Complimentary Author PDF: Not for Broad Dissemination

JSLHR

Research Forum

Risk Factors Associated With Language

in Autism Spectrum Disorder:
Clues to Underlying Mechanisms
Helen Tager-Flusberga

Purpose: Identifying risk factors associated with Results: A wide range of risk factors has been found for
neurodevelopmental disorders is an important line of research, ASD, including demographic (e.g., male, family history),
as it will lead to earlier identification of children who could behavioral (e.g., gesture, motor) and neural risk markers (e.g.,
benefit from interventions that support optimal developmental atypical lateralization for speech and reduced functional
outcomes. The primary goal of this review was to summarize connectivity). Environmental factors, such as caregiver
research on risk factors associated with autism spectrum interaction, have not been found to predict language
disorder (ASD). outcomes. Many of the risk markers for ASD are also found
Method: The review focused on studies of infants who in studies of risk for specific language impairment, including
have older siblings with ASD, with particular emphasis on demographic, behavioral, and neural factors.
risk factors associated with language impairment that Conclusions: There are significant gaps in the literature
affects the majority of children with ASD. Findings from this and limitations in the current research that preclude direct
body of work were compared to the literature on specific cross-syndrome comparisons. Future research directions
language impairment. are outlined that could address these limitations.

A
utism spectrum disorder (ASD) is a neurodeve- neural, and environmental factors that have been found
lopmental disorder involving deficits in communi- in studies of infants at risk for ASD, with an emphasis on
cation, social functioning, and behavior that factors associated with language, and then compare these
emerge during the toddler or preschool years. ASD is diag- factors to those found in studies of infants at risk for specific
nosed on the basis of behavioral observations and assess- language impairment (SLI).
ments, but there are significant disparities between the
children who are identified by clinicians in the early years
and those who are not (Daniels & Mandell, 2013). These ASD
disparities have important consequences for access to inter-
Diagnosis
ventions that can be highly effective in reducing the severity
of impairment and especially for promoting language According to the Diagnostic and Statistical Manual
development (Kasari, 2015). To address these issues, one of Mental Disorders, Fifth Edition (DSM–5; American
recent line of research has explored the infancy period using Psychiatric Association, 2013), ASD is diagnosed on the
prospective longitudinal designs in an effort to identify the basis of two symptom clusters: (a) deficits in social commu-
earliest risk factors for ASD and related neurodevelopmental nication and (b) the presence of repetitive behaviors and
disorders, even before the onset of symptoms (Zwaigenbaum restricted interests. For children to be diagnosed with ASD,
et al., 2007). In this Research Forum article I review some of they must meet all three criteria under social communica-
the work in this area focusing on demographic, behavioral, tion: deficits in (a) social-emotional reciprocity; (b) nonverbal
communicative behaviors; and (c) developing, maintaining,
and understanding relationships. This domain includes
several impairments specific to communication: failure in
a
back-and-forth conversation; deficits in nonverbal com-
Boston University, MA
municative behaviors; and difficulties adjusting behavior
Correspondence to Helen Tager-Flusberg: htagerf@bu.edu to suit various social contexts, which includes adjusting
Editor and Associate Editor: Mabel Rice
Received April 16, 2015
Revision received June 4, 2015
Accepted June 9, 2015 Disclosure: The author has declared that no competing interests existed at the time
DOI: 10.1044/2015_JSLHR-L-15-0146 of publication.

Journal of Speech, Language, and Hearing Research • Vol. 59 • 143–154 • February 2016 • Copyright © 2016 American Speech-Language-Hearing Association 143
Research Forum: SLI, ADHD, ASD, CI, Bilingualism, and Bidialectism
 Complimentary Author PDF: Not for Broad Dissemination

language to different listeners and other pragmatic aspects tests; some acquire spoken language after delays in onset
of language. For the second symptom cluster a child must but never reach the normal range, thus having comorbid
have at least two different atypical behavioral patterns, such language impairment; and some never acquire functional
as stereotyped or repetitive motor movements, insistence spoken language even when they have had access to good
on sameness, highly restricted interests, or atypical sensory interventions. These children are referred to as minimally
sensitivities. Diagnosis also requires rating the severity level verbal, but little is known about the source of their pro-
on the basis of the amount of support needed. ASD is found language deficits (Tager-Flusberg & Kasari, 2013). Es-
almost always accompanied by one or more co-occurring timates of the proportions of children within these subgroups
conditions that may develop during different stages of vary depending on ascertainment methods, but the majority
the life span. These conditions include intellectual disability, of affected children acquire spoken language but remain
language disorder, medical conditions (e.g., genetic syndromes, delayed relative to their peers (Kjelgaard & Tager-Flusberg,
epilepsy, sleep or gastrointestinal problems), and other 2001). It has been claimed that this group of children with
psychiatric conditions (e.g., attention-deficit/hyperactivity language impairment and ASD has comorbid SLI, but this
disorder, tic disorders, anxiety, or depression). Recent proposal is still controversial (Norbury, 2013; Williams,
evidence suggests that this set of comorbidities may be Botting, & Boucher, 2008).
useful for defining meaningful subgroups within the ASD The early developmental profiles of language in
population that could serve as a basis for stratifying samples ASD are highly variable. Most children are delayed in
for research on etiology and pathophysiology (Doshi-Velez, standard milestones, especially the onset of words and
Ge, & Kohane, 2014). phrases. On standardized measures, receptive language
ASD has a strong genetic basis. It is highly heritable appears relatively more impaired than expressive language,
(Colvert et al., 2015), and current research has led to the though this may be related more to lack of overall social
rapid ongoing discovery of a relatively large number of de responsiveness than to language processing deficits (Tager-
novo and transmitted rare and common genomic events Flusberg, 2000). After early delays, some children show
that are associated with the diagnosis (Jeste & Geschwind, accelerated language development in the third or fourth
2014). The etiology of ASD is complex, involving non- year, no longer meeting criteria for language impairment
genetic or environmental risk factors, gene–environment (Szatmari et al., 2000). Another group of children shows
interactions, and epigenetic mechanisms, as is true for all a pattern of regression: At 12 to 15 months they begin to
neurodevelopmental disorders (Tordjman et al., 2014). The use words to communicate with others, but then later in
genetic etiology of ASD contributes to alterations in brain the second year they stop speaking (Pickles et al., 2009).
development that may be traced back to the prenatal pe- This loss of language (and social) skills marks the onset
riod (Bae, Jayaraman, & Walsh, 2015). Many studies of of ASD (Ozonoff et al., 2010). As they develop, some of
children and adults with ASD have demonstrated differences these children will regain some language, but others will
in brain structure and function, using a variety of imaging not.
technologies (Ecker & Murphy, 2014; Lainhart, 2015). One important influence on the development of lan-
Atypical patterns in neural connectivity across the brain guage in children with ASD is effective early behavioral
have been highlighted in numerous studies (e.g., Doyle- intervention. Indeed, across several studies and different
Thomas et al., 2015; Lisiecka et al., 2015). In other studies, types of behavioral interventions, the most significant
associations between atypical connectivity and cortical gains observed in children are in receptive and expressive
organization in specific neural systems have been found in language. For example, Dawson et al. (2010) found that
relation to behavioral impairments associated with ASD toddlers receiving a comprehensive behavioral program
(Ameis & Catani, 2015). For example, several studies have (the Early Start Denver Model) for 20 hours a week gained
found relationships among atypical structure, lateraliza- on average almost 20 standard score points in receptive
tion, and functional connectivity in language regions (e.g., language and 12 points in expressive language after 2 years
Knaus et al., 2010; Verly et al., 2014; Williams et al., in the program. Briefer, more targeted interventions also
2013); however, there is still no consistent pattern of find- lead to significant gains in language, as demonstrated, for
ings, largely because of heterogeneity among participants example, in studies by Kasari and colleagues, who found
and differences in methodology across studies. that training joint attention skills in toddlers and preschoolers
with ASD for 30 min a day over a 5- to 6-week period led
to significant gains in expressive language (Kasari, Freeman,
Language in ASD & Paparella, 2006) that were still evident several years after
In the DSM–5, delayed or impaired language is no the intervention was provided (Kasari, Gulsrud, Freeman,
longer included as a core symptom, though clinicians are Paparella, & Hellemann, 2012). Despite the importance
required to note whether a child has a comorbid language and efficacy of early intervention, in every study there are
disorder (American Psychiatric Association, 2013). There children who make little or no progress at all, but little
is enormous variability in the language profiles of children is known about predictors of response to intervention. In
with ASD (Tager-Flusberg, Edelson, & Luyster, 2011). sum, in ASD there is enormous variability in language
Some have intact structural language skills, scoring within reflected in a range of developmental trajectories, response
(or above) the normal range on standardized language to treatment, and longer term outcomes.

