Lexical and semantic ability in groups of children with cochlear implants, language impairment and autism spectrum disorder

https://doi.org/10.1016/j.ijporl.2013.11.017Get rights and content

Abstract

Objective

Lexical-semantic ability was investigated among children aged 6–9 years with cochlear implants (CI) and compared to clinical groups of children with language impairment (LI) and autism spectrum disorder (ASD) as well as to age-matched children with normal hearing (NH). In addition, the influence of age at implantation on lexical-semantic ability was investigated among children with CI.

Methods

97 children divided into four groups participated, CI (n = 34), LI (n = 12), ASD (n = 12), and NH (n = 39). A battery of tests, including picture naming, receptive vocabulary and knowledge of semantic features, was used for assessment. A semantic response analysis of the erroneous responses on the picture-naming test was also performed.

Results

The group of children with CI exhibited a naming ability comparable to that of the age-matched children with NH, and they also possessed a relevant semantic knowledge of certain words that they were unable to name correctly. Children with CI had a significantly better understanding of words compared to the children with LI and ASD, but a worse understanding than those with NH. The significant differences between groups remained after controlling for age and non-verbal cognitive ability.

Conclusions

The children with CI demonstrated lexical-semantic abilities comparable to age-matched children with NH, while children with LI and ASD had a more atypical lexical-semantic profile and poorer sizes of expressive and receptive vocabularies. Dissimilar causes of neurodevelopmental processes seemingly affected lexical-semantic abilities in different ways in the clinical groups.

Introduction

Language skills develop during a child's early years and include both an expressive and a receptive aspect. Language delays and disorders may occur in one or both of these aspects, and are complex because they may also involve cognitive processes and disorders. The most common cause of specific language impairment (LI) is unknown [1]. Other clinical populations with known language delays and deficits are children with autism spectrum disorder (ASD), neurological disorders, brain injury, and mental retardation. Also, hearing impairment (HI) is an important cause of spoken language delay and deficits [2].

For children with cochlear implants (CI) the development of listening skills and spoken language is quite different than that of children with normal hearing (NH), as their hearing occurs via an electric stimulation of the auditory nerve instead of acoustic hearing [3], [4], [5], [6]. The population of children with CI is heterogeneous in many respects, including etiological factors, age at diagnosis of deafness and age at implantation [3], [7]. Language acquisition among children with CI might also be affected by genetic factors and interpersonal factors, such as communication mode, maternal education and degree of parental involvement [8], [9], [10]. In addition to these factors, their hearing impairment prior to implantation as well as their subsequent lack of opportunity for optimal incidentally learning spoken language through audition, especially in challenging conditions will have an impact on spoken language learning [11], [12].

Children with LI have relatively poor expressive and receptive vocabularies compared to children with typical development [13]. Many children with LI also demonstrate difficulties with retrieving words from their long-term memory [14]. Previous studies have revealed that children with LI have more difficulties with learning semantic relations between words compared to typically developing children [15]. In a comparative study of word association among children with LI and typically developing children aged 6–8 years [16], it was found that the responses from children with LI were less semantically relevant to the target words. Children with ASD typically demonstrate deficits with regard to social interaction and communication, and they often have limited areas of interests and stereotypical behaviors [17]. They have difficulty learning words due to socio-pragmatic deficiencies, including deficits in “theory of mind”, but they also have trouble analyzing words and concepts semantically [18]. There are, in addition, some studies showing that children with ASD perform as well as typically developed children on tests of semantic processing [19], [20].

Language is an instrument for conveying meaning and semantics is what language is all about [21]. Vocabulary is an important part of the language domain and includes both word knowledge and word retrieval ability. Lexical ability is referred to as receptive and expressive word use and the actual or interpreted meaning of words is defined as semantic ability in the current study. Lexical-semantic ability refers to the meaning and use of individual words in a person's vocabulary and the relation in between words.

It has been shown that some children with CI are able to close the gap and reach the same level as children with NH on vocabulary tests [22]. Other researchers have reported that children with CI do not reach the same levels as children with NH [23]. In a study of word fluency ability i.e. the ability to retrieve and generate words within a restricted time limit among children with CI and with NH aged 6–9 years, it was found that age was an important factor for the children in both groups [24]. Two word fluency tasks were used: (1) a phonemically based task (words beginning with the letters F, A, and S and (2) a semantically based task (words from the category “animals”). The 6–7 years old children with CI performed as well as age-matched NH children on both word fluency tasks. One group-specific difference was that the 8–9 years old children with CI retrieved significantly fewer words and used less efficient strategies in the retrieval process as compared to the 8–9 years old children with NH. This difference was especially pronounced on the phonemically based word fluency task. On the semantically based word fluency task there were no statistical differences between the two groups for this age.

