Each class, students will be responsible for reading three research articles and completing written responses to thereadings for that day. These responses are meant to facilitate class discussion of the assigned material. For eachresponse, students will be required to write a one-page paper synthesizing the three articles and include at least 1discussion question per article that they are prepared to bring up in class that day. Article Response assignmentsmust be turned in online via Canvas by 12:00 PM (noon) on the day of class (Thursday). Only assignmentssubmitted ON TIME via the designated link on Canvas will be graded. NO EXCEPTIONS.The following factors will be considered in grading: relevance, accuracy, synthetization of the reading materials,degree to which the responses show understanding/comprehension of the material, and quality of writing.Times New Roman 12Remember citations and intext citationsPlease find the 3 articles attached
INT J LANG COMMUN DISORD, JULY
VOL. 50, NO. 4, 488–498
Assessing early communication skills at 12 months: a retrospective study of
Autism Spectrum Disorder
Nathaniel Robert Swain†, Patricia Ann Eadie†, Margot Ruth Prior‡ and Sheena Reilly§¶
†Department of Audiology & Speech Pathology, University of Melbourne, Parkville, VIC, Australia
‡School of Behavioural Sciences, University of Melbourne, Carlton, Melbourne, VIC, Australia
§Murdoch Childrens Research Institute, Hearing, Language and Literacy, Parkville, Melbourne, VIC, Australia
¶Royal Children’s Hospital, Parkville, Melbourne, VIC, Australia
Department of Paediatrics, University of Melbourne, Parkville, Melbourne, VIC, Australia
(Received February 2014; accepted November 2014)
Background: Early identification of Autism Spectrum Disorder (ASD) is currently limited by the absence of reliable
biological markers for the disorder, as well as the reliability of screening and assessment tools for children aged
between 6 and 18 months. Ongoing research has demonstrated the importance of early social communication skills
in differentiating children later diagnosed with ASD from their typically developing (TD) peers, but researchers
have not yet investigated whether these differences can be detected using community-ascertained systematic
observation data as early as 12 months.
Aims: To investigate whether differences in early social communication skills can be detected at 12 months of age,
comparing children later diagnosed with ASD, and TD peers; and to determine whether differences remain when
groupings are based on age of subsequent ASD diagnosis.
Methods & Procedures: From a prospective community-ascertained sample, we collected data on children in early
life, then conducted retrospective analyses for those children who were later diagnosed with ASD by the age of
7 years, compared with matched TD peers. We analysed standardized observational data of early communication
skills, collected using the Communication and Symbolic Behavior Scales—Developmental Profile (CSBS-DP)
Behavior Sample, when participants were 12 months of age.
Outcomes & Results: Children in the ASD group exhibited significantly lower social communication skills than the
TD group, including on the Total score and Social and Symbolic Composite scores of the CSBS-DP Behavior
Sample. Differences on the Total score and Social Composite were also detected for both early and late ASD
diagnosis groups when compared with the TD group.
Conclusions & Implications: These findings give further support for the importance of social communication in
assessing children at risk of ASD as early as 12 months of age. Future research could evaluate the sensitivity and
specificity of direct observation of these early communication skills as diagnostic indicators for ASD at 12 months,
and investigate whether it is possible to distinguish between ASD and other high-risk groups (e.g. developmental
delay) at this age.
Keywords: autism spectrum disorders (ASD), early communication behaviours, early identification, assessment.
What this paper adds?
What is already known on the subject?
Previous retrospective and ‘at-risk’ clinical studies have highlighted the importance of early social communication
skills as factors that can differentiate ASD and TD groups. In community-based prospective studies, significant group
differences for children between 18 and 24 months of age have also been found, but research has not previously
utilized prospective methods with community-ascertained samples to detect differences at 12 months.
Address correspondence to: Nathaniel Robert Swain, Department of Audiology & Speech Pathology, University of Melbourne, Parkville, VIC
3052, Australia; e-mail: firstname.lastname@example.org
International Journal of Language & Communication Disorders
C 2015 Royal College of Speech and Language Therapists
ISSN 1368-2822 print/ISSN 1460-6984 online
ASD communication skills at 12 months
What this paper adds?