144 Journal of Speech, Language, and Hearing Research • Vol. 59 • 143–154 • February 2016
 Complimentary Author PDF: Not for Broad Dissemination

Infants at Risk have ASD may be diagnosed with language delay (Drumm
& Brian, 2013), social-communication delay (Miller et al.,
Studying Infants at Risk for ASD 2015), or developmental delay or intellectual disability
How early can we identify the emergence of atypical (Messinger et al., 2013). This range of outcomes highlights
behavioral or brain patterns associated with symptoms or the complexity and overlap among neurodevelopmental
degree of language impairment in ASD? This question has disorders, reflecting common underlying etiology and neuro-
led to a surge of interest over the past decade in studying pathology (Doherty & Owen, 2014).
infants who are at risk for ASD, beginning in the first few Third, across most studies significant differences are
months of life or even prenatally, long before the onset of found between high- and low-risk infants, even when the
symptoms (Zwaigenbaum et al., 2007). This line of research infants with ASD outcomes are excluded from the analyses
is important for discovering early biomarkers that may be (Tager-Flusberg, 2010). At the behavioral level, studies
used to parse heterogeneity in outcomes and predict response have found that a significant number of high-risk infants
to treatment. In almost all studies that focus on this issue, carry features of the broader autism phenotype: traits
risk is defined as familial risk—the presence of an older associated with ASD that differentiate these infants from
sibling diagnosed with ASD—thus taking advantage of the low-risk infants in the control groups (Messinger et al.,
high heritability of the disorder. The standard research 2013; Ozonoff et al., 2014). Other differences between
design compares these high-risk infant siblings to low-risk high- and low-risk infants are especially striking in studies
members of a control group, usually infants who have an of structural and functional brain development and in ex-
older sibling but no family history of ASD. The infants are perimental eye-tracking studies that can target fundamental
followed prospectively until the time when a diagnosis of cognitive mechanisms (E. J. H. Jones et. al., 2014). These
ASD can be confirmed, at the age of 2 or 3 years. neurocognitive differences are taken as evidence for early
The risk recurrence rate for infant siblings is close emerging endophenotypes associated with ASD (Gottesman
to 1 in 5, on the basis of data accrued through the Baby & Gould, 2003), which are defined as heritable characteris-
Sibling Research Consortium, which brings together re- tics associated with a disorder that are more commonly
searchers from around the world who are conducting infant- found in relatives of individuals who have been diagnosed
sibling studies of ASD (Ozonoff et al., 2011). These rates with the disorder but may be independent of having the dis-
are higher for males than females and for infants who have order. Thus, endophenotypes facilitate the identification of
more than one older sibling with ASD. In addition to these points of neurocognitive vulnerability to the disorder itself.
demographic risk factors (male, family history), epidemio- As a final point, studies of high-risk infants with and
logical studies have found that parental age is another without ASD outcomes have consistently found that they
important risk factor: Older parents, especially fathers, are follow different behavioral and neural developmental
more likely to have a child diagnosed with ASD (Lee & trajectories over the first few years of life (E. J. H. Jones
McGrath, 2015). et al., 2014; W. Jones & Klin, 2013; Luyster, Powell,
Studies of high-risk infants provide insight into several Tager-Flusberg, & Nelson, 2014). These differences in
key issues. First, they have the potential to discover early development suggest that no single time point will be
signs and risk markers for the almost 20% of infants who particularly revealing about the roots of ASD; instead, it
are later diagnosed with ASD. This is one of the primary seems that the hallmark of the emergence of ASD, usually
motivations for conducting these longitudinal studies, in the second or third year of life, is alterations in devel-
in the expectation that identifying these markers will lead opment (Ozonoff et al., 2010). These differences in devel-
eventually to earlier diagnoses. Reviews of the findings opment also extend to some unaffected siblings, suggesting
that have been reported over the past decade conclude that they are part of the endophenotype of the disorder.
that few behavioral patterns specific to ASD appear before
12 months; even then, the patterns that have been found
signal risk rather than individually sensitive and specific
Early Concerns and Risk Signs for
predictors of outcome (Gliga, Jones, Bedford, Charman, & Language in ASD
Johnson, 2014; E. J. H. Jones, Gliga, Bedford, Charman, Several studies have found that at 12 months, high-
& Johnson, 2014). Although signs of social-communicative risk infants—in particular, those who are later diagnosed
impairment have been the predominant emphasis in many with ASD—are delayed in language and gestural commu-
studies, risk markers are evident across multiple develop- nication (e.g., Iverson & Wozniak, 2007; Mitchell et al.,
mental domains, including language (e.g., Paul, Fuerst, 2006), and those delays may be closely tied to delays in
Ramsay, Chawarska, & Klin, 2011), attention (e.g., Bedford motor development (Bhat, Galloway, & Landa, 2012;
et al., 2014), motor skills (Libertus, Sheperd, Ross, & Landa, LeBarton & Iverson, 2013; Nickel et al., 2013). Indeed,
2014; Nickel, Thatcher, Keller, Wozniak, & Iverson, 2013), delays in language-related milestones may be among the
and temperament (e.g., del Rosario, Gillespie-Lynch, Johnson, most reliable early signs of ASD, though this signal clearly
Sigman, & Hutman, 2014). has low specificity because children who are not at risk for
Second, studies of high-risk infants have found that ASD or who have other neurodevelopmental outcomes
clinical outcomes are not limited to ASD. By age 3 years also have delays in language (Luyster, Seery, Talbott, &
or older, a significant proportion of infants who do not Tager-Flusberg, 2011).

Tager-Flusberg: Risk Factors for Language in ASD 145

 Complimentary Author PDF: Not for Broad Dissemination

In our ongoing infant-sibling study, a collaboration affect the process of perceptual narrowing. We investigated
between Boston University and Boston Children’s Hospital, this hypothesis using event-related potentials (ERPs) to
we found that as early as 6 months, about one fifth of capture the development of the neural basis for speech
parents of high-risk infants expressed significant concerns perception in infants at high or low risk for ASD (Seery,
about their child’s language in weekly online diaries that Vogel-Farley, Tager-Flusberg, & Nelson, 2013).
they kept between 6 and 18 months (Talbott, Nelson, & We used the double oddball paradigm developed by
Tager-Flusberg, 2015a). By 12 months, over half the parents Rivera-Gaxiola et al. (2005). A standard stimulus, /da/, was
whose infants were later diagnosed with ASD expressed presented on 80% of trials; a native contrast, /ta/, on 10% of
concerns about their child’s language, far more so than the trials; and a nonnative contrast, /a/, on 10% of the tri-
concerns about core social-communication or repetitive- als. This nonnative contrast is phonemic in languages such
behavior symptoms of ASD. For the infants in our study, as Bengali, but English-speaking adults cannot distinguish
the high-risk infants who were later diagnosed with ASD it from the standard /da/. ERPs were recorded while infants
showed significant delays in communicative gestures at sat on their mother’s lap watching someone blow bubbles
12 months (Talbott, Nelson, & Tager-Flusberg, 2015b). In and the speech stimuli played through speakers. The infants
a similar vein, at this age they also vocalized significantly who were part of our larger longitudinal study were seen
less than the low-risk infants or high-risk unaffected siblings at 6, 9, and 12 months of age, with an average of 30 infants
(Chenausky, Nelson, & Tager-Flusberg, 2015), thus con- in each group providing usable data at each time point.
firming findings from other studies that early delays in com- We expected that at 6 months the amplitude of the
munication and language are hallmark features of ASD. It ERP component elicited about 150–300 ms after the onset
is important to note that this is even true for infants with of the stimulus (the so-called P150) would be maximally
ASD whose later language is well within the normal range. sensitive to both the native and nonnative stop-consonant
contrasts compared to the standard /da/, for both groups
of infants, and indeed this is what we found over the frontal
Brain Mechanisms Underlying Early Language brain areas. We also expected that at 12 months, the P150
amplitude of the nonnative contrast would no longer be
Speech Perception significantly different from the standard, as a result of
A great deal is known about the development of perceptual narrowing and again this is what we found, for
infants’ perception of speech sounds in the first year of life. not only the low-risk but also the high-risk infants, including
At birth infants show a strong preference for listening to those who at age 3 years were diagnosed with ASD. These
speech (Butterfield & Siperstein, 1970; Vouloumanos & results suggest that, contrary to our initial hypothesis, ASD
Werker, 2007) and, like adults, they perceive phonemes may not involve delays in perceptual narrowing, at least
categorically (Eimas, Siqueland, Jusczyk, & Vigorito, 1971). not for the high-risk infants in our sample, all of whom be-
By 4 months, infants can distinguish their own language gan speaking before age 2 years (Seery et al., 2013).
from even closely related other languages to which they have We followed up on these findings in the same infants
not been exposed (Bosch & Sebastián-Gallés, 2001), and by by examining group differences in the amplitude of the
10 months they no longer discriminate consonant contrasts P150 to repeated presentations of the standard stimulus
that are not used in their native language (Kuhl, 2004). This /da/ at 9 months (Seery, Tager-Flusberg, & Nelson, 2014).
perceptual narrowing can be measured using behavioral Our motivation was to explore whether we would find
(Werker & Tees, 1983) or neural imaging methods, including group differences in the degree of habituation to the repeated
electrophysiology (Cheour et al., 1998; Rivera-Gaxiola, speech sounds, as were found in an earlier study of 9-month-
Silva-Pereyra, & Kuhl, 2005). Although there is some old high-risk infants in their responses to repeated tones;
bias toward left-hemisphere processing of speech in young unlike low-risk infants in the control group, these failed to
infants (Dubois et al., 2009), more robust left-lateralized re- exhibit neural habituation (Guiraud et al., 2011). Within the
sponses to language emerge at the end of the first year of life constraints of our study design, which limited our analyses
(Minagawa-Kawai, Mori, Naoi, & Kojima, 2007). of habituation to three repeated stimuli, we did not find that
Kuhl (2004, 2010) has argued that this process of either group showed significant changes in amplitude in re-
perceptual narrowing depends on implicit learning that sponse to successive presentations of the standard. This may
takes place in a social context, specifically through interac- be because infants prefer speech sounds to tones (which
tions with social partners who provide a rich and extensive were used by Guiraud et al., 2011) and continue to attend
exposure to the infant’s native language; these interactions to them without significantly attenuating their attention.
promote changes in the neural organization for language. Although there were no group differences in habitua-
Given the social impairments that define ASD, we investi- tion, high-risk infants had significantly higher P150 am-
gated whether infants at risk for ASD, including those later plitudes across all the standards compared with the low-risk
diagnosed with the disorder, would be delayed in losing the infants in the control group. Moreover, for the high-risk
capacity to discriminate native and nonnative contrasts. group only, the amplitude of the P150 was significantly cor-
We hypothesized that their relative lack of interest in social related with later expressive language ability as assessed on
events compared with nonsocial objects (cf. Tager-Flusberg, the Mullen Scales of Early Learning at 18 months of age.
2010) might limit their tuning into language and hence Thus, for infants at risk for ASD, the atypical larger P150

