brain

sciences
Article
Developmental Language Disorder and Autism: Commonalities
and Differences on Language
Natasa Georgiou and George Spanoudis *

Department of Psychology, University of Cyprus, P.O. Box 20537, Nicosia 1678, Cyprus;
[email protected]
* Correspondence: [email protected]

Abstract: Language and communication deficits characterize both autism spectrum disorder and
developmental language disorder, and the possibility of there being a common profile of these is a
matter of tireless debate in the research community. This experimental study addresses the relation of
these two developmental conditions in the critical topic of language. A total of 103 children (79 males,
24 females) participated in the present study. Specifically, the study’s sample consisted of 40 children
with autism, 28 children with developmental language disorder, and 35 typically developing children
between 6 and 12 years old. All children completed language and cognitive measures. The results
showed that there is a subgroup inside the autism group of children who demonstrate language
difficulties similar to children with developmental language disorder. Specifically, two different
subgroups were derived from the autism group; those with language impairment and those without.
Both autism and language-impaired groups scored lower than typically developing children on all
language measures indicating a common pathology in language ability. The results of this study
shed light on the relation between the two disorders, supporting the assumption of a subgroup



with language impairment inside the autism spectrum disorder population. The common picture

 presented by the two developmental conditions highlights the need for further research in the field.
Citation: Georgiou, N.; Spanoudis,
G. Developmental Language Keywords: developmental language disorder (DLD); autism spectrum disorder (ASD); comorbidity;
Disorder and Autism: Commonalities autism spectrum disorder with language impairment (ASD-LI); autism spectrum disorder without
and Differences on Language. Brain language impairment (ALN); language development
Sci. 2021, 11, 589. https://doi.org/
10.3390/brainsci11050589

Academic Editor: Marianthi Georgitsi 1. Theoretical Premises

A growing body of research focuses on the investigation of language and communica-
Received: 1 April 2021
Accepted: 29 April 2021
tion deficits observed in children with developmental disorders, such as Developmental
Published: 30 April 2021
Language Disorder (DLD) and Autism Spectrum Disorder (ASD) (e.g., [1,2]). Individuals
diagnosed with DLD are found to have difficulties in areas related to the structure of a
Publisher’s Note: MDPI stays neutral
language, such as phonology, morphology, syntax, and semantics. It has been suggested
with regard to jurisdictional claims in
that individuals diagnosed with ASD have difficulties in another category associated with
published maps and institutional affil- language, namely pragmatics. Pragmatics resides at the top of the hierarchy of the building
iations. blocks of language and is defined as one’s ability to use the language appropriately in
each context [3]. Deficits in this last component of language led some researchers to the
conclusion that structural language deficits may also be present in ASD [4]. Nowadays, the
research community has raised its interest in whether ASD and DLD disorders overlap, and
examines the possibility that these two disorders may consist of different manifestations of
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
the same underlying cause.
This article is an open access article
1.1. Developmental Language Disorder
distributed under the terms and
conditions of the Creative Commons Developmental Language Disorder (DLD) is not a term included in the Diagnostic and
Attribution (CC BY) license (https:// Statistical Manual of Mental Disorders (DSM-5), however, it is widely accepted amongst
creativecommons.org/licenses/by/ both the scientific community and clinicians. Instead, the DSM-5 includes the term “Lan-
4.0/). guage Disorder” (within neurodevelopmental disorders) to define persistent difficulties in

Brain Sci. 2021, 11, 589. https://doi.org/10.3390/brainsci11050589 https://www.mdpi.com/journal/brainsci

Brain Sci. 2021, 11, 589 2 of 29

the acquisition and use of language, because of impairments presented in comprehension or

production [5]. Until recently, it was widely referred to as “Specific Language Impairment
(SLI)”. Nevertheless, in the recent past the term proposed by Bishop and colleagues [6],
namely “Developmental Language Disorder” (DLD; CATALISE-2 consortium; Criteria and
Terminology Applied to Language Impairments: Synthesizing the Evidence), has been
widely accepted and used extensively. In the present study, we will refer to this disorder
using this term (Developmental Language Disorder; DLD). The symptomatology of DLD
includes difficulties that create obstacles to communication or learning in everyday life,
and those language problems are unlikely to resolve or will have not been resolved by five
years of age and are not associated with a known biomedical condition such as brain injury,
neurodegenerative conditions, genetic conditions or chromosome disorders such as Down
syndrome, hearing loss, autism spectrum disorder or intellectual disability [6].
DLD is a heterogeneous disorder; symptoms can be either expressive (e.g., syntax,
vocabulary, phonology, and motor skills), receptive (i.e., comprehensive skills), or a com-
bination of the two [7]. Recently, more empirical evidence supports that children with
DLD are also deficient in pragmatics, a non-core feature of language [8]. It has also been
suggested that a subtype of DLD exists and presents reduced pragmatic abilities [9].

1.2. Autism Spectrum Disorder

Autism is a disorder characterized primarily by social deficits, as two of the three main
diagnostic features, namely impaired communication and reciprocal social interaction,
exist within the social domain. According to the DSM-5, a diagnosis of ASD has to include
deficits in (1) social communication, and social interaction which occur across multiple
contexts and (2) restricted repetitive behaviors, interests, and activities [5].
Although structural language impairment as part of a communication deficit, is not a
necessity for an ASD diagnosis (according to DSM-5), it is strongly associated with ASD [4].
Specifically, it has been stated that approximately 63% of all children diagnosed with ASD
have language impairment [10]. It has also been found that over half of the individuals
with ASD have additional deficits in other levels of the structure of a language, such as in
phonology, grammar, and semantics [11]. A delay in language acquisition and a lack of the
ability to use language appropriately in social contexts (lack of pragmatic abilities) are the
core features of ASD. Pragmatic abilities are a hallmark feature of the disorder, regardless
of an individual’s level of functioning [2,12].
ASD is a highly heterogeneous disorder, with a wide range of abilities in language
and cognition throughout the spectrum. Understanding the nature of these deficits is
complicated by the great heterogeneity of language and cognitive abilities within this
population [2]. For instance, although many ASD individuals present expressive and
receptive language impairments, there is also an important number of ASD individuals
who do not display any language deficits (ALN subgroup) [13]. This variability complicates
the effort to understand the nature of the cognitive and language deficits of the disorder.
Language impairment is the primary symptom of DLD condition. As previously men-
tioned, studies are supporting the existence of language deficits in ASD individuals [14–16].
This finding triggered the scientific community to further investigate the relation of the two
developmental conditions in the domain of language. Research evidence so far has pro-
vided some interesting findings pointing at commonalities in the language domain [15,16].
Data from the use of language tasks along with data from cognition neuroimaging and
genetics in ASD children provide evidence for the existence of an ASD-LI (ASD plus
Language Impairment) subgroup within the ASD group [14–16]. Figure 1 represents the
ASD-LI “subgroup” of shared symptoms between the two disorders.
ci. 2021, 11, 589 3 of 30
Brain Sci. 2021, 11, 589 3 of 30

Language Impairment) subgroup within the ASD group [14–16]. Figure 1 represents the
Brain Sci. 2021, 11, 589
ASD-LI Language
“subgroup” Impairment)
of shared subgroup
symptoms within
between the disorders.
the two ASD group [14–16]. Figure 1 represents3 of
the29
ASD-LI “subgroup” of shared symptoms between the two disorders.

Figure 1. A depiction of the proposed intermediate category ASD-LI, sharing features from both
Figure 1. A depiction of the proposed intermediate category ASD-LI, sharing features from both
ASD and DLD. Figure 1. A depiction of the proposed intermediate category ASD-LI, sharing features from both
ASD and
ASD and DLD.
DLD.
In this figure, In
an this
overlap is identified
figure, an overlapbetween the two
is identified disorders.
between This
the two overlap, this
disorders. This overlap, this
common area common In this
between figure, an overlap is been
identified betweenby thedifferent
two disorders. This overlap, this
area between the two disorders has been documented bystudies
the two disorders has documented de-studies
different despite
common
spite the widely acceptedareaview
between
of thebeing
them two disorders has been
characterized by documented
different by different
symptoms [14– studies de-
the widely accepted view of them being characterized by different symptoms [14–16]. A
16]. A possiblespite
linkthe widelyASD
between accepted
and viewhas
DLD of them
been being characterized ofby different symptoms [14–
possible link between ASD and DLD hasstated as a matter
been stated as a matterinvestigation
of investigation for addi-
16].
for additional tionalA possible
research. Studies link between
to date ASD and DLD has been stated as a matter of investigation
research. Studies tohad
dateallowed researchers
had allowed to draw
researchers different
to draw conclu-
different conclusions based
sions based on forcontroversial
additional research.
findings. Studies to date had allowed researchers to draw different conclu-
on controversial findings.Currently,
Currently,there
therearearetwo
twokey
keyassumptions
assumptionswithin
within the research
sions
the research communitybased on controversial
addressing the findings.
relation of ASDCurrently,
and there
DLD. The are
twotwo key assumptions
different
community addressing the relation of ASD and DLD. The two different perspectives per- within
are
the research
spectives are presented in community
Figure
presented in Figure 2. 2. addressing the relation of ASD and DLD. The two different per-
spectives are presented in Figure 2.

Figure 2. Models
Figureof2.the relationship
Models between DLD
of the relationship and ASD.
between DLDThey
anddepict
ASD. different theories
They depict explaining
different theex-
theories relation between
ASD and DLD. Figurebetween
plaining the relation 2. Models of and
ASD the relationship
DLD. between DLD and ASD. They depict different theories ex-
plaining the relation between ASD and DLD.
In particular, one part of the research community argues that the two disorders are
In particular, one part of the research community argues that the two disorders are
distinct and attribute
In particular, any
part common
one groundof the ground
research to superficial
community similarities;
argues (e.g., [2,13,17,18])
distinct and attribute any common to superficial similarities; (e.g., that the two
[2,13,17,18]) disorders are
while
distinct the other part of the community argues that the two disorders consist of different
while the other part ofand
theattribute
community anyargues
common thatground
the twotodisorders
superficial similarities;
consist (e.g., [2,13,17,18])
of different
manifestations of theofsame pathology and lie that
on the
thesame continuum [19]. This second
manifestationswhile
of thethesame
otherpathology
part the community
and lie on theargues
same continuum two disorders
[19]. consist
This second of different
assumption has received important support from researchers supporting
manifestations of the same pathology and lie on the same continuum [19]. This second that ASD and
DLD disorders are comorbid (e.g., [15,16]).
The debate regarding whether there is a common etiological phenotype between
the two disorders is still ongoing. Research so far cannot provide conclusions on the
similarities and differences between the two disorders. The need for more research in the
Brain Sci. 2021, 11, 589 4 of 29

area is underlined, to provide evidence either to support or dismiss the assumption of a

common etiology behind DLD and ASD.

1.3. A Potential Overlap between ASD and DLD: Language Development in DLD
Language problems are common in children, with DLD prevalence estimated to be
7.5% [20]. Children with DLD present impairments in language acquisition and use, despite
having normal nonverbal intelligence and in the absence of other significant developmen-
tal/hearing deficits, autism, and severe neurological impairment [21]. Language abilities
of individuals with DLD are assessed through multiple measures of both expressive and
receptive language. Children with DLD demonstrate deficits in verbal working memory
(e.g., [22,23]) and on nonverbal as well [23,24].
Research so far, in an attempt to identify the specific causes of DLD, has pointed out
some clinical markers, related to DLD deficits [25]. Specific errors of tense marking (i.e.,
omission of the past tense marker, 3rd person present tense marker) [26], poor sentence
repetition [27], and poor non-word repetition are all considered as discriminative markers
of DLD children from typically developing (TD) children [28].
Specifically, tense marking ability has been documented as a DLD indicator by the
early studies of Rice et al. [29] and Marchman et al. [30]. It has been identified as a key
feature of children with DLD, examined by an elicitation task, and a task of the use of
third person singular [31]. Children with DLD were also found to demonstrate poor
performance on syntactic tense throughout the primary school years [30]. Furthermore,
sentence repetition is a good indicator of DLD, for children speaking English [27] or dialects
of English [32], also Cantonese [33], French [34,35], and Greek Cypriot [36].
For example, in the first study conducted in Cyprus, measuring and validating Sen-
tence Repetition Task (SRT) in Greek-Cypriot children, results showed significant differ-
ences, as the group of children with DLD scored poorer than the TD group of children
participated in the study [36]. The results from the English speaking population are re-
ported by Conti-Ramsden et al. [27], who investigated different potential clinical markers
of DLD and found that sentence repetition (also third person singular task, tense marking,
and non-word repetition) could be the strongest clinical marker for the identification of
DLD. A repetition task considers contributing to the identification of children’s weaknesses
and strengths in the language domain [37].
Another important clinical marker in the field is non-word repetition. A majority of
research has pointed out a strong connection between consistently poor performance on
non-word repetition tasks and DLD (see [38]). In these tasks, listeners are first asked to
hear a made-up word modeled after their native language, (e.g., “vonk” in English, or
“Bηκα” in Greek), and then, repeat it back immediately. The non-word repetition task is
considering an excellent clinical marker to distinguish DLD individuals from typically
developing individuals [1,28].
Especially, non-word repetition tasks received a great deal of attention in investigations
of language impairments. Indeed, their connection to language impairments occurs reliably
across the literature [39,40]. Non-word repetition tasks have received significant attention
in studies of language impairments because of their ability to comprehend the underlying
deficits of children with DLD and acts as a possible identifier of these children. A meta-
analysis provided evidence from different studies indicating that children with DLD
displayed difficulty repeating not only long non-words but even short nonwords [40].
Tasks such as Non-Word Repetition (NWR) and Sentence Repetition (SNT), have
received great research interest as psycholinguistic markers of DLD [27,41,42]. NWR
and SNT measures are used as phonological short-term memory indicators, and putative
markers for DLD [41]. NWR can discriminate DLD from other childhood communication
disorders [42] and exceeds other language measures such as past tense marking in its
specificity and sensitivity to detect language-impaired individuals [27].
In addition to the above, it is also important to make a brief reference to the field of
pragmatic language and its relation to DLD. Pragmatic language is strongly associated with
Brain Sci. 2021, 11, 589 5 of 29

deficits presented in ASD individuals, whereas, its relation to individuals with DLD ability,
is still not determined. On the one hand, studies are supporting additional pragmatic
impairments in children with DLD [3,43,44] while on the other, evidence of intact pragmatic
ability, is provided [45,46], highlighting the need for more research in the field.
DLD has been extensively studied in several different languages, and interesting
results emerged from studies conducted in the Greek language. Studies in Greek-speaking
children with DLD revealed important results especially in relation to specific domains of
linguistic or cognitive difficulties.
The main findings indicate difficulties of children with DLD in their ability to compre-
hend thematic roles, namely, grammatical subject–object relations in complex main-relative
clauses, by utilizing pro- and post-verb clitics about their peers [47], and in the production
and understanding of pronouns [48,49]. In addition, longitudinal studies inform for persis-
tent impairments in specific aspects of language, such as persistent problems with specific
grammatical operations 5 years later [50,51].
Interestingly, controversial results are reported by studies investigating clitic pro-
duction in Greek-speaking children with DLD. Studies are reporting that children with
DLD omit direct-object clitics [52,53] while others report no differences between children
with DLD and TD peers in similar tasks [54–56]. Narratives have also been tested in
Greek-speaking children and found to be effective in differentiating DLD children from
TD [57]. The authors noted that narratives can indeed be a useful tool to identify and assess
language-impaired children in the Greek language.
Additional findings support that children with DLD have difficulties in understanding
the semantic and pragmatic functions of mental state verbs and working memory [58].
Mainly, research on the Greek language specifies difficulties that children with DLD per-
form in semantics, morphosyntax, and pragmatics and not necessarily in phonology
(see also [59]).
The majority of research in the Greek language focuses on specific areas of linguistic
ability. Nevertheless, more research is needed related to the linguistic abilities of DLD
children on all linguistic domains, especially in comparison to the abilities presented by
ASD children, to effectively locate the similarities and differences of the two disorders and
shed light on the predominate “comorbidity” question.