When comparing ASD and TD groups of children at 12 months, significant differences in social communication
skills were found, with some differences also detected between the TD group and two ASD subgroups (divided into
early and late diagnosis). Assessing social communication skills at 12 months may be equally useful for identifying
children at risk of ASD as it is at later ages.
Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by difficulties in social
communication and the existence of restricted and
repetitive behaviours. Evidence suggests that the optimal
time for intervention of ASD is between 18 months and
4 years (Prior et al. 2011). Current diagnostic practices
do not reflect these findings. A clinical diagnosis of
ASD at 2 years has been shown to remain stable for
3 years for 90% of children (Stone et al. 1999). However, diagnoses of ASD are typically made between 3
and 4 years of age or later (Ingersoll 2011). This is partly
due to lack of biological markers, as well as the low
utilization of age-appropriate diagnostic tools, though
new assessments for younger children are emerging, e.g.
the Autism Diagnostic Observation Schedule—2nd
Edition (ADOS-2), which is appropriate for individuals
who are 12 months and over (Lord et al. 2012).
Social communication skills are critical for the diagnosis of ASD (Wiggins et al. 2012). The core social
communication deficits of young children with ASD include joint attention, capacity for vocal communication,
conventional and symbolic gesture, and symbolic play
(Wetherby and Woods 2002). Typically, this skill set
develops throughout the first 24 months of life, and is
dependent on the formation of pre-linguistic skills in the
first 12–18 months, including ‘gaze shifts, shared positive affect, gaze/point following, communication for
joint attention, conventional gestures, communicative
vocalizations, language comprehension, and symbolic
play’ (Wetherby et al. 2004: 474). This set of early communicative skills has been widely investigated for the
identification of young children at risk of ASD, in lieu
of biological markers which are yet to be reliably established (Volkmar et al. 2004). Research on the social
communication skills of young children has endeavoured to identify skills that can detect ASD, utilizing
various methodologies. However, there is limited research of children at earlier ages using direct observation
data collected prior to diagnosis.
Early behavioural markers
Researchers have attempted to detect and describe the
markers of ASD from as early as 12 months using
retrospective methodologies including parental report
and analysis of ‘home videos’. Differences between children with ASD and typically developing (TD) or developmentally delayed (DD) children have been reported
by parents in areas of social engagement (showing, pointing, eye gaze, interaction), receptive communication and
affective engagement (Watson et al. 2007). Some limitations with retrospective parental report methods include
the interval between observation and recollection; lack
of blinding to diagnosis; and lack of clarity in descriptions of, or ages at which, behaviours were observed
(Wimpory et al. 2000).
Other research groups have utilized home videos to
retrospectively compare groups of young children (e.g.
Werner and Dawson 2005). A review of home-video
studies concluded that signs that differentiated autism
from developmental delay in the first 2 years of life were
difficulties with response to own name, looking at others, eye contact, affect and intersubjective behaviours
such as shared attention (Saint-Georges et al. 2010).
While home-video studies reduce some of the biases of
parental report studies, their limitations include inconsistencies in length, quality, content and context of video
samples, and variable use of control groups (Ozonoff
et al. 2008).
Prospective studies have further investigated the early
social communication profiles of children with ASD,
reducing some of the biases of retrospective research.
Prospective parent report studies have used such tools
as the Modified Checklist for Autism in Toddlers
(M-CHAT) to screen toddlers during well-baby checkups (Robins 2008). Predictive ASD markers have included difficulty or delays with interest in other children,
pointing to objects, imitation of caregiver, showing objects, name orientation and following a point (Robins
2008). Another line of research using parental report
involves measures such as the Communication and
Symbolic Behavior Scales—Developmental Profile
(CSBS-DP) (Wetherby and Prizant 2002), utilizing the
Infant–Toddler Checklist (I-TC)—the parent report
screener component. Veness et al. (2012) reported on a
prospective longitudinal study using the I-TC, drawing
participants from the larger Early Language in Victoria
Study (ELVS) cohort and comparing groups of children
with ASD (n = 18), developmental delay (DD) (n =
16) and specific language impairment (SLI) (n = 20),
as well as TD children (n = 60). In this exploratory
study, Veness et al. demonstrated the importance of social communication skills for identifying children at risk
of ASD at 12 months. They found that the ASD group
was significantly poorer than TD children on each of
the seven CSBS cluster scores, and all individual items
except ‘understanding of words’. In contrast to findings
from home-video and at-risk literature, ‘name orientation’ did not differentiate any groups at 12 or 24 months.