146 Journal of Speech, Language, and Hearing Research • Vol. 59 • 143–154 • February 2016
 Complimentary Author PDF: Not for Broad Dissemination

amplitudes to repetitions of speech stimuli were associated brain systems (Uddin, Supekar, & Menon, 2013). Although
with better language outcomes. Perhaps these infants were there is considerable controversy in this area because of
more focused on the linguistic stimuli, paying less attention of conflicting findings, studies have found more consistent
than the low-risk infants in the control group to the other patterns of underconnectivity between frontal and posterior
sights and background sounds during the experiment. This brain regions important in higher order cognitive domains
enhanced attention then served them well as a foundation including, for example, language processing (Just, Keller,
for language development. Of course, other explanations Malave, Kana, & Varma, 2012).
are possible, and only further investigation will help tease One important question is how early these patterns
apart the source and impact of this atypical enhanced ampli- of underconnectivity develop. We still know little about
tude to speech in high-risk infants. the developmental trajectory of functional brain develop-
ment, not only in children with ASD but also in typically
developing children, as there have been few longitudinal
Lateralized Response to Speech studies that focus on the development of neural systems, in
The early positive peak of the waveform elicited by particular those underlying language development (Uddin,
the consonant–vowel stimuli used in our study, the P150, 2015). We took advantage of our speech-perception experi-
captures the acoustic changes related to stop consonants. ment to begin exploring these questions in 6- and 12 month-
Later negative segments of the waveform, between 300 and old infants at high and low risk for ASD (Righi, Tierney,
600 ms after the onset of the stimulus, are more sensitive Tager-Flusberg, & Nelson, 2014). Our measure of functional
to subtle hemispheric differences in speech processing. We connectivity was linear coherence: an index of synchrony
analyzed the average amplitude of this late slow wave at in gamma-band activity in the electroencephalogram (EEG)
each of the three age points: 6, 9, and 12 months. signal elicited by the speech stimuli in frontal and posterior
At 6 months there were no hemispheric differences in regions of interest. Linear coherence assesses the correlation
either the high- or low-risk group. By 9 months, and again between the phase and power information of two EEG
at 12 months, there was a significant Group × Hemisphere signals: The higher the correlation, the more synchronized
interaction: This was driven by significant differences in and integrated the signals are, giving us a proxy measure
the responses in the left and right temporal/parietal regions for functional neural connectivity.
at both ages in the low-risk infants in the control group. We took the EEG data elicited by the standard and
In contrast, there were no hemispheric differences at any deviant speech stimuli in the same time window as the
age among the high-risk infants (Seery et al., 2013). We P150 ERP (150–300 ms after stimulus onset). We then ex-
have now completed similar analyses of data drawn from tracted the gamma frequency band (30–50 Hz) and com-
a larger group of fifty-seven 12-month-old high-risk infants, puted linear coherence between a small set of electrodes
for whom 36-month outcome data were available; we were over the frontal and posterior (temporal/parietal) language
thus able to divide them into those with and without ASD areas of both hemispheres for our measure of functional
outcomes. We found that the infants who did not develop connectivity. At 6 months there were no significant differ-
ASD showed no lateralized response to speech. In contrast, ences between the high- and low-risk groups in average
those who developed ASD exhibited a lateralized response, linear coherence (across all stimuli and both hemispheres).
but in the direction opposite to what we found in the low- At 12 months, significant group differences were found:
risk infants in the control group. Thus, the low-risk infants The low-risk infants had significantly higher linear coher-
in the control group showed a significant left-hemisphere ence between frontal and posterior language brain areas in
bias to processing speech, whereas the infants later diagnosed both hemispheres compared with the high-risk group. We
with ASD showed a significant right-hemisphere bias. then separated out the five high-risk infants who were later
Because neither of the high-risk infant groups (with diagnosed with ASD and reran the analyses of the 12-month
and without ASD outcomes) showed a typical left-lateralized data on the three groups (low risk, high risk with ASD,
response to speech in the first year of life, this suggests that high risk without ASD). We again found that the low-risk
we have identified an early endophenotype, but one that infants in the control group had significantly higher linear
does not directly relate to differences in behavioral (linguistic) coherence than the high-risk infants. Those who did not
outcomes. It may be that right-lateralized responses at develop ASD had marginally higher coherence than those
12 months serve as an early predictive biomarker for ASD; who did, indicating that the infants with ASD outcomes
however, these findings are still preliminary, given our rela- had the lowest degree of functional connectivity by 1 year
tively small sample sizes. old. The developmental trajectories of linear coherence
differed across the groups: Whereas the low-risk infants in
the control group showed an expected increase in linear co-
Functional Connectivity for Language herence between 6 and 12 months, the high-risk infants—
There is a general consensus in the literature that ASD both with and without ASD outcomes—showed a decrease.
involves disruptions in cortical connectivity at both the In a second study we explored the development of
structural and functional levels (Geschwind & Levitt, 2007). neural functional connectivity for language in the first year
Neuroimaging studies of functional connectivity patterns of life in high- and low-risk infants using different language
have found both under- and overconnectivity of large-scale stimuli and brain-imaging technology (Keehn, Wagner,