1.4. Language Ability in ASD

Language ability in autism was described from early on as impaired, especially con-
cerning the social use of language and to the quality of the spoken language (e.g., [60,61]).
However, not much investigation conducted in the following years related to the language
deficits in autism, possibly because early studies did not reveal any significant differences
in linguistic domains of phonology, syntax, and morphology, between children with ASD
and TD children, matched either on age, IQ or language (e.g., [62–65]).
Research so far provided evidence of higher-verbal children with ASD, who score
similar to TD children, based on a range of structural language measures [66]. The ability
in phonology, semantics, morphology, and syntax of a subgroup of ASD individuals
is furthermore supported by different studies [67,68]. It has been found that there are
individuals with ASD who present good or even elevated structural language abilities on
formal testing, by using sophisticated vocabulary and sentence structure [69,70].
The great heterogeneity observed within the ASD population is also apparent in the
extent of structural language difficulties that co-occur with ASD [4,66,71]. It is argued that a
percentage of 25% and 30% of children with ASD are minimally verbal or nonverbal, based
on studies ascertained via population-based or clinical samples [72,73]. The incidences
presented may vary according to individuals’ age, different study criteria of language
ability definitions, and by the way the sample was ascertained [74]. Nevertheless, impaired
language development is often the first irregular symptom to be identified by parents of
children who are later on diagnosed with ASD [75,76]. It has also been found that over half
of the individuals with ASD have additional deficits in different levels of the structure of a
Brain Sci. 2021, 11, 589 6 of 29

language, such as in phonology, grammar, and semantics [11]. In addition, the progress of
lower verbal children with ASD, found to be slower, and also presenting flatter language
development [66].
As it concerns the communication domain, diagnostic features refer to delays and
impairments in language acquisition, and qualitative deficits in discourse and other prag-
matic aspects of language as well. A great emphasis in the research community about
language in autism has focused its interest on the pragmatic impairments, which are treated
as universal among ASD individuals and are specific to this disorder [4,77]. The pragmatic
impairment, is a hallmark feature of the disorder, regardless of an individual’s level of
functioning [2,12].
In the Greek language, only limited studies included children with ASD. In a study
investigating the narrative skills of 30 Greek-speaking children with ASD (ALN and ASD-
LI) and a TD group, their narrative ability was measured in terms of both microstructural
and macrostructural properties. Microstructural properties are referring to lexical and
syntactic measures of complexity and macrostructure to the diversity in the use of internal
state terms (ISTs) and the complexity of the story’s structure. The results showed that
increased language ability and syntactic complexity are positively connected to ASD
children’s narrative performance. However, both ALN and ASD-LI groups of children
with ASD scored lower than the TD controls in measures of the production of Theory of
Mind-unrelated ISTs, modifier clauses, and story structure complexity [78].
In an additional study in the Greek language with high-functioning ASD children,
results were indicative of difficulties in the ASD group at the interface of syntax with
pragmatics and prosody. Specifically, they produced significant deficits in their ability
to distinguish a discourse prominent element and to consider intonation relevant for a
particular interpretation that excludes clitics, in relation to their TD peers [79].
The difficulties of ASD-LI children in the Greek language are also supported by a
study with Greek-Cypriot children investigating comprehension and production of two-
constituent compound words. The results of the study indicate that despite the ASD-LI
children’s ability to recognize compound constituents they demonstrated impairments to
understand the compound meaning. In addition, they produced semantically inappropriate
responses when attempted to explain the meaning of the compounds, indicating limitations
in their conceptual-intentional system [80].
To sum up, ASD is a highly heterogeneous disorder, with impaired cognitive-language
abilities throughout the spectrum. Understanding the nature of these language deficits
is complicated by the great heterogeneity of language and cognitive abilities within this
population [2]. For instance, although many ASD individuals present expressive and
receptive language impairments, there is, simultaneously, an important number of ASD
individuals who do not display any language deficit [13]. This variability complicates the
effort to understand the nature of the cognitive and language deficits of the disorder. More
research is necessary to clarify the nature of the language impairments in (subgroups of)
children with ASD, and how it may be related to that presented in other developmental
conditions of language pathology, such as DLD.

1.5. ASD and DLD Children’s Relation on Language Ability

Many studies have examined any potential phenotypic overlap between autism and
DLD, especially in the language domain. Numerous studies have found that there are
many similarities in the language abilities of children with autism and children with DLD
and argue that the reason for these similarities is a shared etiology [14,15].

1.5.1. Similarities
Impaired language development is apparent in DLD and—at least—in subgroups of
ASD individuals. The relationship though, of these two developmental conditions with
language deficits, is still undetermined. In the last twenty years, several assumptions have
been made suggesting that a clear diagnostic border separating the two disorders may not
Brain Sci. 2021, 11, 589 7 of 29

emerge as clear as expected [15,16,81]. This hypothesis is supported by studies concluding

that there are common features between the two disorders [15,16]. Furthermore, Lindgren
and colleagues in their 2009 study discovered no statistically compelling differences be-
tween the two groups in a variety of language domains. Those discoveries of language
similarities, gave voice to the researchers to contemplate that there is a possible shared
etiology at the background of the two disorders [82].
In addition, Lloyd and colleagues [83] administered a standardized language test to
three groups of children, ASD, DLD, and a shared group with common symptoms from
both disorders. The results provided mixed results. Despite that, the composite scores
of the “shared” and ASD groups had the most similarities, and the DLD group had the
lower expressive language scores, all three groups had the least difficulty with “Listening
to Paragraphs” (which involves abstracting and remembering information from two short
texts). It also showed that all three groups had difficulty with “Recalling Sentences” (SRT;
which involves repeating sentences of increasing length and complexity). Comparable
performance in expressive language between the ASD and DLD groups measured by the
task of Repeating Sentences has also been stated by Manolitsi and Botting study [84]. In
addition, others demonstrated similar phonology [85] and vocabulary delays [85,86].
The assumption of a common phenotypical path linking the two developmental
conditions is furthermore strongly supported by McGregor and colleague’s findings [86].
In their study, they compared an ASD group of children, who were free of syntactic deficits
and had age-appropriate word knowledge, a group of children with ASD plus concomitant
syntactic language impairments (ASD-LI), a DLD group of children, an age-matched
group (AM), and a syntax-matched group (SM), both consisted by unaffected peers. Their
results showed that ASD-LI children, performed similarly to the DLD group, and both
showed sparse lexicons with partial word knowledge and immature knowledge of word—
relationships. The authors state for a behavioral overlap related to the syntax–lexicon
interface and point to the possibility of a commonality in the ASD-LI and DLD phenotypes.
The grammar ability of the two groups has also been studied by Tuller and col-
leagues [87], who stated that both ASD and DLD participants make mistakes with clitic
pronouns and complex sentences. They also reported that the ASD-LI group (and the chil-
dren with DLD) were particularly sensitive to consonants in syllable final position, notably
liquids and obstruents in internal coda position when examined on a standardized word
repetition task. ASD-LI and DLD groups were identified to develop the same strategies
in an attempt to avoid syllabic complexity, such as the omission of a segment. Riches and
colleagues [18] also reported that syllabic structure is problematic for adolescents with
ASD plus language impairment.
More recent data also highlight the possible overlap between the two conditions.
Specifically, the CELF-4 standardized test was administered to assess ASD, DLD, and
TD children’s language skills. The results showed no significant differences in language
between the DLD and ASD groups, suggesting an overlap in the linguistic profiles of
children with DLD and children with ASD [88].
Recent evidence supports the existence of common ground between the two conditions
in relation to the field of semantics. Gladfelter and Barron [89] explored whether global–
local processing differences influence the type of semantic features children with ASD,
DLD, and typically developing peers learn to produce when learning new words. The
results showed that the children from the clinical groups produced more global, in relation
to local, semantic features in their definitions than the TD group. This finding further
supports a common feature in semantic language and may reflect common deficits in depth
of word knowledge.
In addition, a study by Modyanova and colleagues [90] investigating the performance
of ALN and ASD-LI children in their ability in tense marking and morphosyntax revealed
a clear distinction in the morphosyntactic abilities of the two subgroups. ASD-LI scored
significantly worse with grammatical ability. The ASD-LI profile was closer to the one
usually performed by the DLD children, and perhaps even more impaired. A study
Brain Sci. 2021, 11, 589 8 of 29

comparing ASD and DLD groups on tasks of working memory and morphosyntactic
processing showed that children in both groups were more accurate and faster at detecting
errors occurring late, rather than early, in the sentence, indicating no evidence of different
patterns of performance for the DLD and ASD groups [91].
There is also research evidence supporting the existence of a common ground related
to the field of pragmatics. In a recent study, where ASD-high function (HFA), DLD and
TD groups of children participated, the groups were experimentally tested on various
components of grammar, pragmatics, and nonverbal cognition. The results showed similar
performance of DLD and HFA in pragmatics, lower than their TD peers. However, the
DLD group performed significantly lower than the TD children on grammar and several
cognition tests, while the HFA group did not [92]. Thus, although a similar profile is
identified on pragmatics, differences are also noted by this study.
Additionally, Norbury and colleagues [93] also investigated pragmatic language
and the narratives of ASD, DLD and TD children and revealed no significant differences
between children with ASD and DLD, who produced similarly simple narratives that
lacked semantic richness and omitted important story elements, when compared to TD
peers. Importantly, pragmatic errors were common across groups.
As it concerns the longitudinal outcomes of the proposed overlap, data from ado-
lescents come to add to the “commonality” hypothesis. Specifically, Conti-Ramsden and
colleagues [94], in their study, found that 3.9% of teenagers previously diagnosed with
DLD, showed sufficient characteristics of autistic behavior and met the criteria for an
autism diagnosis. Leyfer and colleagues [95] determined that a percentage of 41% of a DLD
diagnosed group of children, reached the diagnostic boundaries for ASD in social and com-
munication domains, and also displayed impairments in structural language. In line with
the previously mentioned discoveries, data from long-term studies provide evidence that
individuals who were previously diagnosed with DLD, as teenagers and adults, express
characteristics that are similar to autism [96,97]. Those findings could imply a possible
relation between the two disorders, which becomes more apparent with children’s growth,
as social expectations grow as well, and their DLD symptoms alone cannot explain their
social deficiencies. Then it seems that their symptoms might become more autistic-like.

1.5.2. Differences
Despite the above-mentioned studies arguing in favor of possible comorbidity of the
two conditions, another theoretical position states that the similarities in the language
and memory performance of children with autism and children with DLD are simply
superficial, and do not suggest a common etiology [98].
The evidence for this position gains support from studies in which children with
DLD and children with autism demonstrate different patterns of deficits between the two
conditions (e.g., [2,13]). For instance, Demouy and colleagues [85] examined receptive and
expressive language in children with ASD, DLD, and Pervasive Developmental Disorder
Not Other Specified (PDD-NOS) and reported that skills in ASD and DLD rely on different
mechanisms, while PDD-NOS showed an intermediate profile sharing some characteristics
of both AD and DLD. The results identified significant differences between three groups in
syntax, pragmatics, and prosody. DLD demonstrated correlations between chronological
age and raw scores in all language tasks, while the two other groups did not. In addition,
DLD showed correlation within all raw scores in language tasks, the majority of correlations
were also found in PDD-NOS but not in the ASD group. The authors suggest that language
skills in ASD and DLD are not comorbid, however, they argue about the existence of some
common symptoms, as they consider PDD-NOS as an intermediate profile sharing some
characteristics of both ASD and DLD.
In another study with adolescents with DLD and ASD-LI, the findings showed the
second groups’ participants to outperform DLDs’ in lexical tasks involving word asso-
ciations and structures [83]. In an additional study with adolescents, Loucas and col-
leagues [99] investigated three groups of ASD-LI, DLD, and TD adolescents, in tasks of
Brain Sci. 2021, 11, 589 9 of 29

spoken word recognition with frequency (low/high) differences and several phonological
onset neighbors. They found that adolescents with ASD-LI needed more speech to identify
low-frequency words with opponents of low density, than the other two groups, which
showed a similar performance, pointing to differences between clinical conditions.
Furthermore, data derived from the field of non-word and sentence repetition tasks
provide important evidence related to the nature of the commonalities observed. Some
studies have been found different patterns of errors related to ASD and DLD children, on
non-word and tasks which require repetition of sentences [13,18]. Specifically, Whitehouse
and colleagues [13] claim that even if some common deficits do exist in both disorders,
there are different underlying causes behind those. They have disclosed different kinds of
non-word repetition mistakes between the two groups, concluding that there is no evidence
supporting that those two disorders may have a common behavioral or etiological result.
However, this result should be viewed with caution, because the number of participants
in the ASD group who performed poorly enough on the non-word repetition measure
and was included in the analysis was small (n = 8). An additional study [99], despite
concluding on the same findings as Whitehouse and colleagues [13], did not find any
compelling differences between the two groups concerning their overall error rates, and
their percentage of phonemic mistakes that specifically affected the syllable structure of
the stimuli, on word-repetition performance. Though, again, the results should be seen
with discretion because ASD-LI children were not matched on their chronological or verbal
mental age (VMA) to DLD children [99].
The assumption that ASD participants demonstrate higher performance than DLD
participants in nonword tasks is furthermore supported by other scientists [17,18]. For
instance, Durrleman and Delage [17] reported that both groups, ASD and DLD, performed
worse than TD on NWR tasks, of the same chronological age. Group differences reveal
the most pronounced impairments relate to the DLD group since the children with ASD
performed superior to the group with DLD. These authors also reported similar deficits
for third person accusative clitic production and general morphosyntax between the two
groups and stated the existence of a subgroup of children with ASD, which showed intact
grammar except for 1st person accusative clitic production, where children with DLD
showed good performance.
A more recent study compared the performance of individuals with ASD with indi-
viduals with Syntactic-Developmental Language Impairment (SyDLI) and a TD group.
their results support that although the two clinical groups show similar performance in
syntactic tasks, they produced different error patterns [100]. Creemers and Schaeffer [101]
showed another demonstration of the differences between the two groups in a Dutch-
speaking population. ASD and DLD participants were compared to a lexical-syntactic task
of mass-count distinction, and a pragmatic task (testing the use of definite markers). The
results showed that the ASD participants perform better than the DLD participants on the
grammatical mass-count task, similar to the TD level. They performed, however, worse
than the DLD group when they were asked to provide a definite determiner, a task that
requires pragmatic abilities [102,103].
In addition, data from pragmatic language and narrative tasks reveal different levels of
performance between the ASD and DLD children. For instance, in a study using pragmatic
language and narrative tasks, different levels of performance were produced between
the ASD and DLD children, and narratives showed that ASD children produced more
expressive mistakes during their story-telling compared to DLD children. Furthermore,
ASD children were found to produce more receptive but not expressive mistakes compared
to DLD children [84]. This finding is also supported by Hudry and colleagues [104] who
found the receptive language abilities are more impaired than the expressive language
abilities, in a group of pre-schoolers with autism. These findings led researchers to hypoth-
esize that there are distinct etiologies explaining these deficits. In a review of the literature,
Williams and colleagues [1] concluded at this second position, arguing that there is not
much evidence supporting a shared etiology explaining their language deficits.
Brain Sci. 2021, 11, 589 10 of 29