Also, the authors did not find any specific items that
differentiated between the ASD groups and SLI or DD
groups. However, in an update of this study incorporating children who had received ASD diagnoses by age
7, Veness et al. (2014) demonstrated that children with
ASD could be differentiated from TD and LI groups using prospective parental report at 8, 12 and 24 months
of age, particularly in the ‘use of gestures’ and ‘use of
communication’ cluster scores. The ASD group could
also be differentiated from all groups including DD,
again on the ‘use of communication’ cluster but only at
24 months of age.
Prospective research has also studied those at risk of
developing ASD, such as siblings of children with an
ASD diagnosis, or children referred for developmental
evaluation. A comprehensive review by Zwaigenbaum
et al. (2009) summarized the findings of these reportbased or observational studies, which showed that at
12–18 months old children with ASD could be distinguished from other clinical groups by deficits in any one
or more of the following: visual, motor, play, socialcommunication, language and cognitive domains.
At-risk studies have provided valuable information
about the development of autism. However, preselection and recruitment biases are possible given that recruitment relies on either family background of ASD,
or children referred for evaluation of DD. Also, participants in these studies may only represent ‘high risk’
cases, and findings may therefore not generalize to the
whole population (Zwaigenbaum et al. 2009).
Prospective population-based studies incorporating
direct observation have the most robust methodologies,
and are therefore the gold standard for the study of
ASD in early life. The ‘One-year well-baby check-up
approach’ (Pierce et al. 2011), and the Social Attention and Communication Study (SACS) (Barbaro and
Dissanayake 2012) are examples of this approach. Both
research groups have developed tools to screen prospectively for ASD. The SACS predictors of ASD diagnosis
included failure to exhibit pointing, waving, imitation,
eye contact and response to name at 12 months, while
factors relating to speech development were not significant (e.g. babbling, early words, attending to sounds)
(Barbaro and Dissanayake 2012).
Nathaniel R. Swain et al.
In contrast to less formal protocols, the CSBS-DP
Behavior Sample (BS) is one of the few structured observational procedures available for infants at 12 months
(Wetherby and Prizant 2002). As part of the FIRST
project, Wetherby et al. (2004) used BS as a
secondary evaluation for children, screened with I-TC.
Video recorded samples of three groups (ASD, DD,
TD) of eighteen 21-month-old children were used to
identify red flags for ASD in the second year of life. The
results verified the BS’s application as a secondary evaluation to I-TC, but did not show significant differences
between ASD and DD groups. A subsequent analysis of
the data using the Systematic Observation of Red Flags
(SORF) for Autism Spectrum Disorders in Young Children (Wetherby and Woods 2002), revealed nine indicators that did differentiate ASD from DD and TD: (1)
lack of appropriate gaze; (2) lack of warm, joyful expressions with gaze; (3) lack of sharing enjoyment or interest;
(4) lack of response to name; (5) lack of coordination
of gaze, facial expression, gesture, and sound; (6) lack of
showing; (7) unusual prosody; (8) repetitive movements
of the body; and (9) repetitive movements with objects.
The authors identified four additional flags that differentiated ASD from TD but not DD including a lack of
responding to contextual cues, pointing, vocalizations
with consonants and conventional toy play (Wetherby
et al. 2004: 485). Used together to predict group membership, the 13 red flags correctly categorized 94.4% of
the participants, showing that children with ASD can
be distinguished from DD and TD around 21 months
on the basis of these observed behaviours. A subsequent
study by Wetherby et al. (2007) also used BS on a similar cohort of participants (mean age of 21 months)
comprising 50 children with ASD, 23 DD and 50 TD.
This study found that children with ASD performed
significantly poorer than DD on five communicative
measures (gaze shifts, gaze/point follow, rate of communication, acts for joint attention, inventory of conventional gestures), and lower than the TD group on all
measures (Wetherby et al. 2007). Studies investigating
the early social communication skills of children with
ASD have utilized a variety of methodologies, producing
both complementary and contrasting findings. Increasingly rigorous and reliable evidence has been gathered,
demonstrating the importance of these skills for the
identification of ASD in early life.