Tager-Flusberg: Risk Factors for Language in ASD 147

 Complimentary Author PDF: Not for Broad Dissemination

Tager-Flusberg, & Nelson, 2013). For this study we used participating in these interactions they in turn influence
functional near-infrared spectroscopy, a noninvasive method their parents’ behavior. It is therefore important to ask
for measuring the concentration of oxy- and deoxyhemo- the question whether some of differences between high-
globin in the cortex using near-infrared light probes placed and low-risk infants in the onset and development of lan-
on the scalp. Changes in these blood oxygenation levels guage could be related to differences in how their mothers
provide an indirect measure of neural activity, similar to (the parent most likely to be the primary caregiver) interact
functional MRI but with much coarser spatial localization with them and, reciprocally, whether high-risk infants in-
(Gervain et al., 2011). We compared the two groups of in- fluence their parents’ social communication and language.
fants at 3, 6, 9, and 12 months as they listened to nonsense
words composed of patterned (ABB—e.g., ba-lo-lo) or
unpatterned (ABC—e.g., ba-lo-ti) trisyllables. As the infants Studies of Maternal Behavior
listened to the language stimuli, they watched a silent video A few studies have investigated the quality of mothers’
of moving shapes or watched someone blowing bubbles interactions with their infants at high risk for ASD. In one
while they sat on their mother’s lap. Clusters of probes early study, Yirmiya et al. (2006) analyzed the interactions
were placed over four regions of interest: frontal and poste- between a small group of mothers of high-risk infants and
rior language regions on the left and right hemispheres. their 4-month-old infants. They found that the mothers
Our measures of functional connectivity were correla- had less emotionally synchronized interactions with their
tions between the average time course of the hemodynamic infants compared with mothers in the control group; how-
response for all possible region-of-interest pairs (within ever, there was no significant impact of these subtle differ-
and across hemispheres). At 3 months, the high-risk infants ences in interaction style on later language development,
showed significantly higher connectivity between frontal assessed when the infants were 14 months old. Wan et al.
and posterior regions in the left hemisphere compared with (2012, 2013) have reported that caregivers of high-risk in-
low-risk infants in the control group. No group differences fants were rated as more directive and less sensitive during
were found at either 6 or 9 months, but by 12 months the play interactions with their 6- and 12-month-old infants
high-risk infants showed significantly lower connectivity compared with caregivers of low-risk infants. In these studies,
compared with the low-risk infants in the control group. the influence of caregiver–infant interactions on child lan-
These 12-month group differences with lower connec- guage were not reported, only the relationship to later di-
tivity in the high-risk infants mirror what we found in agnoses of ASD. At 6 months, none of the rated measures
the EEG connectivity analysis (Righi et al., 2014). The predicted outcomes. At 12 months, the most significant
interesting differences at 3 months reflecting higher functional predictors of ASD were two infant behaviors: (a) attention
connectivity in high-risk infants suggest that these altered to mother and (b) positive affect. Another significant fac-
patterns of neural connectivity are there prenatally or soon tor was dyadic mutuality—a global rating of shared en-
after birth. Moreover, the findings point to strikingly dif- joyment and togetherness—but it is likely that this factor
ferent developmental trajectories in the two groups of infants was heavily influenced by the infants’ behavior (less atten-
that are independent of ASD outcome. Together, our studies tion and less positive affect) during play with their caregiver,
of functional connectivity of high-risk infants in the first particularly as all three measures were highly correlated with
year of life suggest that differences in the development of one another.
neural-system connectivity related to speech and language Leezenbaum, Campbell, Butler, and Iverson (2014)
processing are an early-emerging endophenotype. assessed mothers’ verbal responsiveness to their 13- and
18-month-old infants’ vocal and gestural communication
as they played together in their homes. Mothers of high-
Maternal Contributions to and low-risk infants were equally responsive to their infants’
Early Communication communication at both ages. At 18 months, however, the
high-risk infants produced fewer pointing gestures; this in
Transactional Approaches to Development turn altered their mothers’ linguistic behavior. The vari-
In much of the literature on high-risk infants, it is as- ability in the frequency of infants’ vocal and gestural com-
sumed that differences in behavioral or brain development municative attempts significantly influenced the mothers’
between these infants and low-risk infants in control groups opportunities to respond in ways that are known to support
are related to the genetic risk traveling in families that early word learning. Over time, these reciprocal influences
already have at least one child with ASD. But, taken from could affect language development because opportunities
a broader transactional perspective (Sameroff, 2010), devel- for learning may be limited if the children themselves make
opment is a process that is integrally linked to interactions few attempts to communicate.
between infants and their environment. For language, the As part of our infant-sibling project, we too investi-
critical environmental factor is engagement with caregivers. gated the relationship between maternal and infant gesture
According to this view, parents’ communicative interactions and word usage at 12 months, along with the influences
with infants, which begin long before infants understand on language outcomes at 18 months (Talbott et al., 2015b).
intentional communication, are important for shaping We separated the mothers of high-risk infants later diag-
language development. As infants become more capable of nosed with ASD from those whose infants did not develop

148 Journal of Speech, Language, and Hearing Research • Vol. 59 • 143–154 • February 2016
 Complimentary Author PDF: Not for Broad Dissemination

ASD. At 12 months, all three groups of mothers (high risk several clinical linguistic markers that characterize chil-
with ASD outcome, high risk with no ASD, low risk) were dren with SLI: impairments in nonword repetition, sen-
comparable in the amount of language they spoke to their tence repetition, and, for English speakers, marking
infants. As expected, the mothers also communicated with grammatical tense (Conti-Ramsden, Botting, & Faragher,
their infants using gestures. The mothers of high-risk infants 2001; Tager-Flusberg & Cooper, 1999).
who did not develop ASD gestured significantly more Like ASD, SLI is heterogeneous in both the core
frequently than the other two groups of mothers, who were phenotypic expression (e.g., presence or absence of recep-
comparable to each other in their communications with tive language deficits) and co-occurring conditions, includ-
their infants. We did not find significant correlations between ing, for example, speech sound disorders, attention-deficit/
maternal vocal or gestural communication at 12 months hyperactivity disorder, dyslexia, and social impairment
and later language outcomes for any of the groups. Per- (Leonard, 2014). It is highly heritable, and so far several
haps we lacked statistical power or variability in the socio- risk genes and copy-number variants have been identified
economic backgrounds of our participating families, most (Deriziotis & Fisher, 2013; Simpson et al., 2015). Neuroim-
of whom were well educated. Nevertheless, the important aging studies of individuals with SLI have revealed struc-
finding is that mothers of high-risk infants do talk and tural and functional differences in left frontal and temporal
gesture to their infants, as much as or more than mothers regions associated with language as well as in basal ganglia
of low-risk infants. Indeed, the significantly higher levels structures (Badcock, Bishop, Hardiman, Barry, & Watkins,
of gesturing we found among the mothers of high-risk un- 2012; van der Lely & Pinker, 2014). Despite how common
affected infants suggest that they may be extra vigilant and the disorder is, there has been relatively little systematic
aware of the risks their infants face. research on the neural systems that underlie language im-
In sum, there is no evidence that caregivers of high- pairment in SLI, and most published studies include small
risk infants are providing a suboptimal social-affective sample sizes and different methodologies.
context for language development. When differences have
been found in these mothers’ interactions with their infants,
they seem to be driven by infant behavior, including re- Risk Factors for SLI
duced attention, positive affect, or communicative attempts. For several decades there was considerable interest
In some cases, mothers of high-risk infants provide even in finding risk factors for SLI. This was driven in part
richer language-promoting environments than mothers of because diagnosis before age 4 years is complicated by the
low-risk infants, which may well contribute positively to finding that although many toddlers experience significant
their babies’ developmental trajectories. delays in early language development, in most cases the
delays are resolved during the preschool years with no
long-term enduring language deficits (Paul, 1996; Rescorla,
Risk Factors for SLI and ASD Roberts, & Dahlsgaard, 1997). These so-called late talkers
were studied extensively in an effort to identify which
SLI toddlers would be more likely to go on to receive a diagno-
ASD is not the only complex neurodevelopmental sis of SLI (Moyle, Stokes, & Klee, 2011).
disorder involving language deficits that emerges during The most consistent findings across several case-
the toddler years. To evaluate the specificity of the risk control and epidemiological studies is that family history
factors summarized so far, it is important to compare the and male gender are two important risk factors that raise
findings with ASD to those with other disorders. SLI is the probability that a late-talking toddler will experience
one such candidate, particularly in light of the argument enduring deficits in language (Bishop, Price, Dale, & Plomin,
that children with ASD and language impairment have 2003; Zubrick, Taylor, Rice, & Slegers, 2007) or that school-
comorbid SLI (Tager-Flusberg & Joseph, 2003). age children without knowledge of their language history
SLI is diagnosed on the basis of delays and slowed will meet criteria for SLI (Tomblin, 1989). Socioeconomic
rate of development of language in the absence of hearing factors, including maternal education, may also raise a
impairment, frank neurological damage, intellectual dis- child’s risk for later language impairment, according to
ability, or social deprivation (Leonard, 2014). There is several studies (Christensen, Zubrick, Lawrence, Mitrou,
considerable controversy still over the terminology and def- & Taylor, 2014; Rescorla, 2011).
inition of SLI, which, despite several decades of research, At the behavioral level, several candidate risk markers
has not yet been resolved (Bishop, 2014; Reilly, Bishop, & have been found. Late-talking toddlers who also have poor
Tomblin, 2014). In the DSM–5 the term language disorder receptive language skills or more limited gestural communi-
is used to encompass persistent difficulties in receptive cation are more likely to go on to receive a diagnosis of SLI
and/or expressive language. There is general agreement (Ellis & Thal, 2008). Another important predictor is motor
that SLI involves core deficits in grammar, verbal memory, development: Toddlers with poorer motor skills are also
and vocabulary and that it can be diagnosed on the basis at greater risk for significant delays in language development
of standardized language tests, though there is no agree- that could later meet criteria for SLI (Zubrick et al., 2007).
ment on what the cutoff scores should be (Reilly et al., Two groups of researchers have systematically investi-
2014). Nevertheless, there is some consensus that there are gated infants at risk for SLI, defined on the basis of family