Given that the above studies point to a different phenotypical path, researchers are
left to hypothesize that distinct etiologies are explaining these deficiencies. Still, there is
not enough evidence to explain the origin of the similarities presented by the two develop-
mental conditions. Recently, Taylor and Whitehouse [105] reviewed studies investigating
overlaps in the phenotypes of DLD, ASD, and social (pragmatic) communication disorder
(SPCD). They concluded that some children with language impairments will fall between
the boundaries of conventional diagnostic criteria for ASD and DLD and could meet the
criteria for SPCD. The authors argue that it may be the case for DLD, SPCD, and ASD to
be related disorders that vary according to the degree of the presented deficits in different
domains; namely structural language, pragmatic language, and circumscribed interests
(also see [106]).
This possibility of relatedness as it derives from their similarities enhances the question
as to whether DLD and ASD are parts of the same continuum, as Bishop supports [19].
The observed similarities between the two developmental conditions, let Bishop outbid in
favor of a shared etiology of the two. She then argues that DLD and ASD constitute points
on a continuum of the same disorder instead of separate conditions [19].
Considering all the above mentioned, we can conclude that, the commonality picture
is blurred and the available research data provide conflicting results. We can conclude that
there is important room for research, to deeply comprehend the relation between the two
disorders, especially, in the language domain.

1.6. The Present Study

1.6.1. Purpose of the Study
The importance of locating the similarities and differences of DLD and ASD develop-
mental conditions is critical, to clarify the specific symptoms of each disorder and intervene
accordingly. Despite the research efforts which have been made so far, the question is still
open. Putting the findings together, it seems that there is some evidence supporting the
possibility of an overlap between the two conditions in language domain.
Hence, this study aims to further investigate whether the observed similarities be-
tween the two disorders are surface manifestations of the same underlying neuropsy-
chological dysfunction, or whether they derive from distinct pathologies. The relation
between DLD and ASD will be extensively investigated in relation to the critical domain
of language.

1.6.2. Hypotheses
We hypothesize that an ASD-LI subgroup of children will be formed, based on partici-
pants’ performance on the Expressive and Receptive Language Evaluation (EREL) [107],
a measure which represents the Greek version of the CELF instrument and is suitable to
identify children with DLD.
We also hypothesize that ASD (ASD-LI and ASD without Language Impairment;
ALN) and DLD children will score lower than typically developed group (TD) on language
measures. Specifically, we expect that ASD-LI and DLD groups to demonstrate poorer
performance than ALN and TD groups on language tasks, and also the ALN group to score
poorer than the TD group.
We hypothesize, especially, that the “Sentence Repetition Task” (SRT or RS; EREL
instrument) will differentiate significantly the three clinical groups from the TD children,
as it is considering to be an important instrument in the diagnostic process of DLD [2]. SRT
can also provide information about the language profile of children and their language
strengths and weaknesses [36]. Similarly, the Non-Words Repetition task (NWR) along with
the SRT, is expected to accurately locate language impairments, if any, in the experimental
groups, as both measures have received great research interest as psycholinguistic markers
of DLD [27,41,42]. If there is a linguistic overlap in ASD-LI and DLD, then, in comparison to
ALN and TD children, we expect both the ASD-LI and DLD groups to perform significantly
worse on NWR and SRT tasks.
Brain Sci. 2021, 11, 589 11 of 29

We also expect children’s symptoms in each group to reflect on the primary deficits
of each disorder. Especially, DLD children are expected to present difficulties primarily
on their general communication ability, while ASD children will demonstrate additional
impairments in the social interaction ability, based on the primary deficits of each dis-
order. These abilities will be tested by the Children Communication Checklist Second
Edition [108].

2. Materials and Methods

2.1. Participants’ Recruitment
A total number of 103 children participated in the study (79 males, 24 females). Partici-
pants were 40 children (38 males, 2 females) referred by special educators and speech pathol-
ogists as ASD children (initial number tested 41 children, one child excluded), 24 children
as DLD children (14 males, 10 females), and 39 TD children (27 males, 12 females) were
recruited from the general population referred as a control group, aged 6–12 years old.
ASD and DLD participants were all recruited from private practices in Cyprus, located
in different parts of the main cities in Cyprus to represent as much as possible different
demographics in our study. The speech pathologists and special educators of children
informed their parents if interested in attending the study. Specialists were specifically
informed about the child characteristics which fitted this study’s profile. They were,
specifically, instructed to request the parents’ availability for participation if their children
were already diagnosed with ASD or DLD, or if the child’s characteristics were resembling
such a profile. If parents were interested the researcher contacted them providing them
with further information about the study. In the ASD referred group, an ASD diagnosis was
needed, from a licensed psychologist or psychiatrist, trained in the clinical administration
of the Autism Diagnostic Observation Schedule, Second Edition [109].
The children in the two clinical groups were recipients of special education services.
From the initial DLD group, all of the participants were receiving speech therapy. From the
ASD group, the majority of children (n = 32) were involved in speech therapy, to address
different deficits in language, related to different components of language, such as in
structural and pragmatics domains.
In addition, a comparison group of children from the general population was formed.
These children were selected randomly, from children aged 6–12 years, attending public
schools in Cyprus.
For all the participants’, their relatives (mothers, n = 103) were also included in the
study. Permission from the Cyprus National Bioethics Committee and parental consent for
participation in the study were obtained before testing. Children’s verbal agreement was
also obtained before testing.

2.1.1. Demographic Characteristics

Participants are all Greek native, from mixed socio-economic backgrounds.

2.1.2. Group Selection

DLD: The children met the inclusion criteria for Developmental Language Disorder
described in the Diagnostic and Statistical Manual of Mental Disorder (5th edition) [5].
Specifically, participants in the DLD group were able to participate, if they fitted on DLD
diagnostic criteria of the study when administered the EREL measure. Participants who
scored 1.25 standard deviation below the mean, were suitable for inclusion in this group.
The children’s cognitive ability was also assessed, to reinsure that no participant fell in
an intellectual disability score (75 or above), based on their performance on the Wechsler
Abbreviated Scale of Intelligence First Edition (WASI) [110]. Two participants had a score
of 71, however, given that the IQ score obtained was general and not only their Nonverbal
ability, but we also reached the maximum criterion of inclusion in the study.
Our decision to include participants with IQ scores around 75 is supported by several
studies in which scores below and including 75 were allowed in language impairment
Brain Sci. 2021, 11, 589 12 of 29

samples [111]. In addition, it remains an unresolved question whether would have been
more effective to expand the lower level of IQ score as low as 70 or below, thereby intro-
ducing greater variability within the group of children with language impairments. One
statement is that there are no interesting language differences between groups defined
according to the conventional criterion (Nonverbal IQ of 85 and above) and the expanded
criterion (Nonverbal IQ 70 and above) [112]. Importantly, under the application of DLD
criteria, children with nonverbal IQ scores below-average are not excluded, unless those
scores are so low as to merit the diagnosis of intellectual disability [113]. Nevertheless, in
this study, the threshold of 75 score which was set, was assessed for the general IQ ability
of the participants, and consequently, we can assume that for this deficient group, their
Nonverbal ability will be greater.
ASD: Children have to meet the inclusion criteria for Autism Spectrum Disorder
described in the Diagnostic and Statistical Manual of Mental Disorder [5]. Participants in
the ASD group were able to attend the study if they fitted an ASD diagnosis (A formal
ASD diagnosis by licensed psychologists or psychiatrists, trained in the clinical use of the
instrument ADOS-2). ASD children’s cognitive ability was within the normal range (75 or
above), based on the WASI [110].

2.1.3. Exclusion Criteria

Exclusion criteria were IQ scores under the average and any other history of psy-
chiatric or medical conditions. Additionally, children with other neurologic impairment,
global developmental delay, or significant hearing impairment were excluded from the
sample. Children whose parents are unwilling to complete the caregiver’s measures were
also excluded from the study. Based on the above criteria, one child was excluded from the
research, because it was not able to complete the WASI test.
The following section describes the different instruments and measures administered.
All of the administrations were performed by the researcher, who has the qualifications to
assess and provide diagnosis, as a licensed school psychologist. The overall examination
had a duration of approximately 2.5 h, with a 10 min break in between. The mothers of the
participants were simultaneously completing the measure administered to them unless
they needed some additional help or clarifications from the researcher. In the second case,
after the completion of the child’s participation, the researcher provided the appropriate
guidance and help to the mothers to complete their measures.

2.2. Measures
2.2.1. Pre-Test Examination Measure
Wechsler Abbreviated Scale of Intelligence First Edition (WASI) [110]. This scale is
used to measure participants’ IQ and to ensure that no child presents intellectual disability
among participants. The two-subtest short version will be utilized (vocabulary and matrix
reasoning), and it is administered to provide a comprehensive picture of the participants’
general cognitive ability.

2.2.2. Assessment Measures

Expressive and Receptive Language Evaluation [107] (EREL; Greek standardized test
similar to Clinical Evaluation of Language Fundamentals, Fourth Edition, CELF-4 [114].
This test is a suitable instrument to assess language ability and it is administered to
individuals from 5 to 12 years old. In this study, participants will be assessed with this
instrument to detect their linguistic abilities. Individuals need to respond verbally to
stimuli presented in pictures. The Core Language score for the respective CELF-4 presents
high sensitivity of 0.86 and specificity of 0.96 [114]. The reliability index of the EREL was
ranging from a = 0.78 to a= 0.96 on the Core Language score for the age groups represented
in our study [107]. The Core Language subtests of the EREL were administered and are
referring to:
 (a)sensitivity of 0.86 and
interpret spoken specificity
directions of 0.96 [114].
of increasing Theand
length reliabil
com
ing of
order from a = 0.78
mention of to a= 0.96shapes
familiar on the with
Corevarying
Language score fo
characteris
our study
cation; and (c)[107]. The Core
identify from Language
among several subtests of the
choices EREL
the pic
Brain Sci. 2021, 11, 589 ferring
tioned. to: abilities reflect short-term and procedural mem
These 13 of 29

An(i)example
Conceptofand suchFollowing
a task is:Directions,
(instructions:a subtest
“showthat the es
(a) go”).
Let’s interpret spoken directions of increasing length and c
(i) Concept and Following order of mention
Directions, a subtestof familiar
that examinesshapes with ability
participants’ varyingto (a)characte
interpret spoken directions ofcation;
increasing length and complexity; (b) follow the stated
and (c) identify from among several choices the p order
of mention of familiar shapes with varying characteristics such as color, size, or location;
tioned. These abilities reflect short-term and procedural m
and (c) identify from among several choices the pictured objects that were mentioned.
These abilities reflect short-term andAnprocedural
examplememory
of suchcapacities.
a task is: (instructions: “show th
Let’s go”).
An example of such a task is: (instructions: “show the sock and then show the bird.
Let’s go”).

(ii) Word Structure, which aims to evaluate the participa

structure rules (morphology) to mark inflections, derivatio
select and use appropriate pronouns to refer to people, obje
ships.aims to evaluate the participants’ ability to (a) apply word
(ii) Word Structure, which
structure rules (morphology) to Anmark inflections,
example derivations,
follows where and
the comparison;
administerand (b) aloud
reads
select and use appropriate pronouns to refer to people, objects, and possessive relationships.
ticipant(ii) Word
only see Structure,
the picturewhich aims to
presented evaluate
below. the partica
Instructions
An example follows where the administer reads aloud the instructions and the partici-
pant only see the picture on structure
the left,
presented rules
“Here
below. (morphology)
we havearea given
Instructions to mark
fox”, first
andfor inflections,
then we show
the picture on deriva
the pi
haveselect and
two….” use
(and appropriate
we expect pronouns
the child toto refer
say to
“foxes”).
the left, “Here we have a fox”, and then we show the picture on the right. “Here we have people, o
ships.
two . . . .” (and we expect the child to say “foxes”).
An example follows where the administer reads alou
ticipant only see the picture presented below. Instruction
on the left, “Here we have a fox”, and then we show the
have two….” (and we expect the child to say “foxes”).

(iii) Formulated Sentences,(iii) Formulated

to investigate Sentences,
the ability to investigate
to formulate the ability to f
complete, semanti-
cally, and grammatically cally,
correct,and
spokengrammatically correct, length
sentences of increasing spoken and sentences
complexityof incre
(i.e., simple, compound, (i.e., simple, compound,
and complex andgiven
sentences), using complex
wordssentences),
(e.g., car, if,using
be- given
cause) and contextual constraints imposed by illustrations. These abilities
and contextual constraints imposed by illustrations. These areflect the
capacity to integrate semantic, syntactic, and pragmatic rules and constraints while using
working memory.
integrate semantic, syntactic, and pragmatic rules and cons
Brain Sci. 2021, 11, 589
memory.
In the following example, (iii) Formulated
participants are askedSentences, to investigate
to create a sentence by giventhe theability t
word “book” (“βιβλίo” in the cally,
In the
Greek and grammatically
following
language). example, correct, spoken sentences
participants are asked of to inc
cr
(i.e.,“book”
word simple,(“βιβλίο”
compound, andGreek
in the complex sentences), using gi
language).
and contextual constraints imposed by illustrations. Thes
integrate semantic, syntactic, and pragmatic rules and co
memory.
In the following example, participants are asked to
word “book” (“βιβλίο” in the Greek language).

