Genet. 39, 319328
42 Cook, E. and Scherer, S.W. (2008) Copy-number variations associated
with neuropsychiatric conditions. Nature 455, 919923
43 Constantino, J.N. and Todd, R.D. (2005) Intergenerational
transmission of subthreshold autistic traits in the general
population. Biol. Psychiatry 57, 655660
44 Zhao, X. et al. (2007) A unified genetic theory for sporadic and inherited
autism. Proc. Natl. Acad. Sci. U S A 104, 1283112836
45 Abrahams, B.S. and Geschwind, D.H. (2008) Advances in autism
genetics: on the threshold of a new neurobiology. Nat. Rev. Genet. 9,
341355
46 Newbury, D.F. et al. (2005) Genetic influences on language impairment
and phonological short-term memory. Trends Cogn. Sci. 9, 528534
47 Snowling, M.J. et al. (2007) Children at family risk of dyslexia: a followup in early adolescence. J. Child Psychol. Psychiatry 48, 609618

Trends in Cognitive Sciences

Vol.14 No.2

48 Karmiloff-Smith, A. (2009) Nativism versus neuroconstructivism:

rethinking the study of developmental disorders. Dev. Psychol. 45, 5663
49 Karmiloff-Smith, A. (1998) Development itself is the key to
understanding developmental disorders. Trends Cogn. Sci. 2, 389398
50 Elman, J. et al. (1996) Rethinking Innateness: A Connectionist
Perspective on Development, MIT Press
51 Waddington, C.H. (1966) Principles of development and differentiation,
Macmillan Company
52 Helt, M. et al. (2008) Can children with autism recover? If so, how?.
Neuropsychol. Rev. 18, 339366
53 Bettelheim, B. (1967) The Empty Fortress: Infantile Autism and the
Birth of the Self, Free Press
54 Dawson, G. (2008) Early behavioral intervention, brain plasticity, and
the prevention of autism spectrum disorder, Dev. Psychopathol 20,
775803
55 Green, J. and Dunn, G. (2008) Using intervention trials in
developmental psychiatry to illuminate basic science. Br. J.
Psychiatry 192, 323325
56 Zwaigenbaum, L. et al. (2009) Clinical assessment and management of
toddlers with suspected autism spectrum disorder: insights from
studies of high-risk infants. Pediatrics 123, 138313891
57 Charman, T. et al. (1997) Infants with autism: an investigation of
empathy, pretend play, joint attention, and imitation. Dev. Psychol. 33,
781789
58 Swettenham, J. et al. (1998) The frequency and distribution of
spontaneous attention shifts between social and non-social stimuli
in autistic, typically developing and non-autistic developmentally
delayed infants. J. Child Psychol. Psychiatry 39, 747753
59 Brian, J. et al. (2008) Clinical assessment of autism in high-risk 18month-olds. Autism 12, 433456
60 Hazlett, H.C. et al. (2007) Magnetic resonance imaging and head
circumference study of brain size in autism: birth through age 2
years. Arch. Gen. Psychiatry 62, 13661376
61 Murias, M. et al. (2007) Resting state cortical connectivity reflected in
EEG coherence in individuals with autism. Biol. Psychiatry 62, 270
273
62 Lewis, J.D. and Elman, J.L. (2008) Growth-related neural
reorganization and the autism phenotype: a test of the hypothesis
that altered brain growth leads to altered connectivity. Dev. Sci. 11,
135155
63 Lanphear, B. and Bearer, C. (2005) Biomarkers in paediatric research
and practice. Arch. Dis. Child 90, 594600
64 Skuse, D. (2001) Endophenotypes and child psychiatry. Br. J.
Psychiatry 178, 395396
65 Losh, M. and Piven, J. (2007) Social-cognition and the broad autism
phenotype: identifying genetically meaningful phenotypes. Child
Psychol. Psychiatry 48, 105112
66 Gotham, K. et al. (2009) Standardizing ADOS scores for a measure of
severity in autism spectrum disorders. J. Autism Dev. Disord. 39,
693705
67 Bishop, D.V. and Snowling, M.J. (2004) Developmental dyslexia and
specific language impairment: same or different? Psychol. Bull. 130,
858886

87