Tager-Flusberg: Risk Factors for Language in ASD 149

 Complimentary Author PDF: Not for Broad Dissemination

history, with an emphasis on neural mechanisms underlying There are, however, significant limitations in compar-
auditory and speech processing. Friederici and colleagues ing risk factors for ASD and SLI, which makes it difficult
focused on very young infants with a family history of SLI. to discern whether there are any clear risk factors that dis-
Using ERPs, they found that at 2 months, a group of tinguish these disorders. One major problem is that studies
14 high-risk infants showed a delayed mismatch response on risk factors for these disorders have relied on somewhat
to changes in syllable length (Friedrich, Weber, & Friederici, different primary research designs and methods. Most of the
2004). At 4 to 5 months, nine infants at risk for SLI showed work on SLI grew out of small-scale studies of toddlers with
a delayed mismatch response for discriminating the stress language delays or infants at familial risk, or population-
pattern of two-syllable words (Weber, Hahne, Friedrich, based epidemiological studies of poor language outcomes
& Friederici, 2005), suggesting that deficits in processing in children. In contrast, research on risk factors associated
speech duration very early in life may be a marker for later with ASD has been driven primarily by detailed, relatively
language impairment. Benasich and colleagues studied well-powered longitudinal studies of infants at familial risk.
slightly older infants at familial risk for SLI. In one behav- There is a significant imbalance in the number of publica-
ioral study, they found that eleven 7-month-old high-risk in- tions that address risk factors associated with these disorders.
fants were less able to discriminate tones presented in rapid Most of the research on SLI is older, and there is now only
succession compared with 16 low-risk controls (Benasich a trickle of studies coming out in the literature on SLI.
& Tallal, 2002; Choudhury, Leppanen, Leevers, & Benasich, In contrast, since 2010 about 100 papers a year have been
2007). Their ERP data showed that the amplitude of the published on ASD; most of these are based on infants at
mismatch ERP response was smaller and delayed in onset risk, a design that offers the greatest promise to expand our
when listening to tones with brief interstimulus intervals knowledge of how a neurodevelopmental disorder unfolds
(Benasich et al., 2006). They also found that between 6 and over time. One of the main drivers of this significant imbal-
12 months, high-risk infants have atypical lateralization ance in the studies of infants at risk for ASD and SLI is
of response to tone pairs compared with low-risk controls. in the funding available for different neurodevelopmental
These ERP measures predicted language outcomes, but disorders. Bishop (2010) demonstrated that in the first
they were assessed only when the children were 2 years old, decade of the current century, funding from the National
long before a clinical diagnosis of SLI can be made. Institutes of Health increased 65 times more for ASD re-
Taken together, these studies suggest that neuro- search compared with SLI research. Indeed, ASD was the
cognitive differences, affecting auditory as well as speech fastest growing neurodevelopmental disorder in terms of
perception, are present very early in life for infants at risk research funding, which goes a long way toward explaining
for SLI. One significant limitation of these studies is the why high-powered studies of infants at risk for ASD, many
small number of infants who were included in the high-risk of which were begun during this period, were possible.
group: Groups ranged from nine to 14 infants. A second Even when researchers have taken similar approaches
limitation is that the infants who participated in these stud- to the study of early risk factors, differences in methodology
ies were not followed through the preschool years, when and choice of paradigms limit the ability to make cross-
a diagnosis of SLI could be made. Thus, these early differ- syndrome comparisons. One clear example is studies of
ences in neural responses to tones and speech may be part neural processing of sounds and speech in the first year of
of the endophenotype for SLI; their potential as significant life. Research on ASD has focused on habituation to tones
predictors of risk beyond family history is not known. or speech, atypical lateralization for speech on the basis
of amplitude measures, and reduced functional connectivity
between major cortical language regions (Guiraud et al.,
2011; Keehn et al., 2013; Seery et al., 2013, 2014; Righi
Comparisons of Research on Risk
et al., 2014). Research on SLI has investigated delays in the
Factors for SLI and ASD timing of ERP responses to speech or differences in process-
There are obvious parallels in several of the risk factors ing tones (Benasich et al., 2006; Friedrich et al., 2004;
that have been found for ASD and SLI. At the demographic Weber et al., 2005). We do not know, therefore, whether
level, the most significant predictors for both disorders are the atypical patterns found in infants at risk for ASD are
family history and male gender. Behavioral characteristics, the same as or different from those found in infants at risk
including paucity of gestural communication, poor recep- for SLI.
tive language, and motor delays, are associated with both Some of the most exciting findings from behavioral
ASD and SLI. Also, infants at familial risk for either ASD and brain studies of infants at risk for ASD are the differ-
or SLI exhibit atypical neural responses to auditory or ences seen in the developmental trajectories over the first
speech stimuli, as well as atypical lateralization. To some year of life that have been analyzed for high-risk infants
extent these parallels reflect commonalities in risk profiles with and without a clinical outcome at age 3 years. The
for a broad range of neurodevelopmental disorders; to longitudinal nature of the research designs opens up the
some extent they reflect overlap in etiology, for example, opportunity to carry out hierarchical modeling of develop-
shared risk genes for ASD and SLI (Eicher & Gruen, 2015); ment during a time of greatest change. This has led to
and to some extent they reflect core foundational precur- a deeper understanding of how ASD emerges primarily
sors for language (Iverson, 2010). from differences in behavioral or brain trajectories of change

150 Journal of Speech, Language, and Hearing Research • Vol. 59 • 143–154 • February 2016
 Complimentary Author PDF: Not for Broad Dissemination