(iv) Recalling Sentences,(iv)

is a Recalling Sentences,
subtest by which is a subtest
the participant by which
is asked to listenthe
to partici
spoken sentences of increasing length and complexity, and repeat the sentences withoutand rep
spoken sentences of increasing length and complexity,
changing word meaningchanging word
and content, wordmeaning
structure and content, word
(morphology), structure
or sentence (morpholo
structure
(syntax). Semantic, morphological, and syntactic competence is examined, as it facilitates
(syntax). Semantic, morphological, and syntactic competence is
immediate recall (short-term memory).
immediate recall (short-term memory).
Some examples of sentences to recall are:
Some examples of sentences to recall are:
“Ζήτησε να έρθει μαζί μας γεμάτος χαρά” (He aske
joy).
“H κατάσταση ήταν πραγματικά πολύ άσχημη για ό
Brain Sci. 2021, 11, 589 14 of 29

“Zήτησε να έρθει µαζί µας γεµάτoς χαρά” (He asked to come with us full of joy).
“H κατάσταση ήταν πραγµατικά πoλύ άσχηµη για óλoυς” (The situation was very
bad for everyone).
“Aπoϕάσισαν óτι τo χωράϕι θα πoυληθεί και τα λεϕτά θα µoιραστoύν στα τρία
αδέλϕια και τις δύo αδελϕές” (They decided that the field will be sold and the money
will be divided to the three brothers and the two sisters), and
(v) Word Classes, aiming to evaluate the participants’ ability to analyze words for
their meaning features, define words by referring to class relationships and shared mean-
ings, and describe meanings that are unique to the reference or instance. In this task, at
first, the participant has to choose two of four words provided, that he feels match each
other. This subtest was completed only by children 9 years or older, instead of the Word
Classes subtest.
For instance, the participant is given the words “fence”, “window”, “glass”, “carpet”
and is asked to choose the two that better fit each other. Then, participants are asked, for
instance, “how the words window and glass are related”?
Non-Word (NWR) task [115]. NWR task is considered an important clinical marker
of DLD [27,41], and it was used in this study to detect potential language deficits of
participants. The NWR is consisted of 32 non-words. The nonwords are divided into
four categories of two, three, four, and five syllable pairs each. Non-words were selected
from a list by a group of 58 undergraduate psychology students who were asked to judge
the wordlikeness of non-words on a scale ranging from 1 (very unlikely to be rated as
a real Greek word) to 5 (very likely to be rated as real Greek word) [115]. Instead of
single nonwords, pairs were adopted to avoid ceiling effect, especially in older children.
Instructions and test stimuli were read out loud by the experimenter at a normal rate. The
children were instructed to listen carefully to the stimuli and repeat the pairs of nonwords
as accurately as they could. Repetition responses were scored either as correct for each
nonword pair (score 1) or incorrect (score 0). The maximum correct score that could be
achieved was 16. The Cronbach’s αs of the scale ranged from α = 0.76 to α = 0.81, for
different grades. An example of Nonwords of three syllables in Greek is “Bλυχηθµóς-
Iσκάτης”, and an example of four syllables nonwords is “Λαθηκιδής- Mελίκρατo”.
Children’s Communication Checklist-2 [108]. Is a 70-item questionnaire completed by
a caregiver or school teacher and is suitable to detect communication difficulties in children
from 4 to 16 years old. The caregiver/school teacher rates how frequently several types
of behavior are observed. The response scale is treated as a 4-point Likert scale; less than
once a week (or never). At least once a week, but not every day (or occasionally), Once or
twice a day (or frequently), and Several times (more than twice) a day (or always).
It is a very useful tool for screening for language impairment, pragmatic impairment
and is an indicator for further investigation of autism disorder. It consists of 10 different
subscales (speech, syntax, semantics, coherence, inappropriate initiation, scripted language,
use of context, nonverbal communication, social relations, and interests). The test manual
reports internal consistency and inter-rater agreement for all scales. Internal consistency
was 0.65 or more for all scales.
The test was not available in the Greek language and for this reason, it was translated
back and forth from English to Greek by two native language speakers of each language
before used. Checklist translated version agrees with the translation given from a thesis
titled “The Adaptation of Children’s Communication Checklist [108] written in Greek
language and implemented in parents of 4–7 years old children” [116].
The two composite scores of CCC-2 were used, namely General Communication
Composite (GCC) and Social-Interaction Deviance Composite (SIDC). GCC is designed to
discriminate between children with communication impairments and typically developing
children and SIDC is derived to give optimal discrimination between children with typical
DLD and those with pragmatic difficulties that are disproportionate to their structural
language abilities. Usually, children in the autism spectrum score high in GCC (55 or
below), while they present deficits in the SIDC (scores 8 or below). Children with DLD also
Brain Sci. 2021, 11, 589 15 of 29

typically present impairments on GCC (score of 55 or below), however, they score higher
on the SIDC (9 or above) [117,118]. This proposed categorical division of symptoms was
decided that fits this study’s hypothesis and selected for use.

2.3. Statistical Analysis

2.3.1. Preliminary Analysis
Before comparing data between groups, univariate normality was tested for the four
groups’ performance on each task and scale. Participants’ performances were normally
distributed. Specifically, the histogram of standardized residuals indicated that the data
contained approximately normally distributed errors, as did the normal Q–Q plot of
standardized residuals, which showed points that were generally on the line. By visually
checked the Q–Q plots for the model, the assumption of normality of the residuals is
met. However, there were some minus individual outliers detected but they produced
only small deviations of the normality. For this reason, outliers are not excluded from the
analysis. Multivariate normality was also checked and heteroscedasticity was acceptable
in all cases.
Furthermore, Cronbach’s alpha was calculated to be informed about the internal
consistency of the scales used. The CCC-2 subscale consisted of 70 items, α = 0.884, which
indicates a high level of internal consistency of the scale with this specific sample.
We also tested to see if the data meet the assumption of collinearity indicated that
multicollinearity was not a concern as the VIF’s of the scales was under 10 (Field, 2013).
IQ-Total, VIF = 1.54; CCC-2 (speech) VIF = 4.58; CCC-2 (syntax) VIF = 5.3; CCC-2 (semantic)
VIF = 3.84; CCC-2 (coherence) VIF = 4.76; CCC-2 (Inappropriate initiation) VIF = 4.61; CCC-
2 (stereotyped) VIF = 4.87; CCC-2 (use of context) VIF = 4.60; CCC-2 (social) VIF = 2.62;
CCC-2 (nonverbal) VIF = 2.63; CCC-2 (interests) VIF = 3.99); Multicollinearity was not also
a concern for Words-Non Words Total (VIF = 8.5).
Next, multiple comparisons were contacted to detect highly correlated measures and
items. It appeared that some factors of the CCC-2 were highly correlated and therefore it
was decided that the two basic subcategories of CCC-2 will be used [117]. Specifically, the
General Communication Composite (GCC; VIF = 1.05) of CCC-2, designed to discriminate
between children with communication impairments and typically developing children and
also the Social-Interaction Deviance Composite (SIDC; VIF = 1.05), derived to give optimal
discrimination between children with typical DLD and those with pragmatic difficulties
that are disproportionate to their structural language abilities, were used. This proposed
categorical division of symptoms was decided that fits this study’s hypothesis and selected
for use. A full correlation matrix among measures is provided (Table 1).

Table 1. Correlations among measures.

Variable CFD WS RS FS WC NWR SIDC GCC

CFD —
WS 0.632 ** —
RS 0.712 ** 0.698 ** —
FS 0.806 ** 0.741 ** 0.891 ** —
WC 0.328 — 0.184 0.425 —
NWR 0.679 ** 0.674 ** 0.662 ** 0.694 ** 0.393 —
SIDC −0.101 0.026 −0.149 −0.055 −0.195 −0.177 —
GCC 0.172 0.208 0.342 ** 0.123 0.139 0.206 −0.324 —
** Significant at p = 0.001. Note: CFD = Concept and Following Directions; WS = Word Structure; RS = Recalling
Sentences; FS = Formulated Sentences; WC = Word Classes EREL measures; NWR = Non-Word Repetition
Task; SIDC = Social-Interaction Deviance Composite, Children’s Communication Checklist-2; GCC = General
Communication Composite, Children’s Communication Checklist-2.

2.3.2. Power Analysis

An a priori power analysis was conducted using G*Power3 [119] to test the difference
between three independent group means using analysis of covariance, a large effect size
Brain Sci. 2021, 11, 589 16 of 29

(d = 0.40), and an alpha of 0.05. The result showed that a total sample of approximately
n = 110 is acceptable for the study, to achieve a power of 0.80. Our total sample consists of
103 children, who participated in this study.

2.3.3. Main Analysis

Furthermore, to study whether the relation between our participants is explained
by group differences, or if there are differences within the groups of participants, we
decided to proceed to a linear mixed model (LMM) analysis [120]. LMM technique pro-
vides the researcher with the opportunity to simultaneously study both within person
(intra-individual) systematic change (level 1) and also between-person (inter-individual)
differences (level 2) of the participants in different measures [121].
Undoubtedly, participants’ performance differs significantly and heterogeneity is
observed among their responses on several tasks. Via this technique we can examine the
total between- and within-person variance in the dependent variables. We adopted this
analytic strategy to provide answers to the amount of outcome variation located in intra-
and inter-individual levels. The intra-class correlation coefficient (ICC) provides the oppor-
tunity to examine to what extent the total outcome variation is related to inter-individual
differences. Technically speaking, ICC refers to the proportion of the total variation of the
dependent variable that can be attributed to between-person differences. The ratio of the
between-cluster variance to the total variance is what is called the Intraclass Correlation
and provides information of the proportion of the total variation that is accounted by
the clustering.
A high ICC (>0.10) implies that the between-group variability dominates the within-
group variance, meaning that, most of the differences that we observe across individuals on
a variable are stemming from group differences [122]. On the contrary, a low ICC indicates
that the variation seen in a variable is the result of individual differences within groups.
Participants IQ total score and age (in months) were treated as covariates, as one-way
ANOVA revealed significant group differences of the two variables. Specifically, there
appears to be a statistically significant effect of group on the dependent variables of IQ
total variable, F (3,99) = 6.3, p < 0.001 (between groups DLD and TD children, and ASD-LI
and TD children. Age variable was significantly differed as well [F (3,99) = 11.21, p < 0.001].
ASD group differed from ASD-LI group, DLD group form the group ASD-LI and ASD-LI
from TD children’s group. The means and standard deviation scores are presented in
Table 2. Sex was not treated as a covariate as there was no effect of sex or interaction
between group and sex. In addition, the educational level of the parents (mothers, fathers)
of the participants was also examined and did not produced significant differences between
groups, p = 0.962 and p = 0.298, respectively. Similarly, the socioeconomic status of the
family, was also not significantly different among groups, p = 0.975. Descriptive statistics
of participants’ characteristics are presented in Table 3.

Table 2. Descriptive statistics of group formation.

ASD (n = 16) ASD-LI (n = 24) DLD (n = 28) TD (n = 35)

Mean SD Range Mean SD Range Mean SD Range Mean SD Range
Age 108.81 23.33 73–140 78.29 6.85 72–99 97.21 19.98 72–193 105.71 23.79 72–144
IQ Total 92.38 10.56 85–126 86.88 5.36 80–103 90.18 9.54 71–111 96.43 8.8 76–113
CFD 37.69 7.41 24–51 17.04 5.44 8–29 24.79 9.89 6–48 40.37 8.33 25–52
WS 14.57 5.16 9–22 10.46 4.25 4–21 12.2 6.09 0–27 18.06 3.4 14–23
FS 22.31 11.66 4–44 9 6.49 0–24 16.61 8.74 0–32 29.51 10.71 9–49
RS 39.13 17.47 3–68 18.33 10.31 0–43 30.82 16.08 0–56 45.89 12.43 15–69
WC 18.33 9.06 4–36 - - 12.75 5.47 8–25 23.16 6.32 8–35
Note: CFD = Concept and Following Directions; WS = Word Structure; RS = Recalling Sentences; FS = Formulated Sentences; WC = Word Classes.
Brain Sci. 2021, 11, 589 17 of 29

Table 3. Participants’ Characteristics.

ASD (n = 16) ASD-LI (n = 24, girls = 1) DLD (n = 28, girls = 10) TD (n = 35, girls = 11)
Mean SD Range Mean SD Range Mean SD Range Mean SD Range
Age * 108.81 23.33 73–140 78.29 6.85 72–99 97.21 19.98 72–193 105.71 23.79 72–144
IQ-Total 92.38 10.56 85–126 86.88 5.36 80–103 90.18 9.54 71–111 96.43 8.8 76–113
SES 1.56 0.51 1–2 1.58 0.50 1–2 1.61 0.50 1–2 1.60 0.50 1–2
EducM 1.75 0.58 1–3 1.79 0.78 1–3 1.68 0.72 1–3 1.71 0.72 1–3
EducF 1.63 0.72 1–3 1.80 0.88 1–4 1.67 0.68 1–3 1.44 0.56 1–3
Note: SES = Socio-Economic Status, EducM = Mothers’ Educational Level, EducF = Fathers’ Educational Level. * Age is presented
in months.

Our concern was that any group of differences may not be due to the group factor,
but instead may simply reflect differences in total IQ score and/or age. To address this
concern, we performed two LMM analyses for each of the dependent variables (language
measures and the composites scores of the CCC-2). In all models A we included measures
of total IQ and age as covariates, so that any fixed effect of group then reflects differences
due to group membership above and beyond any individual differences in IQ score or age.
In all models B we included only the factor of the group. Additionally, in all analyses we
considered subjects as random effects. All models A showed that total IQ score and age
are not significant predictors of the slope. In other words, the effect of factor group on the
dependent variable was constant across the values of total IQ and age.