We suggest that there are three important factors that might have a negative impact on word learning among children with CI. The first one is the impact of having no hearing or insufficient hearing during the initial stage of life, which lessens a person's ability to perceive spoken language as well as the ability to develop phonological and semantic representations during the pre-lexical period. Word learning has shown to be strongly correlated with age at implantation and phonological working memory in a study of 15 children with CI (5–11 years) [25]. The second factor is that children with CI develop their spoken language skills through artificial hearing, which might affect incidental word learning, especially in noisy real-world situations like classrooms [26]. In a study of word learning in noisy environments among school-aged children with NH it was found that noise made the learning of expressive words, in particular, more difficult [27]. The researchers claimed that the noise made it harder for the children to perceive new words, making it difficult for them to establish robust memory traces and store new words in their long-term memory. For children with CI this could potentially affect both receptive and expressive word learning. The third factor, which can affect spoken language development in children with CI, is the type and degree of spoken language stimulation received from caregivers [8], [9], [28]. Percy-Smith and colleagues [8] showed that parental use of an auditory-verbal approach in their communication resulted in statistically better auditory capabilities and spoken language outcome for their child. Similar result were shown in a recent study by Dettman et al. [9] comparing the influence of different communication modes in well-matched groups of children with CI. Szagun [28] showed that environmental factors and especially the quality use of spoken language like longer mean length utterances (MLU) used by the parents was the main factor that explained most of the observed variability in the sample, not age at implantation.

Children with LI and ASD are interesting populations to compare with children with CI, because they also experience an atypical neurodevelopmental process that might influence their development of phonological and semantic representations i.e. mental representations of speech sounds and underlying meaning of words affecting their vocabulary ability. There are different theories of how children acquire language, both in typical and atypical cases. The neuroconstructivistic theory has a developmental approach and spokesmen argue that genes, brain growth, cognition and environmental factors interact multi-directionally, both in typical and atypical cases [29]. The developmental process itself is in focus in this theory and this means that there are both similarities and differences that might be found in groups of children with conditions like for instance HI, LI and ASD. The origin of the neurodevelopmental processes are different but the consequences might, at least on the surface, overlap or resemble each other in abilities like receptive and expressive vocabulary. Furthermore, there are many children with HI who also have additional diagnoses like LI or ASD, which additionally motivates comparison studies of these clinical groups as a way of gaining knowledge of the specific phenotypes.

Fast mapping difficulties, as can result from having a less efficient phonological memory, have been found among both children with CI and children with LI [30], [31]. These difficulties are manifested as less efficient learning of new concepts after limited auditory exposure that affect word learning and the development of receptive vocabulary in a negative way [25]. In one of the few comparative studies of children with LI and children with mild-to-moderate HI it was found that both groups (5–10 years old) showed significantly worse results on tests of phonological awareness compared to age-matched controls with NH [32]. Worth noting, however, was that the children with HI had better semantic ability than the children with LI, despite their phonological deficits.

There are very few, if any, normative tests of semantic networking ability that are designed for typically developed children or clinical groups. One way of investigating depth and breadth in vocabulary, e.g., lexical-semantic networks, is to conduct more qualitative analyses, such as error response analysis, of normative vocabulary tests. This has been done both in clinical groups and in groups of typically developed children [33], [34]. Storms et al. [35] conducted a normative study of picture naming ability among children 6–12 years old using the Boston Naming Test (BNT) [36]. Their results showed that age and gender were significantly important factors for picture naming ability in the normal population. The age effect showed a linear increase from 6 to 12 years and boys performed better than girls. The results on the BNT also correlated very strongly with results regarding cognitive ability measured with Ravens Colored Progressive Matrices [37]. Storms et al. [35] also performed an error response analysis, which showed that the most common errors typically developed and NH children made were verbal semantic paraphasias and omitted responses, expressed as “I don’t know”. In a similar normative study by Brusewitz and Tallberg [34] of picture naming ability among typically developed Swedish children aged 6–15 years, it was found that Swedish children performed slightly lower on the BNT than the North American norms for children. When analyzing the error types made by the Swedish children, the researchers found that the younger children had significantly more omitted responses and used more unspecific responses than the older children when they were unable to name words.

In the present study we predicted that deaf children with CI would have a poorer vocabulary than age-matched children with NH [23], [24], [25]. The semantic knowledge of children with CI was predicted to be at the same level as typically developing children with NH [24] and better than that of children with LI and ASD, since previous studies have found specific semantic difficulties among the latter two groups [15], [16], [18]. For this reason, we also predicted that children with CI would use more relevant strategies and make less serious errors in comparison to children from the two clinical groups (LI and ASD). We expected that the groups of children with LI and ASD would perform differently from each other with regard to their lexical-semantic abilities due to their different neurodevelopmental profiles.

