Methods for Assessing Language in School-Age Children with Autism and Intellectual Disability

Background: The language phenotype of autism is characterized by delays in expressive language vocabulary, pragmatics, and grammar. This phenotype is closely tied to general language comprehension, and has been compared to other disorders like Specific Language Impairment and disorders with similar behavioral phenotypes (e.g., fragile X). Little has been done on the best method for assessing language in children with autism with an intellectual disability.  Research in fragile X has found children perform at nonverbal mental age expectations on standardized tests, but below expectations on a language sample. Language samples require social engagement, and can be particularly challenging for individuals with poor pragmatic skills or those with high levels of social anxiety, both found in autism. 

Objectives: To examine the best assessment method for language in children with autism who also have an intellectual disability using a variety of methods commonly used in both the research and clinical domains. We included a comparison group of nonverbal mental age matched children with fragile X syndrome.

Methods: Seven boys with autism have completed the study (study is ongoing, anticipated number of participants 26), as well as 14 boys with fragile X syndrome between the ages of 9-16 years. Assessments included standardized language assessments: receptive and expressive vocabulary (PPVT and EVT), receptive and expressive grammar (i.e., TEGI, CELF), as well as a nonverbal IQ test (Leiter brief IQ), a language sample, and a sentence imitation task. Additionally each participant completed the Autism Diagnostic Observation Schedule (ADOS) and parents completed the Autism Diagnostic Interview-Revised (ADI-R). The Childhood Autism Rating Scale was completed post-assessment. The language samples, sentence imitation task, and ADOS were transcribed using standard language transcription procedures, and analyzed using the Systematic Analysis of Language Transcripts (SALT; Miller & Chapman, 2000). Transcripts were analyzed for standard language measures including number of utterances, number of words, grammatical complexity (MLU), and measures of dysfluencies (e.g., repetitive speech, incomplete sentences).

Results: Preliminary results indicate the boys with autism show a relative strength on the sentence imitation task (100% accuracy) and standardized assessments, but lower language levels on the conversation sample (mean MLU: 4.47). The boys with FXS demonstrated more impairment on the sentence imitation task (75% accuracy) and in the language sample (FXS MLU: 2.62).

Conclusions: The boys with autism demonstrated a relative strength on both the sentence imitation task as well as the standardized tests. The language samples yielded less complex language compared to what would be expected based on standardized test performance, but more complexity compared to FXS. We will compare the language used in the language samples to the ADOS in order to look at contextual differences. The boys with FXS struggled with the sentence imitation task, but like autism, demonstrated a relative strength on the standardized assessment. While language samples are an important part of both clinical practice and research, preliminary evidence suggests they are not as informative in terms of the true picture of strengths and weaknesses for some aspects of language, in this case grammar.