3. Results
In this study, the measures used to answer the research questions were the EREL, the
Non-Words Repetition task (NWR), and also the Children’s Communication Checklist-
Second edition (CCC-2). The Non-Words Repetition task (NWR) data was treated as
raw scores because we wanted to include in our methodological design the factor of age.
However, we were not able to manipulate all data in this way, since there are measures that
provide their diagnostic algorithms, and transformation to z scores is already constructed
by the manual provided, as in the case of WASI, and CCC-2. For these instruments, the
manual’s diagnostic instructions were followed.
The administration of EREL language assessment revealed a language-impaired sub-
group of children within the ASD group. The ASD group was then sub-divided upon
participants’ performance on the EREL instrument. A group of children with language im-
pairment (ASD-LI), and a group of children with normal language (ALN) were identified;
ASD participants whose scores were −1.25 SD or more below the normative mean on two
out of four Core Language scores [20] formed the ASD-LI group. The means and standard
deviations of groups’ performance on EREL are provided in Table 2.
These results transformed the experimental groups and an additional group emerged
by the analysis. Especially, the final groups were formed as follows: 16 children consisted
the ASD group (without additional language impairment) (16 males, 0 females), 28 children
the DLD group (17 males, 11 females), 24 children consisted the ASD-LI group (22 males,
2 females), and 35 children (24 males, 11 females) the TD group. Four children from
the general population were also found to meet the DLD diagnostic criteria. The group
descriptions that emerged by the language ability analysis are presented in Table 2.
In addition, the multilevel model was preferred for the analysis, as we assumed the
existence of heterogeneity between individuals. The results are presented in Table 4. In
addition, boxplots upon all the language measures, support this assumption (see Figure 3).
Specifically, boxplots indicate important differences between the different groups and, in
addition, provide a significant picture of important individuals differences in each group.
The results are further supported by the model ICC.
 by the manual provided, as in the case of WASI, and CCC-2. For these instruments, the
manual’s diagnostic instructions were followed.
The administration of EREL language assessment revealed a language-impaired sub-
group of children within the ASD group. The ASD group was then sub-divided upon
Brain Sci. 2021, 11, 589 participants’ performance on the EREL instrument. A group of children with language 18 of 29
impairment (ASD-LI), and a group of children with normal language (ALN) were identi-
fied; ASD participants whose scores were −1.25 SD or more below the normative mean on
two out of four Core Language scores [20] formed the ASD-LI group. The means and
Table 4. Intra-class correlation and confidence intervals (CI) for Language Measures and
standard deviations of groups’ performance on EREL are provided in Table 2.
CCC-2 Composites.
These results transformed the experimental groups and an additional group emerged
by the analysis.
Measure Especially, the final groups ICCwere formed as follows: 16 childrenCI consisted
the ASD group (without additional language impairment) (16 males, 0 females), 28 chil-
Language Domain
dren the DLD NWR group (17 males, 11 females), 0.36 24 children consisted the ASD-LI group (22
1.827–9.694
males, 2 females),
CFD and 35 children (24 males, 11
0.57 females) the TD group. Four children from
17.883–42.053
the general WS population were also found to meet the DLD diagnostic
0.24 criteria. The group
10.053–17.516
descriptionsRS that emerged by the language 0.34
ability analysis are presented20.438–46.656
in Table 2.
FS 0.37 9.746–28.988
In addition, the multilevel model was preferred for the analysis, as we assumed the
WC 0.23 11.253–25.089
existence of heterogeneity between individuals. The results are presented in Table 4. In
GCC 0.27 33.032–67.143
addition, boxplots
SIDC upon all the language 0.11 measures, support this assumption
2.585–13.449(see Figure
Note: ICC = Intra-Class Correlation Coefficient, CI = 95% Confidence Interval. NWR different
3). Specifically, boxplots indicate important differences between the = Non-Word groups and,
Repetition
in CFD
task; addition, provide
= Concept a significant
and Following Directions;picture of Structure;
WS = Word important RS =individuals differences
Recalling Sentences; in each
FS = Formulated
Sentences,
group. WCThe=results
Word Classes; GCC = General
are further supportedCommunication Composite
by the model ICC.Children’s Communication Checklist-
2; SIDC = Social-Interaction Deviance Composite Children’s Communication Checklist-2.

Brain Sci. 2021, 11, 589 19 of 30

Figure 3. Boxplots for all language measures. CFD = Concept and following directions; WS = Word structure; RS = Recalling
Figure 3. Boxplots for all language measures. CFD = Concept and following directions; WS = Word structure; RS = Re-
sentences; FS = Formulated
calling sentences; sentences,sentences,
FS = Formulated WC = WordWCclasses;
= WordNWR = Non-word
classes; repetitionrepetition
NWR = Non-word task. task.

The model ICC was also significant for all the language measures (see Table 3), as the
Table 4. Intra-class correlation and confidence intervals (CI) for Language Measures and CCC-2
ICC percentages ranged between 23% and 57% on different language tasks, supporting that
Composites.

Measure ICC CI
Language Domain
NWR 0.36 1.827–9.694
CFD 0.57 17.883–42.053
Brain Sci. 2021, 11, 589 19 of 29

an important amount of the total variance, can be explained by inter-individual differences.

Important differences have been noted between groups at all language measures as well.
For the Concept and Following Directions (CFD) subscale of EREL, results revealed
important group differences, p < 0.001. Specifically, the ALN group scored significantly
higher (p < 0.01) from ASD-LI, p < 0.001, and DLD children, p < 0.001. Interestingly, there
were not any important differences between ALN and TD groups of children (p = 0.33). No
important differences were also observed between participants in the ASD-LI and DLD
groups p = 0.90, while ASD-LI and also DLD groups, scored significantly lower than the
TD groups, p < 0.001. The ICC model for CFD measure was large (57%), further supporting
the existence of important inter-individual differences among the participants.
Similarly, important group differences were found for the Word Structure (WS) sub-
scale of EREL, p < 0.001. Important differences were found between the ASD-LI group,
which scored significantly lower (p < 0.01) than the TD group, and also the DLD group
scored significantly lower than the TD group, p < 0.001. No important differences were
found among ALN and TD groups (p = 0.07), DLD group (p = 0.89), nor ASD-LI group
(p = 1.0). Similarly, no important differences were observed between ASD-LI and DLD par-
ticipants (p = 0.78). In addition, the ICC model for the WS measure was high (24%), further
supporting the existence of important inter-individual differences among the participants.
As it concerns the Recalling Sentences (RS or SRT) subscale, important group differ-
ences were also found, p < 0.001. As in the case of the WS subscale, important differences
were located in the groups of ASD-LI group which scored significantly lower (p < 0.001)
than the TD group, and also DLD group scored significantly lower than the TD group,
p = 0.02. No important differences were found among ALN and TD groups (p = 0.50),
DLD group (p = 0.66), nor ASD-LI group (p = 0.07). Similarly, no important differences
were observed between ASD-LI and DLD participants (p = 0.29). Even not all significantly
important, differences revealed for all groups, boxplots presented, indicate for different
performances between groups.
This finding is moreover sustained by the ICC model, which declares a high percentage
for RS measure (34%) and outbids the existence of important inter-individual differences
in groups. Formulating Sentences (FS) EREL measure has also provided important group
differences on language domain, p < 0.001. Specifically, important differences were found
between the ASD-LI group which scored significantly lower (p < 0.01) than the TD group,
and also the DLD group (M = 16.61, SD = 8.74) scored significantly lower than the TD
group, p < 0.001. Important differences were also found between the ALN and TD groups,
p < 0.01., DLD group (p = 0.66), nor ASD-LI group (p = 0.07). However, no important
differences were observed between ASD-LI and DLD participants (p = 0.99), ALN and
DLD groups (p = 0.97), nor ASD-LI and ALN participants (p = 0.98). The ICC model for
the subscale of FS was large enough (37%), further supporting the existence of important
inter-individual differences among the participants.
The last subscale of EREL measure, Word Classes (WC) also showed important dif-
ferences between groups, p < 0.01. Important differences came up between DLD children,
which scored significantly lower (p < 0.01) than TD children. No important differences
emerged between ALN and DLD, p = 0.34, or TD group, p = 0.19. Importantly, not an
ASD-LI group was formed in this subscale of EREL. WC is a subscale administered only in
children between the ages of 9 and 12 years old. None of our participants were found to
belong in the ASD-LI group in this age range. Our sample consists of 36 children (from 103)
belonging to that age range, of which 19 children are TD participants. The variability of
participant’s responses in each group is demonstrated, and it is further supported by a high
ICC model for this variable (23%) and indicates important inter-individual differences.
Important differences between groups on language domain are also found on the
NWR task, p < 0.001. Specifically, differences are observed in ASD-LI children (M = 2.16,
SD = 2.73), who scored significantly worse than TD children (M = 10.2, SD = 4.71), p < 0.001.
Likewise, DLD children (M = 3.79, SD = 3.45), also performed worse than TD children on
the same task, p < 0.001, and controversially, no significantly different from the ASD-LI
Brain Sci. 2021, 11, 589 20 of 29

group, p = 0.99. No differences were also found in the ALN group (M = 6.88, SD = 4.54),
related to the ASD-LI (p = 0.53), DLD (p = 0.30), and TD (p = 0.05) group, respectively.
The ICC model for NWR is also high (36%), informing once again for inter-individual
differences among participants.
As it concerns the two composite scores of CCC-2 important findings emerged. GCC
subscale (see Figure 4) significantly differentiated between groups, p < 0.001. ALN group
(M = 37.5, SD = 30.64), the score was significantly lower than TD group (M = 72.06,
SD = 17.60), p < 0.001, indicating increased deficits in the communication domain. In
addition, the ASD-LI group (M = 41.91, SD = 15.74), also obtained significantly elevated
communication deficits in comparison to the TD group of children, p < 0.001. Finally, impor-
Brain Sci. 2021, 11, 589 tant differences were found for DLD children (M = 48.41, SD = 21.8), who demonstrate21 more
of 30
impairments in communication in comparison to TD children, p < 0.001. No important
differences are observed among participants on ASD-LI and DLD groups (p = 0.16), who
scored similarly on this subscale. No important differences are also found between ALN
and ASD-LI (p = 0.87), or ALN and DLD (p = 0.67) participants.

Figure 4. Boxplots for Children’s Communication Checklist-2 composites. GCC = General Com-
Figure 4. Boxplots
munication for Children’s
Composite Children’sCommunication
CommunicationChecklist-2
Checklist-2;composites. GCC = GeneralDeviance
SIDC = Social-Interaction Com-
munication Composite Children’s Communication
Composite Children’s Communication Checklist-2. Checklist-2; SIDC = Social-Interaction Deviance
Composite Children’s Communication Checklist-2.
SIDC subscale of CCC-2 (see Figure 4), also produces important differences among
groups,SIDC subscale
p < 0.01. ALLof CCC-2(M
children (see Figure
= 1.47, SD 4), also produces
= 11.58) important
differ significantly differences
from DLD (M =among 13.44,
groups,
SD p < p0.01.
= 12.06), ALL
< 0.01, andchildren
also TD (M(M == 1.47, SD == 9.61)
9.97, SD 11.58) differ significantly
children, from DLDgroup
p = 0.03. No important (M =
13.44, SD =were
differences 12.06), p < between
found 0.01, andASD-LI
also TD (M(M = 9.97,
= 6.73, SD SD = 9.61)
= 7.96) andchildren,
ALN, DLD, p =or0.03. No im-
TD groups.
portant
DLD group
group alsodifferences were found
is not importantly differentbetween
from theASD-LI (M =(p6.73,
TD group SD = 7.96) and ALN,
= 0.75).
DLD,Related
or TD to groups. DLD group also is not importantly different
the two composites of CCC-2, the model ICC was significant from thefor
TDthegroup (p =
variable
0.75).
GCC (27%), indicating that the General Communication Composite contributes to the group
Relatedwhile
differences, to the
thetwo composites
SIDC’s of CCC-2,
contribution is alsothe model ICC
important, evenwas significant
if not that largefor(11%).
the vari-
In
able GCC
sum, the ICC(27%), indicating
model showedthat thatthe General
there Communication
are important Composite
between-group contributes
differences, to the
explain a
group differences,
significant proportionwhile theobserved
of the SIDC’s contribution
variance (Tableis also
3). important, even if not that large
(11%). In sum, the ICC model showed that there are important between-group differences,
4. Discussion
explain a significant proportion of the observed variance (Table 3).
The results from this study are important and thought-provoking. We had investigated
a4.potential
Discussion relationship between ASD and DLD, and our findings are indicative of its
existence.
The results from we
Specifically, thishad expected
study that an ASD-LI
are important subgroup would have
and thought-provoking. Wederived from
had investi-
the general ASD group, indicating for language impairments within
gated a potential relationship between ASD and DLD, and our findings are indicative of the ASD population,
resembling
its existence.a language profile
Specifically, we similar to DLDthat
had expected children. We had
an ASD-LI also hypothesized
subgroup would have that ASD
derived
(ASD-LI and ALN) and DLD groups of children would have scored
from the general ASD group, indicating for language impairments within the ASD popu- lower than the TD
group
lation,on languageameasures,
resembling language and alsosimilar
profile that ASD-LI
to DLDand DLD groups
children. We had would demonstrate
also hypothesized
poorer performance than ALN and TD groups on EREL measures.
that ASD (ASD-LI and ALN) and DLD groups of children would have scored lower In addition, we have
than
predicted
the TD group on language measures, and also that ASD-LI and DLD groupsWe
that the ALN group would have scored poorer than the TD group. also
would
hypothesized
demonstrate poorerthat both the ASD-LIthan
performance and ALNDLD andgroups
TD would
groupshave performed
on EREL significantly
measures. In addi-
worse on NWR and SRT tasks.
tion, we have predicted that the ALN group would have scored poorer than the TD group.
We also hypothesized that both the ASD-LI and DLD groups would have performed sig-
nificantly worse on NWR and SRT tasks.
Most of these hypotheses have been met. Specifically, as it concerns language tasks,
a common profile of ASD and DLD is supported by our results, indicating the presence of
Brain Sci. 2021, 11, 589 21 of 29

Most of these hypotheses have been met. Specifically, as it concerns language tasks,
a common profile of ASD and DLD is supported by our results, indicating the presence
of a language phenotype in ASD similar to that of DLD. Additionally, ASD-LI and DLD
groups scored worse than ALN and TD groups on language measures. However, despite
our projections, ALN and TD groups revealed no important differences in language tasks.
Controversially, ASD-LI and DLD groups produced significantly more deficits than TD
groups. Concerning NWR and SRT tasks, our findings on both tasks showed that ASD-LI so
as in the DLD group, produced more deficits than the TD group. Nonetheless, despite our
expectations, there were not found any important differences between ALN and ASD-LI or
DLD groups, or between ALN and TD groups.
Our results are strongly supported by previous studies. Specifically, there is a common
profile of ASD and DLD, which indicates a possible common feature between the two
developmental disorders, namely, ASD-LI. These findings are in line with previous studies,
reporting that a statistically important number of children with ASD, score similarly to
DLD children on different language tasks [15], and also that an ASD-LI group exists within
the ASD population, which has found to be more impaired on language tasks than ALN
children [14,68]. Our results further support the presence of a language phenotype in ASD
similar to that of DLD, that is, presenting numerous and serious dysfunctions in several
aspects of language [15,68,123].
Moreover, not that surprisingly, ALN and TD groups revealed no important differ-
ences on language tasks, providing further support at the assumption of the existence of a
subtype inside the ASD population, referred to as the “ASD-LI individuals”. This finding is
further supported by De Fossé and colleagues’ [124] study, who found similar asymmetry
of frontal language cortex between the ALN and TD control group, indicating that the
Broca’s area asymmetry reversal, is stronger connected with language impairment than
autism itself.
In the majority of language tasks administered, the ASD-LI group scored lower than
the DLD group. This finding could be explained by the fact that ASD-LI children except
for the language impairment presented, have additional deficits to manage (e.g., social
deficits). This complexity of symptoms with additional impairments in different fields
constitutes an additional obstacle in their effort for success on different tasks.
A special reference is worth making on the RS (SRT) subscale of the EREL instrument.
The group differences showed that ASD-LI as well as the DLD group produced more
deficits than the TD group. The finding that the clinical groups with LI perform worse
than the typical groups is in line with other studies (e.g., [27,32]). Surprisingly, though,
there were not found to be any important differences between the ALN and ASD-LI or
DLD groups, or between the ALN and TD groups. The means and standard deviations of
the task show differences between the groups, whereas the ALN group scores were higher
than both the language-impaired groups, though not statistically important. However, our
results are in line with Conti-Ramsden et al.’s [27] statements, who argue that sentence
repetition could be the strongest clinical marker for the identification of DLD, and can
significantly differentiate language-impaired groups from typical language ability, as did
in our study. A repetition task considers contributing to the identification of children’s
weaknesses and strengths in the language domain [37]. Indeed, in our study SRT managed
to reveal differences in both language-impaired groups in comparison to the TD group.
Similarly, the NWR task produced important differences among participants in differ-
ent groups. Both the language-impaired groups demonstrated worse performance than the
TD group. The results of the present study are consistent with several previous studies,
supporting that children with ASD-LI and those with DLD have similarly poor perfor-
mance on the NWR task. Evidence supporting NWR deficits in ASD-LI is compelling and
the results of this study confirmed previous findings in the area [13,15,18,125].
Furthermore, the ICC model for all language tasks showed that there are important
between-group differences, explaining a significant proportion of the observed variance.
This finding provides additional support to the assumption that DLD and ASD constitute
Brain Sci. 2021, 11, 589 22 of 29

