Lexical and semantic development in children with cochlear implants

Sammanfattning: Introduction: Two important events have changed the prerequisites for children with severe-profound hearing impairment (HI). The first is the universal newborn hearing screening (UNHS) that enables early detection and identification of congenital or early- acquired HI. The second factor is the possibility of cochlear implant (CI) intervention at younger ages. There has been little previous interest in studying lexical-semantic ability in the new generation of children with CI who are implanted at a fairly young age, often bilaterally, in relation to the cognitive capacities that influence on use and knowledge of words. Aim: The general aim was to explore lexical-semantic ability and development in a group of school-aged children with CI and in comparison with age-matched normal hearing (NH) children. Method: The four studies examined different aspects of lexical-semantic ability in children aged 6-9 years. The cohort consisted of 34 children with CI and 39 age- matched children with NH (Study I-IV). In addition, two other clinical groups: children with language impairment (LI) (n=12) and children with autism spectrum disorder (ASD) (n=12) participated in Study II. A broad test battery was used, especially examining lexical-semantic ability but also non-verbal cognitive ability as well as phonological and learning ability. An error response analysis was conducted on a picture-naming test and a linguistic cluster analysis was performed on two different word fluency tasks with the purpose of examining strategies that the participants used when they retrieved words from their long-term-memory (LTM). Study IV had a partly longitudinal approach where a subgroup of children with CI (n=18) were examined more closely over time and in comparison with other age-matched groups at the age of 8-9 years. Statistical analyses were made primarily to examine group differences. Results: Many children with CI reached age-equivalent lexical-semantic ability at the age of 8-9 years. Semantic knowledge and non-verbal cognitive ability predicted grammatical sentence understanding (GSU) in both groups (CI and NH). Children with CI used age-appropriate learning strategies and had similar cognitive capacities necessary for managing use and knowledge of words as NH controls. In addition, children with CI showed better outcome than children with LI or ASD. However, an atypical developmental pattern was found in Study II where children with CI had better expressive than receptive vocabulary compared to typically developed children with NH. Also, the variation of the outcome was overall greater in children with CI (Study I- IV). An age-related CI subgroup-pattern was found in Study I that lead to the planning of a follow-up study, also examining retrospective, early spoken language development (Study IV). Children in the subgroup with higher group mean age at 1st CI had significantly poorer results on receptive vocabulary, phonemically based letter word fluency and GSU than NH controls. Despite of these results, the whole sample of children with CI had an age-appropriate level of expressive vocabulary and semantic feature knowledge. Semantic knowledge was demonstrated as an adequate ability to recognize semantic features and to use semantically relevant responses when lacking the lexical term while naming pictures (Study III). Children with an younger group mean age at 1 CI had better early expressive language use, two years after CI-operation, than the other subgroup, and they were also able to catch up after school entry on receptive vocabulary and phonemically based letter word fluency ability (Study IV). Conclusions: Children with CI and typical non-verbal cognitive ability did not have specific deficiencies in cognitive processing of lexical-semantic items, but some children had deficiencies with phonological and lexical-semantic knowledge. The results indicate that age at 1st CI is important as a starter engine of spoken language development through listening, but that other more linguistic-related factors and strategies also are important for development of lexical-semantic ability. There was a greater variation of the spoken language outcome in children with CI, and some atypical developmental patterns were found in the sample. Future studies of lexical-semantic ability in children with CI should therefore have a longitudinal approach and also explore the influence of environmental factors.