rather than in differences that can be measured at any sin- less attention has been paid to SLI. This is ironic, given
gle point in time. There are fewer comparable developmen- that SLI affects far more children than does ASD.
tal studies of SLI and none that include infants before One of the main conclusions we can draw from the
the age of 18 months. We therefore cannot compare the research conducted so far is that many risk factors are
emergence of these two disorders to investigate whether, shared across these two disorders, and these factors perhaps
even though there are many shared risk factors, the key extend to other neurodevelopmental disorders of known or
differences between them might be found in their develop- unknown etiology. Another important conclusion is that
ment over the first 3 years of life. risk factors for poor language outcomes can be found at
Another important difference between studies of risk the level of genes, brains, and behavior, and in some cases
factors for ASD and SLI is at the level of clinical outcomes. even the environment, broadly construed. No one factor
Most research on infants at high risk for ASD focuses on can be singled out; instead, a complex, cumulative model
the same gold-standard objective assessments and clinical of risk is the most likely direction to take in developing a
best-estimate measures of diagnosis at age 2 or 3 years. comprehensive understanding of how the full range of
The literature on SLI is less consistent. None of the studies potential risk factors interacts over the first few years of
on infants at familial risk for SLI followed their participants life to shape language outcomes for all children. Future
longitudinally to evaluate whether their findings were research should build on the accomplishments that have
endophenotypes or more specific risk markers for SLI. Also, already been made. Cross-syndrome comparisons will be
some of the epidemiological studies relied on test score out- important for highlighting shared and distinct risk factors.
comes rather than clinical evaluations. One key difference Our ultimate goal is to develop preventive interventions
between ASD and SLI is that proportionately far fewer that may be individually designed around every infant’s
preschool-age children with SLI are identified in clinical unique cluster of risk markers in order to offer them all the
caseloads compared with toddlers or preschoolers with best opportunity to reach their full linguistic potential.
ASD, a disorder for which there is now mandated pediatric
screening at both 18 and 24 months of age. This difference
makes it far harder to find young children at risk for SLI at Acknowledgment
early ages before a confirmed SLI diagnosis can be made. Funding for this Research Forum article was provided by
Although a number of studies have followed late the National Institute on Deafness and Other Communication
talkers over several years to evaluate their outcomes, the Disorders Grant RO1 DC 10290, awarded to Helen Tager-Flusberg
results are surprising in that so few late talkers in these and Charles A. Nelson.
studies ended up with SLI—significantly fewer than would
be predicted simply on the basis of the prevalence of the
disorder in the general population (Rescorla, 2011). Several References
researchers have discussed this paradox (Rescorla & Dale, Ameis, S. H., & Catani, M. (2015). Altered white matter connec-
2013). Leonard (2013) argued that a number of factors tivity as a neural substrate for social impairment in autism
may have led to the potential exclusion of children with spectrum disorder. Cortex, 62, 158–181.
American Psychiatric Association. (2013). Diagnostic and statistical
early language delays from the older studies on late talkers,
manual of mental disorders (5th ed.). Washington, DC: Author.
including the use of only vocabulary size as a definition of Badcock, N. A., Bishop, D. V. M., Hardiman, M. J., Barry, J. G.,
delayed language, thus leading to underestimates of children & Watkins, K. E. (2012). Co-localisation of abnormal brain
with SLI at later ages. Indeed, Rice, Taylor, and Zubrick structure and function in specific language impairment. Brain
(2008) recently found that when expanded definitions of and Language, 120, 310–320.
late talkers are used at 24 months in a large epidemiological Bae, B.-I., Jayaraman, D., & Walsh, C. A. (2015). Genetic changes
sample—specifically, definitions that include use of word shaping the human brain. Developmental Cell, 32, 423–434.
combinations—late language emergence does predict later Bedford, R., Pickles, A., Gliga, T., Elsabbagh, M., Charman, T.,
diagnoses of SLI. Johnson, M. H., & The BASIS Team. (2014). Additive effects
of social and non-social attention during infancy relate to later
autism spectrum disorder. Developmental Science, 17, 612–620.
Conclusions Benasich, A. A., Choudhury, N., Friedman, J. T., Realpe-Bonilla, T.,
Chojnowska, C., & Gou, Z. (2006). The infant as a prelinguistic
Language skill is the single most important predictor model for language learning impairments: Predicting from
of long-term educational, social, and vocational achieve- event-related potentials to behavior. Neuropsychologia, 44,
ment for all children, including those with ASD (Tager- 396–411.
Flusberg et al., 2011). Understanding the full range of risk Benasich, A. A., & Tallal, P. (2002). Infant discrimination of rapid
factors that predict language outcomes is therefore of great auditory cues predicts later language impairment. Behavioural
importance for clinical practice because it will allow clini- Brain Research, 136, 31–49.
Bhat, A. N., Galloway, J. C., & Landa, R. J. (2012). Relation
cians to identify children in need of targeted language inter- between early motor delay and later communication delay in
ventions at a much younger age than would be possible if infants at risk for autism. Infant Behavior & Development, 35,
we waited until a full-blown language disorder could be 838–846.
diagnosed. Although significant progress has been made in Bishop, D. V. M. (2010). Which neurodevelopmental disorders get
identifying risk factors for language in ASD, comparatively researched and why? PLoS One, 5, e15112.

Tager-Flusberg: Risk Factors for Language in ASD 151

 Complimentary Author PDF: Not for Broad Dissemination

Bishop, D. V. M. (2014). Ten questions about terminology for Drumm, E., & Brian, J. (2013). The developing language abilities
children with unexplained language problems. International and increased risks of “unaffected” siblings of children with
Journal of Language & Communication Disorders, 49, 381–415. autism spectrum disorder. Neuropsychiatry, 3, 513–524.
Bishop, D. V. M., Price, T. S., Dale, P. S., & Plomin, R. (2003). Dubois, J., Hertz-Pannier, L., Cachia, A., Mangin, J. F., Le Bihan,
Outcomes of early language delay: II. Etiology of transient D., & Dehaene-Lambertz, G. (2009). Structural asymmetries
and persistent language difficulties. Journal of Speech, Language, in the infant language and sensori-motor networks. Cerebral
and Hearing Research, 46, 561–575. Cortex, 19, 414–423.
Bosch, L., & Sebastián-Gallés, N. (2001). Evidence of early lan- Ecker, C., & Murphy, D. (2014). Neuroimaging in autism—From
guage discrimination abilities in infants from bilingual envi- basic science to translational research. Nature Reviews Neurology,
ronments. Infancy, 2, 29–49. 10, 82–91.
Butterfield, E. C., & Siperstein, G. N. (1970). Influence of contin- Eicher, J. D., & Gruen, J. R. (2015). Language impairment and
gent auditory stimulation upon non-nutritional suckle. In J. F. dyslexia genes influence language skills in children with autism
Bosma (Ed.), Third symposium on oral sensation and percep- spectrum disorders. Autism Research, 8, 229–234. doi:10.1002/
tion: The mouth of the infant (pp. 313–334). Springfield, IL: aur.1436
Charles C Thomas. Eimas, P. D., Siqueland, E. R., Jusczyk, P., & Vigorito, J. (1971,
Chenausky, K., Nelson, C. A., & Tager-Flusberg, H. (2015). Early January 22). Speech perception in infants. Science, 171,
vocalization types in infants at high and low risk for autism. 303–306.
Unpublished manuscript. Ellis, E. M., & Thal, D. J. (2008). Early language delay and risk
Cheour, M., Ceponiene, R., Lehtokoski, A., Luuk, A., Allik, J., for language impairment. Perspectives on Language Learning
Alho, K., & Näätänen, R. (1998). Development of language- and Education, 15, 93–100.
specific phoneme representations in the infant brain. Nature Friedrich, M., Weber, C., & Friederici, A. D. (2004). Electro-
Neuroscience, 1, 351–353. physiological evidence for delayed mismatch response in
Choudhury, N., Leppanen, P. H. T., Leevers, H. J., & Benasich, infants at-risk for specific language impairment. Psychophys-
A. A. (2007). Infant information processing and family history iology, 41, 772–782.
of specific language impairment: Converging evidence for Gervain, J., Mehler, J., Werker, J. F., Nelson, C. A., Csibra, G.,
RAP deficits from two paradigms. Developmental Science, 10, Lloyd-Fox, S., . . . Aslin, R. N. (2011). Near-infrared spectros-
213–236. copy: A report from the McDonnell Infant Methodology
Christensen, D., Zubrick, S. R., Lawrence, D., Mitrou, F., & Taylor, Consortium. Developmental Cognitive Neuroscience, 1, 22–46.
C. L. (2014). Risk factors for low receptive vocabulary abilities Geschwind, D. H., & Levitt, P. (2007). Autism spectrum disorders:
in the preschool and early school years in the longitudinal Developmental disconnection syndromes. Current Opinion in
study of Australian children. PLoS One, 9, e101476. Neurobiology, 17, 103–111.
Colvert, E., Tick, B., McEwen, F., Stewart, C., Curran, S. R., Gliga, T., Jones, E. J. H., Bedford, R., Charman, T., & Johnson,
Woodhouse, E., . . . Bolton, P. (2015). Heritability of autism M. H. (2014). From early markers to neuro-developmental
spectrum disorder in a UK population-based twin sample. mechanisms of autism. Developmental Review, 34, 189–207.
JAMA Psychiatry, 72, 415–423. doi:10.10001/jamapsychiatry. Gottesman, I. I., & Gould, T. D. (2003). The endophenotype con-
2014.3028 cept in psychiatry: Etymology and strategic intentions. The
Conti-Ramsden, G., Botting, N., & Faragher, B. (2001). Psycholin- American Journal of Psychiatry, 160, 636–645.
guistic markers for specific language impairment (SLI). The Guiraud, J. A., Kushnerenko, E., Tomalski, P., Davies, K., Ribeiro,
Journal of Child Psychology and Psychiatry, 42, 741–748. H., Johnson, M. H., & The BASIS Team. (2011). Differential
Daniels, A. M., & Mandell, D. S. (2013). Explaining differences in habituation to repeated sounds in infants at high risk for au-
age at autism spectrum disorder diagnosis: A critical review. tism. NeuroReport, 22, 845–849.
Autism, 18, 583–597. Iverson, J. M. (2010). Developing language in a developing body:
Dawson, G., Rogers, S., Munson, J., Smith, M., Winter, J., The relationship between motor development and language
Greenson, J., . . . Varley, J. (2010). Randomized, controlled development. Journal of Child Language, 37, 229–261.
trial of an intervention for toddlers with autism: The Early Iverson, J. M., & Wozniak, R. H. (2007). Variation in vocal-
Start Denver Model. Pediatrics, 125, e17–e23. motor development in infant siblings of children with autism.
del Rosario, M., Gillespie-Lynch, K., Johnson, S., Sigman, M., & Journal of Autism and Developmental Disorders, 37, 158–170.
Hutman, T. (2014). Parent reported temperament trajectories Jeste, S. S., & Geschwind, D. H. (2014). Disentangling the hetero-
among infant siblings of children with autism. Journal of geneity of autism spectrum disorder through genetic findings.
Autism and Developmental Disorders, 44, 381–393. Nature Reviews Neurology, 10, 74–81.
Deriziotis, P., & Fisher, S. E. (2013). Neurogenomics of speech and Jones, E. J. H., Gliga, T., Bedford, R., Charman, T., & Johnson,
language disorders: The road ahead. Genome Biology, 14, 204. M. H. (2014). Developmental pathways to autism: A review of
Doherty, J. L., & Owen, M. J. (2014). Genomic insights into the prospective studies of infants at risk. Neuroscience & Biobehav-
overlap between psychiatric disorders: Implications for research ioral Reviews, 39, 1–33.
and clinical practice. Genome Medicine, 6, 29. doi:10.1186/ Jones, W., & Klin, A. (2013, December 19). Attention to eyes is
gm546 present but in decline in 2–6-month-old infants later diagnosed
Doshi-Velez, F., Ge, Y., & Kohane, I. (2014). Comorbidity clusters with autism. Nature, 504, 427–431.
in autism spectrum disorders: An electronic health record Just, M. A., Keller, T. A., Malave, V. L., Kana, R. K., & Varma,
time-series analysis. Pediatrics, 133, e54–e63. S. (2012). Autism as a neural systems disorder: A theory of
Doyle-Thomas, K. A. R., Lee, W., Foster, N. E. V., Tryfon, A., frontal-posterior underconnectivity. Neuroscience & Biobehav-
Ouimet, T., Hyde, K. L. . . . The NeuroDevNet ASD Imaging ioral Reviews, 36, 1292–1313.
Group. (2015). Atypical functional brain connectivity during Kasari, C. (2015). Update on behavioral interventions for autism
rest in autism spectrum disorders. Annals of Neurology, 77, and developmental disabilities. Current Opinion in Neurology,
866–876. doi:10.1002/ana.24391 28, 124–129.