points on a continuum of the same disorder instead of separate conditions [19]. Indeed,
the experimental groups present great inter-participants’ heterogeneity, as the ICC model
shows for all language tasks, suggesting that they do not constitute separate solid groups
with distinct traits and symptomatology, but rather are highlighting substantial overlap
amongst groups with “distinct” diagnoses [117].
Especially, our results support the existence of an intermediate case of shared symp-
Brain Sci. 2021, 11, 589 toms, namely ASD-LI, which indicates an important connection between ASD and23DLD. of 30
Interestingly, a great number of our participants fell in this intermediate category, not
only about the language domain. This finding significantly supports the assumption of
existing comorbidity between ASD and DLD. Important relations have been also found to
exist between the clinical groups, as revealed by the majority of the tasks and measures
exist between the clinical groups, as revealed by the majority of the tasks and measures
employed. These relations, further support an overlapping situation between ASD and
employed. These relations, further support an overlapping situation between ASD and
DLD. For instance, ASD-LI and DLD groups were found to have similar performance in
DLD. For instance, ASD-LI and DLD groups were found to have similar performance in
all tasks of the language domain. This finding points to a comorbid relation in linguistics.
all tasks of the language domain. This finding points to a comorbid relation in linguistics.
Our results
Our results allow
allow usus to
to argue
argue inin favor
favor ofof ASD
ASD andand DLD
DLD as as disorders
disorders lying
lying on on aa continuum
continuum
of symptoms, supporting the second theory presented in Figure 2,
of symptoms, supporting the second theory presented in Figure 2, that the two disorders that the two disorders
consist of different manifestations of the same pathology, and lie on
consist of different manifestations of the same pathology, and lie on the same continuum. the same continuum.
As illustrated
As illustrated inin Figure
Figure 5,
5, the
the group
group of of children
children with
with ASD ASD with
with high
high language
language ability
ability
(ALN) resides at the one end of the line while, DLD and ASD-LI
(ALN) resides at the one end of the line while, DLD and ASD-LI groups are located groups are located in
proximity
in proximity at the other
at the endend
other of the lineline
of the of the continuum
of the continuum (low language
(low language ability). It isItim-
ability). is
portant to mention, that the points on the continuum are not related
important to mention, that the points on the continuum are not related to our numeric to our numeric data.
The clinical
data. groups
The clinical are only
groups visually
are only illustrated
visually in this
illustrated in figure to facilitate
this figure the overall
to facilitate un-
the overall
derstanding of this theory. We further expected DLD children to
understanding of this theory. We further expected DLD children to present difficultiespresent difficulties pri-
marily on on
primarily their general
their communication
general communication ability as itasis ittested
ability by the
is tested byCCC-2,
the CCC-2,whilewhile
ASD
children
ASD would
children have have
would demonstrated additional
demonstrated impairments
additional impairments in theinsocial interaction
the social sub-
interaction
scale (SIDC)
subscale (SIDC)of CCC-2
of CCC-2[108], based
[108], basedonon the
theprimary
primarydeficits
deficitsofofeacheachdisorder.
disorder. BothBoth of our
of our
hypotheses have been met.

Figure 5. An
Figure 5. An iconic
iconic representation
representation of
of language
language ability
ability of
of children
children with
with ASD
ASD and
and DLD.
DLD.

The
The CCC-2
CCC-2 instrument
instrumentidentifies
identifieschildren
childrenwithwithdisproportionate
disproportionate pragmatic
pragmaticand social
and so-
impairments.
cial impairments. In this study
In this studythis assumption
this assumption is is
confirmed,
confirmed,asasthe
theASD
ASDgroup
groupofof children
children
were
were more
moredeficit
deficitininthis
thiscomposite,
composite, proposing
proposing increased pragmatic
increased impairments,
pragmatic unrelated
impairments, unre-
to one’s structural language ability. It is of great importance the finding of different
lated to one’s structural language ability. It is of great importance the finding of different
impairments characterized by the fundamental deficits presented by each group, based
impairments characterized by the fundamental deficits presented by each group, based
on CCC-2 composites. The DLD group presented deficits on the General Communication
on CCC-2 composites. The DLD group presented deficits on the General Communication
Composite (GCC), and fewer deficits on the SIDC subscale confirming DLD’s children
Composite (GCC), and fewer deficits on the SIDC subscale confirming DLD’s children
diagnosis, and also the ASD group has demonstrated impairments in GCC and SIDC
diagnosis, and also the ASD group has demonstrated impairments in GCC and SIDC sim-
simultaneously, as it is expected for individuals in the autistic spectrum. The results on
ultaneously, as it is expected for individuals in the autistic spectrum. The results on GCC
GCC show that no important differences are observed in the three clinical groups related
show that no important differences are observed in the three clinical groups related to
to their GCC symptoms. This result was expected, as both developmental conditions
their GCC symptoms. This result was expected, as both developmental conditions are
are proposed to score similarly on GCC (scores below 55), significantly lower than TD
proposed to score similarly on GCC (scores below 55), significantly lower than TD groups
groups [117,118]. The findings for SIDC were partially expected. Specifically, the ALN
[117,118]. The findings for SIDC were partially expected. Specifically, the ALN group of
group of children presented more social deviance deficits, which was excepted and our
children presented
hypothesis was met. more
Still,social deviance
no other deficits, which
experimental group was
did, excepted and our
either ASD-LI hypothesis
group, which
was met. Still, no other experimental group did, either ASD-LI group, which
was assumed to present significantly important deficits in the social deviance domain, as was assumed
to present
ALN did. significantly important deficits in the social deviance domain, as ALN did.
Nevertheless, our
Nevertheless, our results
results areare confirmed
confirmed from from previous
previous studies
studies supporting
supporting that
that the
the
SIDC score is expected to be eight or less in children with ASD [117,118]. Importantly,
SIDC score is expected to be eight or less in children with ASD [117,118]. Importantly, the the
ASD-LI group
ASD-LI group also
also scored
scored below
below eight
eight on
on the
the SIDC
SIDC composite.
composite. AtAt the
the other
other end
end ofof the
the
distribution, children with DLD usually score above eight, as in our study [117,118]. Our
findings gain support from a validation study [117], in which the GCC was effective in
distinguishing between children with communication impairments (including both DLD
and autism) and typically developing children. The SIDC was usually less than eight in
Brain Sci. 2021, 11, 589 23 of 29

distribution, children with DLD usually score above eight, as in our study [117,118]. Our
findings gain support from a validation study [117], in which the GCC was effective in
distinguishing between children with communication impairments (including both DLD
and autism) and typically developing children. The SIDC was usually less than eight in
children with ASD.
Nevertheless, these findings raise questions about the extent to which pragmatic
impairments are a secondary consequence of speech and language disorder. For instance,
a large proportion of cognitively able children with ASD experience additional language
impairments [15,126]. On the other hand, non-autistic children with LI are not only
characterized by structural language deficits, but many also experience difficulties with
pragmatic aspects of language [117]. In this study DLD group of children was found free
of social- pragmatics deficits as it was captured by the SIDC composite of CCC-2.
It is apparent that language impairment adversely affects social interaction. However,
the present data show that some children have pragmatic impairments out of keeping with
their structural language skills. This is in line with previous studies showing that children
with good structural language can nevertheless have pragmatic difficulties [127]. At a
theoretical level, to further investigate these language—and consequently—communicative
abilities is critical as communication deficits have been related to difficulties in understand-
ing other’s minds and in social functioning [128–130].

4.1. Limitations
It is, finally, important to acknowledge some methodological limitations. Evidence
supporting NWR deficits in ASD-LI are numerous and the results of this study are con-
sistent with several previous studies in the area [13,15,18,125]. In more recent studies that
have compared individuals with ASD-LI and those with DLD on NWR tasks, results pro-
vided evidence for significant differences in the error patterns between these two groups.
We did not directly compare the error patterns in the ASD-LI and DLD groups of children.
In addition, we did not measure articulation or oro-motor skills in the ASD and DLD
children. Therefore, the NWR deficits could be attributed to oro-motor deficits in the
ASD-LI or DLD children, rather than explained by cognitive deficits.
Nevertheless, a study comparing NWR performance in ASD and DLD [125] included
only DLD children with no referral of articulation impairments and still found poor NWR
in these children. Therefore, impairments related to the NWR in DLD may not solely be
explained by concurrent articulation deficits. Nevertheless, future research should continue
to attempt to illustrate the specific cognitive and linguistic relation to poor NWR and SNR
in ASD-LI and DLD children, to further inform the debate on etiological overlap in these
two developmental conditions.
Bias can be introduced at any phase of a research process. We tried to prevent
systematic errors and bias during designing, conducting, and reporting the results of our
study. However, we believe some selection bias was inserted in the study due to the fact
that participants, especially children in pediatric groups, were recruited through a call from
speech pathologists and special educators who knew children.

4.2. Strengths of the Study

This is the first study, to our knowledge, conducted in Cyprus investigating the ASD
and DLD population directly. The prevalence of ASD and DLD individuals in Cyprus
is still not determined even if Special Education Services needs are elevated. This study
succeeded to locate and examine an important number of ASD and DLD individuals all
over the country, given the small numbers of the population in Cyprus. This study could
be considered as attaining numbers of participants similar to what we could find in an
epidemiological study in Cyprus, by addressing these two developmental conditions in a
wide range of childhood ages.
In addition, our results are also significant, as they confirm the existence of this inter-
mediate set of symptoms that consists of the ASD-LI group, in a different language such
Brain Sci. 2021, 11, 589 24 of 29

as the Greek language. Greek, is a language with significant peculiarities and complexity
in expression. The view that this middle ground exists is further reinforced by our data.
This category of mixed symptoms deserves to be investigated closely, to obtain optimal
results, both in research and clinical practice. The assumption that this intermediate group
does, indeed, exist will no longer cause stress to specialists working with clinical groups
about what is the most appropriate intervention plan to follow. Instead, they will be
provided with the necessary support and guidance to work in the light of this middle
cluster of symptoms.
To further investigate the existence and relation of the two disorders will provide
important and valuable information scientifically and clinically. The need for the develop-
ment of effective intervention programs for both developmental conditions is highlighted
by the specialists of all fields working with these populations, as the number of the affected
children increases daily.

5. Conclusions
In conclusion, our results further support the assumption of common symptomatology
of the two disorders and are in line with the suggestion of moving beyond the “labels” and
start subgrouping individuals in response to their characteristics, instead of trying to detect
the symptoms following an already given diagnosis. The apparent need to divide the
ASD group into two different categories based on the divergence of symptoms presented,
namely ASD children with additional language impairment and ASD children without,
is an indicative starting point of the need to change our perspectives and methodologies.
Beyond that, if we accept that this category exists, it is important to provide answers to
consequent questions. For instance, at what point in the continuum can we place this
“shared” category of social symptoms? Can anyone argue that it is placed exactly in
the middle of the two predominant disorders? Is it possibly closer to autism disorder,
or perhaps to language disorder? With all this in mind, unquestionably, the research
community has still a lot to discover and many challenges to overcome concerning this
particular category.

Author Contributions: Conceptualization, N.G. and G.S.; methodology, N.G. and G.S.; formal
analysis, G.S.; investigation, N.G.; resources, G.S.; data curation, N.G.; writing—original draft
preparation, N.G.; writing—review and editing, N.G. and G.S.; visualization, G.S.; supervision,
G.S.; project administration, G.S.; funding acquisition, G.S. All authors have read and agreed to the
published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: The study was conducted according to the guidelines of the
Declaration of Helsinki, and approved by the Cyprus National Bioethics Committee) (EEBK/EP/2018/16;
28/06/2018).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author, GS. The data are not publicly available due to their containing information
that could compromise the privacy of research participants.
Acknowledgments: We would like to thank all children and families who participated in this study.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Williams, D.; Botting, N.; Boucher, J. Language in Autism and Specific Language Impairment: Where Are the Links? Psychol. Bull.
2008, 134, 944–963. [CrossRef] [PubMed]
2. Riches, N.G.; Loucas, T.; Baird, G.; Charman, T.; Simonoff, E. Sentence repetition in adolescents with Specific Language
Impairments and autism: An investigation of complex syntax. Int. J. Lang. Commun. Disord. 2010, 45, 47–60. [CrossRef]
Brain Sci. 2021, 11, 589 25 of 29