Differences between subtypes of ASD
With the introduction of DSM 5 (American Psychiatric
Association 2013) there has been much discussion about
subtypes in ASD diagnosis. Evidence from studies of
the social communication skills of children with autism
versus children with Asperger’s syndrome in preschoolers (e.g. Ramberg et al. 1996), and early life through
ASD communication skills at 12 months
retrospective parental report (e.g. Ozonoff et al. 2000),
suggests there is little evidence of qualitative differences
between these subtypes in early life. Rather the groups
tend to differ in severity of symptoms along a spectrum (Macintosh and Dissanayake 2004, Wiggins et al.
One factor on which autism and Asperger’s groups
tend to differ is the age of receiving a diagnosis; with
autism diagnoses generally received significantly earlier
than Asperger’s (Howlin and Asgharian 1999, Mandell
et al. 2005). In order to examine differences between
these subgroups based on severity of symptoms, some
studies have used ‘age of diagnosis’ to delineate between groups, though the cut-off points utilized have
varied based on study design. For example, Landa et al.
(2007) prospectively studied a clinical sample of infants
at high and low risk of ASD, including early and late
ASD diagnosis groups (defined as meeting diagnostic
criteria before or after 14 months). The authors compared these groups with TD children (referred to as low
risk controls) and siblings of children with ASD with
and without a Broader Autism Phenotype classification
(BAP, defined as exhibiting social or language delays,
but without a clinical judgement of ASD). Assessed using the CSBS-DP BS at 14 months, the early diagnosis
group was found to be lower than non-BAP group on all
BS items, and lower than all non-ASD groups (including BAP, non-BAP and TD children) on ‘gaze shifts’,
‘action schema sequences’, ‘initiation of joint attention’,
‘behavior regulatory bids’, and ‘inventory of gestures
and consonants in syllables’. Differences between early
and late ASD diagnostic groups were also detected at this
age, including the early group exhibiting lower scores on
BS items ‘positive affect’, ‘initiation of behaviour regulatory bids’ and ‘initiation of joint attention’. The late
diagnosis group only differed from non-ASD groups on
‘gaze shifts’. However, more significant differences between the late diagnosis group and non-ASD groups
were present by 24 months, including lower scores on
‘shared positive affect’ and ‘inventory of gestures’ (Landa
et al. 2007). In contrast to the cut-off used in Landa et al.
(2007), Twyman et al. (2009) conducted a retrospective
parental report study using an early diagnosis cut-off
of before 36 months, finding that parents in the early
diagnosis group had concerns with social development
at an earlier age than those in the late diagnosis group
(mean of 18 months versus a mean of 25.3 months). As
reviewed above, the previous research investigating the
social communication skills of children with early and
late diagnoses of ASD has either been retrospective and
reliant on parental report, or has utilized ‘at-risk’ clinical
samples—the findings of which may not be generalizable to the whole population. Also, previous studies have
not examined differences between ASD and TD groups
as early as 12 months of age.
While there is now a large body of work that attests
to early social communication skills being crucial for
the detection of ASD, prospective community-based research using a direct observation tool has only compared
children 18 months and older. Research with younger
children has been primarily based on home videos or
parent report measures, or utilizing sampling methods
which may have introduced bias, such as at-risk and sibling studies. Hence, we report here the results from a systematic observation study of the social communication
skills of 12 month olds drawn from a larger prospective
study of language development in a community sample.
We identified children who were diagnosed with
ASD by the age of 7 years, and conducted retrospective
analyses of data collected at 12 months using the CSBSDP Behavior Sample. A group of matched TD children,
who had also completed this assessment at 12 months,
were compared with the ASD group to investigate differences in social communication. All observations and
coding were completed prior to ASD diagnosis. This
study investigated: (1) if differences in early social communication skills can be detected between children with
ASD and a group of matched TD peers at 12 months;
and (2) whether these differences can be detected between the TD group and two ASD subgroups divided
into early or late diagnosis.