There has previously been little interest in investigating the lexical-semantic abilities of children with CI and especially in conducting comparative studies of other clinical groups displaying atypical neurodevelopmental and linguistic processes. The primary aim of the present study was to explore the lexical-semantic ability of children with CI aged 6–9 years in relation to age and non-verbal cognitive ability. Another aim was to compare lexical-semantic ability among the group of children with CI, two other clinical groups of children (with LI and ASD), and typically developed children with NH of the same ages. Finally, we wanted to investigate the importance of background factors like age at implantation, etiology and speech recognition level for lexical-semantic ability in children with CI.

Section snippets

Participants

The participants in the present comparative study comprised 97 (54 boys/43 girls) Swedish-speaking children from three different clinical groups and one control group. All of the participants were aged 5:6–9:0 years, including 34 children with CI, 12 children with LI, 12 children with ASD and 39 children with NH. Background information, including age, gender, and non-verbal cognitive ability for all four groups, is presented in Table 1. We found no significant differences between the four

Results

Descriptive group data for picture naming ability (BNT) and receptive lexico-semantic ability (PPVT III, Semantic Feature Test – Pictures and Semantic Feature Test – Questions) are presented in Table 2. Results of the error response analysis for the BNT for the four groups are presented in Table 3. The language tests are also presented one by one with statistical analyses and group comparisons. Finally, we have investigated the influence of background factors on expressive and receptive

Discussion

The results showed that there were significant differences between the groups for most of the expressive and receptive aspects of lexical-semantic ability. The most surprising outcome was that children with CI demonstrated age-equivalent expressive vocabulary despite lower levels of receptive vocabulary compared with NH peers. This pattern was not seen in the other two clinical groups. Early auditory deprivation and later artificial hearing lead to deficits in auditory and higher cognitive

Conclusions

The results from the present study indicate that children with CI aged 6–9 years, and especially those who were implanted within the early linguistic period of life, can develop lexical-semantic abilities that in general are comparable to age-matched children with NH. The group of children with CI had a normal semantic ability profile and picture naming ability but significantly poorer receptive vocabulary. The reasons behind this unusual finding are still unclear but environmental factors not

Acknowledgments

The authors wish to thank all of the children who participated and their parents for their contribution. The authors also want to thank Madelen Snickars, Viire Kask, Emma Bergström and Idah Lubowa-Mubiru for their help with recruitment and data collection. This research was supported by Karolinska Institutet, Sunnerdahls Handikappfond, Stingerfonden, Aina Börjesonfonden, and HEAD Graduate School (Linnaeus HEAD, Linköping University).

References (51)

  • K.A. Gordon et al.

    Use it or lose it? Lessons learned from the developing brains of children who are deaf and use cochlear implants to hear

    Brain Topogr.

    (2011)
  • L. Percy-Smith et al.

    Parental mode of communication is essential for speech and language outcomes in cochlear implanted children

    Acta Otolaryngol.

    (2010)
  • S. Dettman et al.

    Communication outcomes for groups of children using cochlear implants enrolled in auditory-verbal, aural-oral, and bilingual–bicultural early intervention programs

    Otol. Neurotol.

    (2013)
  • J. Niparko et al.

    Spoken language development in children following cochlear implantation

    JAMA

    (2010)
  • A. Geers

    Factors influencing spoken language outcomes in children following early cochlear implantation

    Adv. Otorhinolaryngol.

    (2006)
  • S. Gray et al.

    The diagnostic accuracy of four vocabulary tests administered to preschool children

    Lang. Speech Hear. Serv. Sch.

    (1999)
  • U. Nettelbladt et al.

    Språkutveckling och språkstörning hos barn

    (2007)
  • M. Alt et al.

    Factors that influence lexical and semantic fast mapping of young children with specific language impairment

    J. Speech Lang. Hear. Res.

    (2006)
  • L. Sheng et al.

    Lexical-semantic organization in children with specific language impairment

    J. Speech Lang. Hear. Res.

    (2010)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorders

    (2000)
  • C. Gillberg et al.

    Autism – Medicinska och pedagogiska aspekter

    (2001)
  • L. Mottron et al.

    A study of memory functioning in individuals with autism

    J. Child Psychol. Psychiatr.

    (2001)
  • M. Toichi et al.

    Long-term memory in high functioning autism: controversy on episodic memory in autism reconsidered

    J. Autism Dev. Disord.

    (2003)
  • A. Wierzbicka

    Semantics: Primes and Universals

    (1996)
  • H. Hayes et al.

    Receptive vocabulary development in deaf children with cochlear implants: achievement in an intensive auditory-oral education setting

    Ear Hear.

    (2009)
  • Cited by (0)

    View full text