3. Bishop, D.V.M. Pragmatic language impairment: A correlate of SLI, a distinct subgroup, or part of the autistic continuum? In
Speech and Language Impairments in Children: Causes, Characteristics, Intervention and Outcome; Bishop, D.V.M., Leonard, L.B., Eds.;
Psychology Press: Hove, UK, 2000; pp. 99–113.
4. Tager-Flusberg, H.; Lord, C.; Paul, R. Language, and communication in autism. In Handbook of Autism and Pervasive Developmental
Disorders, 3rd ed.; Volkmar, F., Paul, R., Klin, A., Cohen, D., Eds.; John Wiley & Sons: Hoboken, NJ, USA, 2005; pp. 335–364.
5. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders: DSM-5; American Psychiatric Publishing:
Arlington, VA, USA, 2013.
6. Bishop, D.V.M.; Snowling, M.J.; Thompson, P.A.; Greenhalgh, T. Phase 2 of CATALISE: A multinational and multidisciplinary
Delphi consensus study of problems with language development: Terminology. J. Child Psychol. Psychiatry 2017, 58, 1068–1080.
[CrossRef] [PubMed]
7. Bishop, D.V.M. Uncommon Understanding: Development and Disorders of Language Comprehension in Children; Psychology Press:
Hove, UK, 1997; p. 278.
8. Roqueta, C.A.; Katsos, N. A distinction between linguistic and social pragmatics helps the precise characterization of pragmatic
challenges in children with autism spectrum disorders and developmental language disorder. J. Speech Lang. Her Res. 2020, 63,
1494–1508. [CrossRef]
9. Spanoudis, G. Theory of mind and Specific Language Impairment in school-age children. J. Commun. Disord. 2016, 61, 83–96.
[CrossRef] [PubMed]
10. Levy, S.E.; Giarelli, E.; Lee, L.C.; Schieve, L.A.; Kirby, R.S.; Cunniff, C.; Nicholas, J.; Reaven, J.; Rice, C.E. Autism spectrum
disorder and co-occurring developmental, psychiatric, and medical conditions among children in multiple populations of the
United States. J. Dev. Behav. Pediatr. 2010, 31, 267–275. [CrossRef] [PubMed]
11. Baird, G.; Simonoff, E.; Pickles, A.; Chandler, S.; Loucas, T.; Meldrum, D.; Charman, T. Prevalence of disorders of the autism
spectrum in a population cohort of children in South Thames: The Special Needs and Autism Project (SNAP). Lancet 2006, 368,
210–215. [CrossRef]
12. Tager-Flusberg, H. Language and understanding minds: Connections in autism. In Understanding Other Minds: Perspectives from
Developmental Cognitive Neuroscience; Baron-Cohen, S., Tager-Flusberg, H., Cohen, D.J., Eds.; University Press: Oxford, UK, 2000;
pp. 124–149.
13. Whitehouse, A.J.O.; Barry, J.G.; Bishop, D.V.M. Further defining the language impairment of autism: Is there a Specific Language
Impairment subtype? J. Commun. Disord. 2008, 41, 319–336. [CrossRef]
14. Tager-Flusberg, H.; Joseph, R.M. Identifying neurocognitive phenotypes in autism. Philos. Trans. R. Soc. Lond. B Biol. Sci. 2003,
358, 303–314. [CrossRef]
15. Kjelgaard, M.M.; Tager-Flusberg, H. An investigation of language impairment in autism: Implications for genetic subgroups.
Lang. Cogn. Process. 2001, 16, 287–308. [CrossRef]
16. Roberts, J.A.; Rice, M.L.; Tager-Flusberg, H. Tense marking in children with autism. Appl. Psycholinguist. 2004, 25, 429–448.
[CrossRef]
17. Durrleman, S.; Delage, H. Autism Spectrum Disorder and Specific Language Impairment: Overlaps in Syntactic Profiles. Lang.
Acquis. 2016, 23, 361–386. [CrossRef]
18. Riches, N.G.; Loucas, T.; Baird, G.; Charman, T.; Simonoff, E. Non-word repetition in adolescents with Specific Language
Impairment and autism plus language impairments: A qualitative analysis. J. Commun. Disord. 2011, 44, 23–36. [CrossRef]
19. Bishop, D.V.M. Overlaps between autism and language impairment: Phenomimicry or shared etiology. Behav. Genet. 2010, 40,
618–629. [CrossRef] [PubMed]
20. Norbury, C.F.; Gooch, D.; Wray, C.; Baird, G.; Charman, T.; Simonoff, E.; Vamvakas, G.; Pickles, A. The impact of nonverbal ability
on prevalence and clinical presentation of language disorder: Evidence from a population study. J. Child Psychol. Psychiatry 2016,
11, 1247–1257. [CrossRef] [PubMed]
21. Leonard, L.B. Children with Specific Language Impairment; MIT Press: Cambridge, UK, 1998.
22. Montgomery, J.W. Working memory and comprehension in children with specific language impairment: What we know so far. J.
Commun. Disord. 2003, 36, 221–231. [CrossRef]
23. Vugs, B.; Cuperus, J.; Hendriks, M.; Verhoeven, L. Visuospatial working memory in Specific Language Impairment: A meta-
analysis. Res. Dev. Disabil. 2013, 34, 2586–2597. [CrossRef] [PubMed]
24. Henry, L.A.; Messer, D.J.; Nash, G. Executive functioning in children with specific language impairment. J. Child Psychol.
Psychiatry 2012, 53, 37–45. [CrossRef]
25. Bishop, D.V.M.; Snowling, M. Developmental dyslexia and specific language impairment: Same or different? Psychol. Bull. 2004,
130, 858–886. [CrossRef]
26. Leonard, L.; Caselli, M.C.; Bortolini, U.; McGregor, K.K. Morphological Deficits, in Children with Specific Language Impairment:
The Status of Features in the Underlying Grammar. Lang. Acquis. 1992, 2, 151–179. [CrossRef]
27. Conti-Ramsden, G.; Botting, N.; Simkin, Z.; Knox, E. Follow-up of children attending infant language units: Outcomes at 11 years
of age. Int. J. Lang. Commun. Disord. 2001, 36, 207–219. [CrossRef]
28. Coady, J.; Evans, J.L. Uses and interpretations of nonword repetition tasks in children with and without Specific Language
Impairment (SLI). Int. J. Lang. Commun. Disord. 2008, 43, 1–40. [CrossRef] [PubMed]
Brain Sci. 2021, 11, 589 26 of 29

29. Rice, M.L.; Wexler, K.; Cleave, P.L. Specific language impairment as a period of extended optional infinitive. J. Speech Lang. Hear.
Res. 1995, 38, 850–863. [CrossRef] [PubMed]
30. Marchman, V.A.; Wulfeck, B.; Ellis-Weismer, S.E. Morphological productivity in children with normal language and SLI: A study
of the English past tense. J. Speech Lang. Hear. Res. 1999, 42, 206–219. [CrossRef] [PubMed]
31. Simkin, Z.; Conti-Ramsden, G. Nonword repetition and grammatical morphology: Normative data for children in their final year
of primary school. Int. J. Lang. Commun. Disord. 2001, 36, 395–404. [PubMed]
32. Oetting, J.B.; McDonald, J.L.; Seidel, C.M.; Hegarty, M. Sentence recall by children with SLI across two mainstream dialects of
English. J. Speech Lang. Hear. Res. 2016, 59, 183–194. [CrossRef]
33. Stokes, S.; Wong, A.; Fletcher, P.; Leonard, L. Nonword Repetition and Sentence Repetition as Clinical Markers of Specific
Language Impairment: The Case of Cantonese. J. Speech Lang. Hear. Res. 2006, 49, 219–236. [CrossRef]
34. Thordardottir, E.; Kehayia, E.; Mazer, B.; Lessard, N.; Majnemer, A.; Sutton, A.; Trudeau, N.; Chilingaryan, G. Sensitivity and
specificity of French language and processing measures for the identification of primary language impairment at age 5. J. Speech
Lang. Hear. Res. 2011, 54, 580–597. [CrossRef]
35. Leclercq, A.; Quémart, P.; Magis, D.; Maillart, C. The sentence repetition task: A powerful diagnostic tool for French children with
specific language impairment. Res. Dev. Disabil. 2014, 35, 2423–2430. [CrossRef]
36. Theodorou, E.; Kambanaros, M.; Grohmann, K.K. Sentence Repetition as a Tool for Screening Morphosyntactic Abilities of Bilectal
Children with SLI. Front. Psychol. 2017, 8, 1–13. [CrossRef]
37. Seeff-Gabriel, B.; Chiat, S.; Dodd, B. Sentence imitation as a tool in identifying expressive morphosyntactic difficulties in children
with severe speech difficulties. Int. J. Lang. Commun. Disord. 2010, 45, 691–702. [CrossRef] [PubMed]
38. Girbau, D. The Non-word Repetition Task as a clinical marker of Specific Language Impairment in Spanish-speaking children.
First Lang. 2016, 36, 30–49. [CrossRef]
39. Bortolini, U.; Arfe´, B.; Caselli, M.C.; Degasperi, L.; Deevy, P.; Leonard, L.B. Clinical markers for Specific Language Impairment in
Italian: The contribution of clitics and non-word repetition. Int. J. Lang. Commun. Disord. 2006, 41, 695–712. [CrossRef] [PubMed]
40. Graf Estes, K.; Evans, J.L.; Else-Quest, N.M. Differences in the Nonword Repetition Performance of Children with and Without
Developmental Language Disorder: A Meta-Analysis. J. Speech Lang. Hear. Res. 2007, 50, 177–195. [CrossRef]
41. Bishop, D.V.M.; Donlan, C. Nonword repetition as a behavioural marker for inherited language impairment: Evidence from a
twin study. J. Child Psychol. Psychiatry 1996, 37, 391–403. [CrossRef] [PubMed]
42. Conti-Ramsden, G.; Hesketh, A. Risk markers for SLI: A study of young language-learning children. Int. J. Lang. Commun. Disord.
2003, 38, 251–263. [CrossRef] [PubMed]
43. Botting, N.; Conti-Ramsden, G. Autism, primary pragmatic difficulties and Specific Language Impairment: Can we distinguish
them using psycholinguistic markers? Dev. Med. Child Neurol. 2003, 45, 515–545. [CrossRef]
44. Bishop, D.V.M.; Chan, J.; Adams, C.; Hartley, J.; Weir, F. Conversational responsiveness in Specific Language Impairment:
Evidence of disproportionate pragmatic difficulties in a subset of children. Dev. Psychopathol. 2000, 12, 177–199. [CrossRef]
[PubMed]
45. Friedmann, N.; Novogrodsky, R. Subtypes of Sli: Sysli, Phosli, Lesli, and Prasli. In Language Acquisition and Development; Gavarró,
A., Freitas, M.J., Eds.; Cambridge Scholars Press/CSP: Newcastle, UK, 2008; pp. 205–217.
46. Van der Lely, H.K. SLI in children: Movement, economy, and deficits in the computational-syntactic system. Lang. Acquis. 1998, 7,
161–192. [CrossRef]
47. Stavrakaki, S. Greek neurolinguistics: The state of the art. J. Greek Linguist. 2005, 6, 187–234. [CrossRef]
48. Petinou, K.; Terzi, A. Clitic misplacement among normally developing and SLI children and the status of Infl Heads. Lang. Acquis.
2002, 10, 1–28. [CrossRef]
49. Stavrakaki, S.; Van der Lely, H. Production and comprehension of pronouns by Greek children with specific language impairment.
Br. J. Dev. Psychol. 2010, 28, 189–216. [CrossRef] [PubMed]
50. Stavrakaki, S. Comprehension of reversible relative clauses in specifically language impaired and normally developing Greek
children. Brain Lang. 2001, 77, 419–431. [CrossRef] [PubMed]
51. Stavrakaki, S. Developmental perspectives on Specific Language Impairment: Evidence from the production of wh-questions by
Greek SLI children over time. Adv. Speech Lang. Pathol. 2006, 8, 384–396. [CrossRef]
52. Tsimpli, I.M.; Stavrakaki, S. The effects of a morphosyntactic deficit in the determiner system: The case of a Greek SLI child.
Lingua 1999, 108, 31–85. [CrossRef]
53. Tsimpli, I.M. LF-interpretability and language development: A study of verbal and nominal features in Greek normally developing
and SLI children. Brain Lang. 2001, 77, 432–448. [CrossRef]
54. Theodorou, E.; Grohmann, K.K. Object clitics in Cypriot Greek children with SLI. Lingua 2015, 161, 144–158. [CrossRef]
55. Manika, S.; Varlokosta, S.; Wexler, K. The lack of omission of clitics in Greek children with SLI: An experimental study. In BUCLD
35: Proceedings of the 35th Boston University Conference on Language Development; Danis, N., Mesh, K., Sung, H., Eds.; Cascadilla
Press: Somerville, MA, USA, 2011; pp. 429–437.
56. Varlokosta, S.; Konstantzou, K.; Nerantzini, M. Clitic production in Greek SLI. In Proceedings of the Conference Specific Language
Impairment—Diagnosis, Prognosis, Intervention, Warsaw, Poland, 5–8 July 2012.
57. Theodorou, E.; Grohmann, K.K. Narratives in Cypriot Greek mono- and bilingual children with SLI. In Proceedings of the 3rd
ISCA Workshop Tutorial and Research Workshop on Experimental Linguistics, Athens, Greece, 25–27 August 2010; pp. 185–188.
Brain Sci. 2021, 11, 589 27 of 29