Overview of ELVS
The Early Language in Victoria Study (ELVS) is a
prospective longitudinal population study of the development of language and literacy problems of a total of 1,910 Australian children, which commenced in
2002. Sampling methods for this study have been outlined in detail elsewhere (e.g. Reilly et al. 2006, Veness
et al. 2012). Briefly, infants were recruited at 7.5–10
months from six of the 31 Local Government Areas
(LGAs), in metropolitan Melbourne, Australia, chosen
to represent geographically non-adjoining areas, which
span the spectrum of disadvantage–advantage, and correspond to the Australian 2001 Census-based SocioEconomic Indexes for Areas (SEIFA) Index of Disadvantage. Participants were recruited through Maternal
and Child Health (MCH) nurses, through universally
available hearing screening sessions, or local newspaper
publicity. Infants with any known developmental delay
or disability (DD) were excluded at time of recruitment,
along with children of caregivers who could not speak
and/or understand English. Parents completed questionnaires about their child’s development at 8 months, and
subsequently at 1, 2, 3 and 4 years of age. This included the CSBS-DP I-TC completed by all participants
Nathaniel R. Swain et al.
(n = 1911) at 8 months, and participants followed up
at 12 months (n = 1725). The attrition between the
8- and 12-month data collections were due to families
failing to return questionnaires at the 12-month collection (Reilly et al. 2006). At 12 months, the CSBS-DP
BS was administered as an additional face-to-face assessment on two subsets of the ELVS cohort (n = 728)
as budget constraints at that time did not allow for assessment of the whole cohort. The first subgroup of
participants comprised a large number of children who
were identified as having a score in the lowest quartile on
the 8-month I-TC, i.e. at or below the 25th percentile
(raw score equivalent ࣘ 32; n = 347). The second subgroup contained randomly selected children, matched
to the first subgroup for age and gender, who achieved
a percentile rank above 25 on the 8-month I-TC (raw
score equivalent > 32; n = 381). The oversampling of
the lowest quartile was completed to ensure that the
research project met the aims of the main ELVS study
(which included identifying potential early signs of language delay). At 4 years of age individual assessments for
all ELVS participants were administered including measures of speech, language and cognition. Further details
of the ELVS methodology and study sample, including
parent report and face-to-face assessments are reported
in Reilly et al. (2006) and Prior et al. (2008).
Participants in current study
The participants for the present study comprised two
groups retrospectively drawn from the ELVS cohort:
children with ASD (n = 22), and a control group of
children with typical development (TD) (n = 22). The
ASD group comprised the children who had been given
a clinical diagnosis of ASD by a community health professional or assessment team between the ages of two and
seven and had completed the CSBS Behaviour Sample
at 12 months.
The TD group comprised participants who were
randomly selected and then matched to the ASD group
on gender and a measure of socio-economic status (the
Australian 2001 Census-based Socio-Economic Indexes
for Areas (SEIFA) Index of Disadvantage). The TD
group had also completed the BS at 12 months, and
had maintained results consistent with typical development through to 4 years. Figure 1 describes the flow of
participants from completion of the I-TC at 8 months
to the sampling in the current study.
The ELVS questionnaires collected demographic and
background data for each participant in the wider study.
Variables were chosen based on factors that are thought
to influence a child’s language and general development.
Figure 1. Flowchart of Early Language in Victoria Study (ELVS)
participants and those included in the current study of Autism Spectrum Disorder (ASD) using the Communication and Symbolic Behavior Scales—Developmental Profile (CSBS-DP) Behavior Sample.
Participant characteristics are summarized in table 1.
The matching process of the TD group controlled for
the potential effects of gender and socio-economic status between the groups. All other salient variables were
consistent between the ASD and TD groups, including
twin or premature birth, birth weight and order, as well
as family factors such as English speaking background,
family history of language problems, maternal education, mental health and vocabulary score, and maternal
age at birth of child.
Measures and procedures
Determining an ASD diagnosis
As previously reported by Veness et al. (2014), information regarding the clinical ASD diagnosis of children was
based on the standard diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders (DSMIV). These diagnoses were made external to the ELVS
research team, according to standard clinical practice in
Australia, by multidisciplinary teams including paediatricians, speech pathologists and psychologists. Parents
reported via questionnaire or telephone if their child had
ASD communication skills at 12 months
Table 1. Summary of participant demographic characteristics
TD (n = 22)
Premature birth (
Purchase answer to see full