152 Journal of Speech, Language, and Hearing Research • Vol. 59 • 143–154 • February 2016
 Complimentary Author PDF: Not for Broad Dissemination

Kasari, C., Freeman, S., & Paparella, T. (2006). Joint attention at three years of age. Journal of the American Academy of
and symbolic play in young children with autism: A randomized Child & Adolescent Psychiatry, 52, 300–308.e1.
controlled intervention study. The Journal of Child Psychology Miller, M., Young, G. S., Hutman, T., Johnson, S., Schwichtenberg,
and Psychiatry, 47, 611–620. A. J., & Ozonoff, S. (2015). Early pragmatic language difficul-
Kasari, C., Gulsrud, A., Freeman, S., Paparella, T., & Hellemann, ties in siblings of children with autism: Implications for DSM–5
G. (2012). Longitudinal follow-up of children with autism social communication disorder? The Journal of Child Psychology
receiving targeted interventions on joint attention and play. and Psychiatry, 56, 774–781.
Journal of the American Academy of Child & Adolescent Psy- Minagawa-Kawai, Y., Mori, K., Naoi, N., & Kojima, S. (2007).
chiatry, 51, 487–495. Neural attunement processes in infants during the acquisition
Keehn, B., Wagner, J. B., Tager-Flusberg, H., & Nelson, C. A. of a language-specific phonemic contrast. The Journal of Neu-
(2013). Functional connectivity in the first year of life in infants roscience, 27, 315–321.
at-risk for autism: A preliminary near-infrared spectroscopy Mitchell, S., Brian, J., Zwaigenbaum, L., Roberts, W., Szatmari, P.,
study. Frontiers in Human Neuroscience, 7, 444. doi:10.3389/ Smith, I., & Bryson, S. (2006). Early language and communi-
fnhum.2013.00444 cation development of infants later diagnosed with autism
Kjelgaard, M. M., & Tager-Flusberg, H. (2001). An investigation spectrum disorder. Journal of Developmental & Behavioral
of language impairment in autism: Implications for genetic Pediatrics, 27, S69–S78.
subgroups. Language and Cognitive Processes, 16, 287–308. Moyle, J., Stokes, S. F., & Klee, T. (2011). Early language delay
Knaus, T. A., Silver, A. M., Kennedy, M., Lindgren, K. A., and specific language impairment. Developmental Disabilities
Dominick, K. C., Siegel, J., & Tager-Flusberg, H. (2010). Research Reviews, 17, 160–169.
Language laterality in autism spectrum disorder and typical Nickel, L. R., Thatcher, A. R., Keller, F., Wozniak, R. H., &
controls: A functional, volumetric, and diffusion tensor MRI Iverson, J. M. (2013). Posture development in infants at height-
study. Brain and Language, 112, 113–120. ened versus low risk for autism spectrum disorders. Infancy,
Kuhl, P. K. (2004). Early language acquisition: Cracking the 18, 639–661.
speech code. Nature Reviews Neuroscience, 5, 831–843. Norbury, C. F. (2013). Sources of variation in developmental lan-
Kuhl, P. K. (2010). Brain mechanisms in early language acquisition. guage disorders: Evidence from eye-tracking studies of sentence
Neuron, 67, 713–727, production. Philosophical Transactions of the Royal Society B,
Lainhart, J. E. (2015). Brain imaging research in autism spectrum 369. Retrieved from http://rstb.royalsocietypublishing.org/content/
disorders: In search of neuropathology and health across the 369/1634/20120393
lifespan. Current Opinion in Psychiatry, 28, 76–82. Ozonoff, S., Iosif, A.-M., Baguio, F., Cook, I. C., Hill, M. M.,
LeBarton, E. S., & Iverson, J. M. (2013). Fine motor skill predicts Hutman, T., . . . Young, G. S. (2010). A prospective study of
expressive language in infant siblings of children with autism. the emergence of early behavioral signs of autism. Journal
Developmental Science, 16, 815–827. of the American Academy of Child & Adolescent Psychiatry,
Lee, B. K., & McGrath, J. J. (2015). Advancing parental age and 49, 256–266.
autism: Multifactorial pathways. Trends in Molecular Medicine, Ozonoff, S., Young, G. S., Belding, A., Hill, M., Hill, A., Hutman,
21, 118–125. T., . . . Iosif, A.-M. (2014). The broader autism phenotype in
Leezenbaum, N. B., Campbell, S. B., Butler, D., & Iverson, J. M. infancy: When does it emerge? Journal of the American Academy
(2014). Maternal verbal responses to communication of infants of Child & Adolescent Psychiatry, 53, 398–407.
at low and heightened risk of autism. Autism, 18, 694–703. Ozonoff, S., Young, G. S., Carter, A., Messinger, D., Yirmiya, N.,
Leonard, L. B. (2013). Alternate routes to language impairment. Zwaigenbaum, L., . . . Stone, W. L. (2011). Recurrence risk for
In L. A. Rescorla & P. S. Dale (Eds.), Late talkers: Language autism spectrum disorders: A Baby Siblings Research Consor-
development, interventions, and outcomes (pp. 363–376). tium study. Pediatrics, 128, e488–e495.
Baltimore, MD: Brookes. Paul, R. (1996). Clinical implications of the natural history of
Leonard, L. B. (2014). Children with specific language impairment slow expressive language development. American Journal of
(2nd ed.). Cambridge, MA: MIT Press. Speech-Language Pathology, 5, 5–21.
Libertus, K., Sheperd, K. A., Ross, S. W., & Landa, R. J. (2014). Paul, R., Fuerst, Y., Ramsay, G., Chawarska, K., & Klin, A. (2011).
Limited fine motor and grasping skills in 6-month-old infants Out of the mouths of babes: Vocal production in infant sib-
at high risk for autism. Child Development, 85, 2218–2231. lings of children with ASD. The Journal of Child Psychology
Lisiecka, D. M., Holt, R., Tait, R., Ford, M., Lai, M.-C., Chura, and Psychiatry, 52, 588–598.
L. R., . . . Suckling, J. (2015). Developmental white matter mi- Pickles, A., Simonoff, E., Conti-Ramsden, G., Falcaro, M., Simkin,
crostructure in autism phenotype and corresponding endophe- Z., Charman, T., . . . Baird, G. (2009). Loss of language in
notype during adolescence. Translational Psychiatry, 5, e529. early development of autism and specific language impairment.
doi:10.1038/tp.2015.23 The Journal of Child Psychology and Psychiatry, 50, 843–852.
Luyster, R. J., Powell, C., Tager-Flusberg, H., & Nelson, C. A. Reilly, S., Bishop, D. V. M., & Tomblin, B. (2014). Terminological
(2014). Neural measures of social attention across the first years debate over language impairment in children: Forward move-
of life: Characterizing typical development and markers of au- ment and sticking points. International Journal of Language &
tism risk. Developmental Cognitive Neuroscience, 8, 131–143. Communication Disorders, 49, 452–462.
Luyster, R. J., Seery, A. M., Talbott, M. R., & Tager-Flusberg, H. Rescorla, L. (2011). Late talkers: Do good predictors of outcome
(2011). Identifying early-risk markers and developmental exist? Developmental Disabilities Research Reviews, 17, 141–150.
trajectories for language impairment in neurodevelopmental Rescorla, L. A., & Dale, P. S. (Eds.). (2013). Late talkers: Lan-
disorders. Developmental Disabilities Research Reviews, 17, guage development, interventions, and outcomes. Baltimore,
151–159. MD: Brookes.
Messinger, D., Young, G. S., Ozonoff, S., Dobkins, K., Carter, A., Rescorla, L., Roberts, J., & Dahlsgaard, K. (1997). Late talkers at
Zwaigenbaum, L., . . . Sigman, M. (2013). Beyond autism: A 2: Outcome at age 3. Journal of Speech, Language, and Hearing
Baby Siblings Research Consortium study of high-risk children Research, 40, 556–566.