58. Spanoudis, G.; Natsopoulos, D.; Panayiotou, G. Mental verbs and pragmatic language difficulties. Int. J. Lang. Commun. Disord.
2007, 42, 487–504. [CrossRef]
59. Spanoudis, G.; Papadopoulos, T.C.; Spyrou, S. Specific language impairment and dyslexia: Categorical distinction or continuum?
J. Learn. Disabil. 2019, 52, 3–14. [CrossRef]
60. Kanner, L. Irrelevant and metaphorical language in early infantile autism. Am. J. Psychiatry 1946, 103, 242–245. [CrossRef]
61. Rutter, M. Concepts of autism: A review of research. J. Child Psychol. Psychiatry 1968, 9, 1–25. [CrossRef] [PubMed]
62. Bartolucci, G.; Albers, R.J. Deictic categories in the language of autistic children. J. Autism Child. Schiz. 1974, 4, 131–141. [CrossRef]
63. Bartolucci, G.; Pierce, S. A preliminary comparison of phonological development in autistic, normal, and mentally retarded
subjects. Br. J. Disord. Commun. 1977, 12, 137–147. [CrossRef]
64. Bartoclucci, G.; Pierce, S.; Streiner, D.; Eppel, P. Phonological investigation of verbal autistic and mentally retarded subjects. J.
Autism Child. Schiz. 1976, 6, 303–315. [CrossRef]
65. Tager–Flusberg, H.; Calkins, S.; Nolin, T.; Baumberger, T.; Anderson, M.; Chadwick–Dias, A. A longitudinal study of language
acquisition in autistic and Downs syndrome children. J. Autism Dev. Disord. 1990, 20, 1–21. [CrossRef]
66. Tek, S.; Mesite, L.; Fein, D.; Naigles, L. Longitudinal analyses of expressive language development reveal two distinct language
profiles among young children with autism spectrum disorders. J. Autism Dev. Disord. 2013, 44, 75–89. [CrossRef]
67. Gernsbacher, M.A.; Morson, E.M.; Grace, E. Language and Speech in Autism. Annu. Rev. Linguist. 2016, 2, 413–425. [CrossRef]
68. Tager-Flusberg, H. Defining language phenotype in autism. Clin. Neurosci. Res. 2006, 6, 219–224. [CrossRef]
69. Boucher, J. Research review: Structural language in autistic spectrum disorder—Characteristics and causes. J. Child Psychol.
Psychiatry 2012, 53, 219–233. [CrossRef]
70. Tager-Flusberg, H.; Caronna, E. Language Disorders: Autism and Other Pervasive Developmental Disorders. Pediatr. Clin. North
Am. 2007, 54, 469–481. [CrossRef]
71. Ellis Weismer, S.; Kover, S.T. Preschool language variation, growth, and predictors in children on the autism spectrum. J. Child
Psychol. Psychiatry 2015, 56, 1327–1337. [CrossRef] [PubMed]
72. Anderson, D.K.; Lord, C.; Risi, S.; DiLavore, P.S.; Shulman, C.; Thurm, A.; Welch, K.; Pickles, A. Patterns of growth in verbal
abilities among children with autism spectrum disorder. J. Consult. Clin. Psychol. 2007, 75, 594–604. [CrossRef]
73. Norrelgen, F.; Fernell, E.; Eriksson, M.; Hedvall, A.; Persson, C.; Sjolin, M.; Gillberg, C.; Kjellmer, L. Children with autism
spectrum disorders who do not develop phrase speech in the preschool years. Autism 2014, 19, 934–943. [CrossRef]
74. Brignell, A.; Morgan, A.; Woolfenden, S.; Klopper, F.; May, T.; Sarkozy, V.; Williams, K. A systematic review and meta-analysis
of the prognosis of language outcomes for individuals with autism spectrum disorder. Autism Dev. Lang. Impair. 2018, 3, 1–19.
[CrossRef]
75. Herlihy, L.; Knoch, K.; Vibert, B.; Fein, D. Parents’ first concerns about toddlers with autism spectrum disorder: Effect of sibling
status. Autism 2015, 19, 20–28. [CrossRef]
76. Hudry, K.; Chandler, S.; Bedford, R.; Pasco, G.; Gliga, T.; Elsabbagh, M.; Johnson, M.H.; Charman, T. Early language profiles in
infants at high-risk for autism spectrum disorders. J. Autism Dev. Disord. 2014, 44, 154–167. [CrossRef]
77. Tager-Flusberg, H. A psychological approach to understanding the social and language impairments in autism. Int. Rev. Psychiatry
1999, 11, 325–334. [CrossRef]
78. Peristeri, E.; Andreou, M.; Tsimpli, I.M. Syntactic and Story Structure Complexity in the Narratives of High- and Low-Language
Ability Children with Autism Spectrum Disorder. Front. Psychol. 2017, 8, 2027. [CrossRef]
79. Terzi, A.; Marinis, T.; Francis, K. The Interface of Syntax with Pragmatics and Prosody in Children with Autism Spectrum
Disorders. J. Autism Dev. Disord. 2016, 46, 2692–2706. [CrossRef] [PubMed]
80. Kambanaros, M.; Christou, N.; Grohmann, K.K. Interpretation of compound words by Greek-speaking children with autism
spectrum disorder plus language impairment (ASD–LI). Clin. Linguist. Phon. 2019, 33, 135–174. [CrossRef] [PubMed]
81. Bishop, D.V.M.; Norbury, C.F. Exploring the borderlands of autistic disorder and Specific Language Impairment: A study using
standardized diagnostic instruments. J. Child Psychol. Psychiatry 2002, 43, 917–929. [CrossRef] [PubMed]
82. Lindgren, K.A.; Folstein, S.E.; Tomblin, J.B.; Tager-Flusberg, H. Language and reading abilities of children with autism spectrum
disorders and Specific Language Impairment and their first-degree relatives. Autism Res. 2009, 2, 22–38. [CrossRef]
83. Lloyd, H.; Paintin, K.; Botting, N. Performance of children with different types of communication impairment on the Clinical
Evaluation of Language Fundamentals (CELF). Child Lang. Teach Ther. 2006, 22, 205–223. [CrossRef]
84. Manolitsi, M.; Botting, N. Language abilities in children with autism and language impairment: Using narrative as an additional
source of clinical information. Child Lang. Teach Ther. 2011, 27, 39–55. [CrossRef]
85. Demouy, J.; Plaza, M.; Xavier, J.; Ringeval, J.; Chetouani, M.; Périsse, D.; Chauvin, D.; Viaux, S.; Golse, B.; Cohen, D.; et al.
Differential language markers of pathology in autism, pervasive developmental disorder not otherwise specified and Specific
Language Impairment. Res. Autism Spectr. Disord. 2011, 5, 1402–1412. [CrossRef]
86. McGregor, K.K.; Berns, A.J.; Owen, A.J.; Michels, S.A.; Duff, D.; Bahnsen, A.J.; Lloyd, M. Associations between syntax and the
lexicon among children with or without ASD and language impairment. J. Autism Dev. Disord. 2012, 42, 35–47. [CrossRef]
87. Tuller, L.; Ferré, S.; Prévost, P.; Barthez, M.; Malvy, J.; Bonnet-Brilhault, F. The effect of computational complexity on the acquisition
of French by children with ASD. In Innovative Investigations of Language in Autism Spectrum Disorder; Naigles, L., Ed.; Walter de
Gruyter: Washington, DC, USA, 2017; pp. 115–140.
Brain Sci. 2021, 11, 589 28 of 29

88. Ramírez-Santana, G.M.; Acosta-Rodríguez, V.M.; Hernández-Expósito, S.A. Comparative study of language phenotypes in
Autism Spectrum Disorder and Developmental Language Disorder. Psicothema 2019, 31, 437–442.
89. Gladfelter, A.; Barron, K.L. How Children with Autism Spectrum Disorder, Specific Language Impairment, and Typical Language
Learn to Produce Global and Local Semantic Features. Brain Sci. 2020, 10, 231. [CrossRef]
90. Modyanova, N.; Perovic, A.; Wexler, K. Grammar is differentially impaired in subgroups of autism spectrum disorders: Evidence
from an investigation of tense marking and morphosyntax. Front. Psychol. 2017, 8, 1–23. [CrossRef]
91. Ellis Weismer, S.; Davidson, M.M.; Gangopadhyay, I.; Sindberg, H.; Roebuck, H.; Kaushanskaya, M. The role of nonverbal
working memory in morphosyntactic processing by children with specific language impairment and autism spectrum disorders.
J. Neurodev. Disord. 2017, 9, 1–16. [CrossRef] [PubMed]
92. Schaeffer, J.; Van Witteloostuijn, M.; Creemers, A. Article choice, theory of mind, and memory in children with high-functioning
autism and children with Specific Language Impairment. Appl. Psycholinguist. 2018, 39, 89–115. [CrossRef]
93. Norbury, C.F.; Gemmell, T.; Paul, R. Pragmatics abilities in narrative production: A cross-disorder comparison. J. Child Lang. 2014,
41, 485–510. [CrossRef]
94. Conti-Ramsden, G.; Simkin, Z.; Pickles, A. Estimating familial loading in SLI: A comparison of direct assessment versus parental
interview. J. Speech Lang. Hear. Res. 2006, 49, 88–101. [CrossRef]
95. Leyfer, O.T.; Tager-Flusberg, H.; Dowd, M.; Tomblin, J.B.; Folstein, S.E. Overlap between autism and Specific Language
Impairment: Comparison of Autism Diagnostic Interview and Autism Diagnostic Observation Schedule scores. Autism Res. 2008,
1, 284–296. [CrossRef]
96. Howlin, P.; Mawhood, L.; Rutter, M. Autism and developmental receptive language disorder—A follow-up comparison in early
adult life II: Social, behavioural, and psychiatric outcomes. J. Child Psychol. Psychiatry 2000, 41, 561–578. [CrossRef]
97. Mawhood, L.; Howlin, P.; Rutter, M. Autism and developmental receptive language disorder—A comparative follow-up in early
adult life. I: Cognitive and language outcomes. J. Child Psychol. Psychiatry 2000, 41, 547–559. [CrossRef]
98. Whitehouse, A.J.O.; Barry, J.G.; Bishop, D.V.M. The broader language phenotype of autism: A comparison with Specific Language
Impairment. J. Child Psychol. Psychiatry 2007, 48, 822–830. [CrossRef]
99. Loucas, T.; Riches, N.; Baird, G.; Pickles, A.; Simonoff, E.; Chandler, S.; Charman, T. Spoken word recognition in adolescents with
autism spectrum disorders and Specific Language Impairment. Appl. Psycholinguist. 2013, 34, 301–322. [CrossRef]
100. Sukenik, N.; Friedmann, N. ASD is not DLI: Individuals with autism and individuals with syntactic DLI show similar performance
level in syntactic tasks, but different error patterns. Front. Psychol. 2018, 9, 279. [CrossRef]
101. Creemers, A.; Schaeffer, J.C. Specific language impairment and high functioning autism: Evidence for distinct etiologies and for
modularity of grammar and pragmatics. In Proceedings of the 6th Conference on Generative Approaches to Language Acquisition North
America (GALANA 2015), University of Maryland, College Park, MD, USA, 19–21 February 2015; Perkins, L., Dudley, R., Gerard, J.,
Hitczenko, K., Eds.; Cascadilla Proceedings Project: Somerville, MA, USA, 2016; pp. 1–12.
102. Balaban, N.; Friedmann, N.; Ariel, M. The effect of theory of mind impairment on language: Referring after right hemisphere
damage. Aphasiology 2016, 30, 1424–1460. [CrossRef]
103. Schaeffer, J. Linguistic and cognitive abilities in children with specific language impairment as compared to children with
high-functioning autism. Lang. Acquis. 2016, 25, 5–23. [CrossRef]
104. Hudry, K.; Leadbitter, K.; Temple, K.; Slonims, V.; McConachie, H.; Aldred, C.; Howlin, P.; Charman, T.; PACT Consortium.
Preschoolers with autism show greater impairment in receptive compared with expressive language abilities. Int. J. Lang.
Commun. Disord. 2010, 45, 681–690. [CrossRef]
105. Taylor, L.J.; Whitehouse, A.J.O. Autism Spectrum Disorder, Language Disorder, and Social (Pragmatic) Communication Disorder:
Overlaps, Distinguishing Features, and Clinical Implications. Aust. Psychol. 2016, 51, 287–295. [CrossRef]
106. Whitehouse, A.J.O.; Line, E.A.; Watt, H.J.; Bishop, D.V.M. Qualitative aspects of developmental language impairment relate to
language and literacy outcome in adulthood. Int. J. Lang. Commun. Disord. 2009, 44, 489–510. [CrossRef]
107. Spanoudis, G.; Pahiti, J. Expressive and Receptive Language Evaluation: 5–12 Years of Age; Department of Psychology, University of
Cyprus: Nicosia, Cyprus, 2014.
108. Bishop, D.V.M. Children’s Communication Checklist, 2nd ed.; The Psychological Corporation: New York, NY, USA, 2003.
109. Lord, C.; Rutter, M.; DiLavore, P.C.; Risi, S.; Gotham, K.; Bishop, S. Autism Diagnostic Observation Schedule, 2nd ed.; Western
Psychological Services: Torrance, CA, USA, 2012.
110. Wechsler, D. Wechsler Intelligence Scale for Children, 1st ed.; The Psychological Corporation: London, UK, 1999.
111. Anns, S. The Role of Declarative Memory in the Acquisition of Conceptual Semantic Knowledge in Autism Spectrum Disorder.
Ph.D. Thesis, University of London, London, UK, 2016; pp. 1–242.
112. Reilly, S.; Tomblin, B.; Law, J.; McKean, C.; Mensah, F.K.; Morgan, A.; Goldfeld, S.; Nicholson, J.M.; Wake, M. Specific language
impairment: A convenient label for whom? Int. J. Lang. Commun. Disord. 2014, 49, 416–451. [CrossRef]
113. McGregor, K.; Goffman, L.; Owen Van Horne, A.; Hogan, T.; Finestack, L. Developmental Language Disorder: Applications for
Advocacy, Research, and Clinical Service. Perspect. ASHA Spec. Interest Groups 2020, 5, 38–46. [CrossRef]
114. Semel, E.; Wiig, E.; Secord, W. Clinical Evaluation of Language Fundamentals-3; The Psychological Corporation: San Antonio, TX,
USA, 1995.
115. Spanoudis, G.; Natsopoulos, D. Memory functioning and mental verbs acquisition in children with Specific Language Impairment.
Res. Dev. Disabil. 2011, 32, 2916–2926. [CrossRef]
Brain Sci. 2021, 11, 589 29 of 29

116. Glarou, E. Adaptation of Children’s Communication Checklist-2 (CCC-2; Bishop, 2003) in Greek Language and Implementation
in Parents of 4-7 Years Old Children. Unpublished. Undergraduate Thesis, University of Patra, Patra, Greece, 2015; pp. 1–102.
117. Norbury, C.; Nash, M.; Baird, G.; Bishop, D.V.M. Using a parental checklist to identify diagnostic groups in children with
communication impairment: A validation of the Children’s Communication Checklist—2. Int. J. Lang. Commun. Disord. 2004, 39,
345–364. [CrossRef] [PubMed]
118. Bishop, D.V.M.; Maybery, M.; Wong, D.; Maley, A.; Hallmayer, J. Characteristics of the broader phenotype in autism: A study of
siblings using the Children’s Communication Checklist—2. Am. J. Med. Genet. B. 2006, 141B, 117–122. [CrossRef]
119. Faul, F.; Erdfelder, E.; Lang, A.G.; Buchner, A. G*Power 3: A flexible statistical power analysis program for the social, behavioral,
and biomedical sciences. Behav. Res. Methods. 2007, 39, 175–191. [CrossRef]
120. Singer, J.D.; Willett, J.B. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence; Oxford University Press: New
York, NY, USA, 2003.
121. Collins, L.M. Analysis of longitudinal data: The integration of theoretical model, temporal design, and statistical model. Annu.
Rev. Psychol. 2006, 57, 505–528. [CrossRef] [PubMed]
122. Hox, J. Multilevel Analysis: Techniques and Applications; Lawrence Erlbaum Associates: Mahwah, NJ, USA, 2002.
123. Tager-Flusberg, H. Defining language impairments in a subgroup of children with autism spectrum disorder. Sci. China Life Sci.
2015, 58, 1044–1052. [CrossRef] [PubMed]
124. De Fossé, L.; Hodge, S.M.; Makris, N.; Kennedy, D.N.; Caviness, V.S., Jr.; McGrath, L.; Steele, S.; Ziegler, D.A.; Herbert, M.R.;
Frazier, J.A.; et al. Language-association cortex asymmetry in autism and specific language impairment. Ann. Neurol. 2004, 56,
757–766. [CrossRef]
125. Williams, D.; Payne, H.; Marshall, C. Non-word Repetition Impairment in Autism and Specific Language Impairment: Evidence
for Distinct Underlying Cognitive Causes. J. Autism Dev. Disord. 2013, 43, 404–417. [CrossRef] [PubMed]
126. Loucas, T.; Charman, T.; Pickles, A.; Simonoff, E.; Chandler, S.; Meldrum, D.; Baird, G. Autistic symptomatology and language
ability in Autism Spectrum Disorder and Specific Language Impairment. J. Child Psychol. Psychiatry 2008, 49, 1184–1192.
[CrossRef]
127. Conti-Ramsden, G.; Crutchley, A.; Botting, N. The extent to Which Psychometric Tests Differentiate Subgroups of Children with
SLI. J. Speech Lang. Hear. Res. 1997, 40, 765–777. [CrossRef] [PubMed]
128. Capps, L.; Kehres, J.; Sigman, M. Conversational abilities among children with autism and children with developmental delays.
Autism. 1998, 2, 325–344. [CrossRef]
129. Tager-Flusberg, H. Current theory and research on language and communication in autism. J. Autism Dev. Disord. 1996, 26,
169–172. [CrossRef]
130. Tager-Flusberg, H.; Cooper, J. Present and future possibilities for defining a phenotype for specific language impairment. J. Speech
Lang. Hear. Res. 1999, 42, 1001–1004. [CrossRef]