Tager-Flusberg: Risk Factors for Language in ASD 153

 Complimentary Author PDF: Not for Broad Dissemination

Rice, M. L., Taylor, C. L., & Zubrick, S. R. (2008). Language Tomblin, J. B. (1989). Familial concentration of developmental
outcomes of 7-year-old children with or without a history of language impairment. Journal of Speech and Hearing Disorders,
late language emergence at 24 months. Journal of Speech, 54, 287–295.
Language, and Hearing Research, 51, 394–407. Tordjman, S., Somogyi, E., Coulon, N., Kermarrec, S., Cohen, D.,
Righi, G., Tierney, A. L., Tager-Flusberg, H., & Nelson, C. A. Bronsard, G., . . . Xavier, J. (2014). Gene × Environment in-
(2014). Functional connectivity in the first year of life in infants teractions in autism spectrum disorders: Role of epigenetic
at risk for autism spectrum disorder: An EEG study. PLoS mechanisms. Frontiers in Psychiatry, 5, 53. doi:10.3389/fpsyt.
One, 9, e105176. 2014.00053
Rivera-Gaxiola, M., Silva-Pereyra, J., & Kuhl, P. K. (2005). Brain Uddin, L. Q. (2015). Idiosyncratic connectivity in autism: Devel-
potentials to native and non-native speech contrasts in 7- and opmental and anatomical considerations. Trends in Neurosci-
11-month-old American infants. Developmental Science, ence, 38, 261–263.
8, 162–172. Uddin, L. Q., Supekar, K., & Menon, V. (2013). Reconceptualiz-
Sameroff, A. (2010). A unified theory of development: A dialectic ing functional brain connectivity in autism from a develop-
integration of nature and nurture. Child Development, 81, mental perspective. Frontiers in Human Neuroscience, 7, 458.
6–22. doi:10.3389/fnhum.2013.00458
Seery, A., Tager-Flusberg, H., & Nelson, C. A. (2014). Event- van der Lely, H. K. J., & Pinker, S. (2014). The biological basis
related potentials to repeated speech in 9-month-old infants of language: Insight from developmental grammatical impair-
at risk for autism spectrum disorder. Journal of Neurodevelop- ments. Trends in Cognitive Sciences, 18, 586–595.
mental Disorders, 6, 43. Verly, M., Verhoeven, J., Zink, I., Mantini, D., Peeters, R., Deprez,
Seery, A. M., Vogel-Farley, V., Tager-Flusberg, H., & Nelson, C. A. S., . . . Sunaert, S. (2014). Altered functional connectivity of
(2013). Atypical lateralization of ERP response to native and the language network in ASD: Role of classical language areas
non-native speech in infants at risk for autism spectrum disor- and cerebellum. NeuroImage: Clinical, 4, 374–382.
der. Developmental Cognitive Neuroscience, 5, 10–24. Vouloumanos, A., & Werker, J. F. (2007). Listening to language
Simpson, N. H., Ceroni, F., Reader, R. H., Covill, L. E., Knight, at birth: Evidence for a bias for speech in neonates. Develop-
J. C., The SLI Consortium, . . . Newbury, D. F. (2015). Genome- mental Science, 10, 159–164.
wide analysis identifies a role for common copy number variants Wan, M. W., Green, J., Elsabbagh, M., Johnson, M., Charman, T.,
in specific language impairment. European Journal of Human Plummer, F., & The BASIS Team. (2012). Parent–infant inter-
Genetics, 23, 1370–1377. doi:10.1038/ejhg.2014.296 action in infant siblings at risk of autism. Research in Develop-
Szatmari, P., Bryson, S. E., Streiner, D. L., Wilson, F., Archer, L., mental Disabilities, 33, 924–932.
& Ryerse, C. (2000). Two-year outcome of preschool children Wan, M. W., Green, J., Elsabbagh, M., Johnson, M., Charman, T.,
with autism or Asperger’s syndrome. The American Journal Plummer, F., & The BASIS Team. (2013). Quality of interac-
of Psychiatry, 157, 1980–1987. tion between at-risk infants and caregiver at 12–15 months is
Tager-Flusberg, H. (2000). The challenge of studying language de- associated with 3-year autism outcome. The Journal of Child
velopment in autism. In L. Menn & N. B. Ratner (Eds.), Methods Psychology and Psychiatry, 54, 763–771.
for studying language production (pp. 311–330). Mahwah, NJ: Weber, C., Hahne, A., Friedrich, M., & Friederici, A. D. (2005).
Erlbaum. Reduced stress pattern discrimination in 5-month-olds as a
Tager-Flusberg, H. (2010). The origins of social impairments in marker of risk for later language impairment: Neurophysio-
autism spectrum disorder: Studies of infants at risk. Neural logical evidence. Cognitive Brain Research, 25, 180–187.
Networks, 23, 1072–1076. Werker, J. F., & Tees, R. C. (1983). Developmental changes
Tager-Flusberg, H., & Cooper, J. (1999). Present and future possi- across childhood in the perception of non-native speech sounds.
bilities for defining a phenotype for specific language impair- Canadian Journal of Psychology, 37, 278–286.
ment. Journal of Speech, Language, and Hearing Research, Williams, D., Botting, N., & Boucher, J. (2008). Language in
42, 1275–1278. autism and specific language impairment: Where are the links?
Tager-Flusberg, H., Edelson, L., & Luyster, R. (2011). Language and Psychological Bulletin, 134, 944–963.
communication in autism spectrum disorders. In D. Amaral, Williams, D. L., Cherkassky, V. L., Mason, R. A., Keller, T. A.,
G. Dawson, & D. Geschwind (Eds.), Autism spectrum disorders Minshew, N. J., & Just, M. A. (2013). Brain function differ-
(pp. 172–185). Oxford, United Kingdom: Oxford University ences in language processing in children and adults with autism.
Press. Autism Research, 6, 288–302.
Tager-Flusberg, H., & Joseph, R. M. (2003). Identifying neurocog- Yirmiya, N., Gamliel, I., Pilowsky, T., Feldman, R., Baron-Cohen,
nitive phenotypes in autism. Philosophical Transactions of the S., & Sigman, M. (2006). The development of siblings of chil-
Royal Society B, 358, 303–314. dren with autism at 4 and 14 months: Social engagement, com-
Tager-Flusberg, H., & Kasari, C. (2013). Minimally verbal school- munication, and cognition. The Journal of Child Psychology
aged children with autism spectrum disorder: The neglected and Psychiatry, 47, 511–523.
end of the spectrum. Autism Research, 6, 468–478. Zubrick, S. R., Taylor, C. L., Rice, M. L., & Slegers, D. W. (2007).
Talbott, M. R., Nelson, C. A., & Tager-Flusberg, H. (2015a). Di- Late language emergence at 24 months: An epidemiological
ary reports of concerns in mothers of infant siblings of children study of prevalence, predictors, and covariates. Journal of Speech,
with autism across the first year of life. Journal of Autism and Language, and Hearing Research, 50, 1562–1592.
Developmental Disorders, 45, 2187–2199. Zwaigenbaum, L., Thurm, A., Stone, W., Baranek, G., Bryson, S.,
Talbott, M. R., Nelson, C. A., & Tager-Flusberg, H. (2015b). Ma- Iverson, J., . . . Sigman, M. (2007). Studying the emergence of
ternal gesture use and language development in infant siblings autism spectrum disorders in high-risk infants: Methodological
of children with autism spectrum disorder. Journal of Autism and practical issues. Journal of Autism and Developmental
and Developmental Disorders, 45, 4–14. Disorders, 37, 466–480.

154 Journal of Speech, Language, and Hearing Research • Vol. 59 • 143–154 • February 2016