The Impact of Bilingual Exposure on the Expressive Language of Children with Autism Spectrum Disorders

Background: Many parents and clinicians assume that early bilingual exposure negatively impacts language development, especially for children with ASDs.  This belief persists despite the fact that studies of bilingualism in normal and language-impaired children have not found differences in the ages at which early language milestones are reached or in total spoken vocabulary size between children from monolingual and bilingual environments.

Objectives: To compare major early language milestones and current spoken vocabulary size in children with ASDs from monolingual and bilingual environments. 

Methods: Families with a child aged 18 months to 6 years of age diagnosed with an ASD were recruited from Quebec and Ontario (target N=110).  Phone interviews were administered to families and included a detailed caregiver language history, questions from the Autism Diagnostic Interview-Revised, and the Vineland Adaptive Behavior Scales-II; these measures generated an estimate of their child’s language exposure and also described developmental history and current general function.  Families also completed questionnaires including the MacArthur Communicative Development Inventories (Words and Sentences; available in 10 languages).  Expressive vocabulary size for children from bilingual environments was calculated both in their dominant language and across languages by summing concepts labeled by a word in one or both languages (total conceptual vocabulary). 

Results: Children were first classified into monolingual (MON) or bilingual (BIL) groups based on exposure to one or two languages from birth to age 2, with a significant difference in estimated language exposure between the groups (MON mean = 100%, BIL = 64%; SD 20%, p=<.01).  Groups were comparable in age (MON=57 months, BIL=58) and demographic variables (parental education, family income).  Data analysis (MON N=27, BIL N=20) revealed no significant differences between the groups’ mean age in months at first words and first phrases.  Family concern regarding a child’s development occurred at the same mean age, but the age at which the possibility of an ASD was first discussed with a professional differed significantly, with families from bilingual environments reporting the first discussion 6 months later than monolingual families (31.5 vs. 25.1 months; p <.01, 95% CI 1.4-11.3).  A second analysis compared current expressive vocabulary size.  Children were re-grouped based on lifetime exposure to a second language, with >10% average lifetime exposure required to be considered bilingual; no significant differences in age (57 vs. 58 months) or demographic variables were noted.  The groups did not differ significantly on either their vocabulary in their dominant language (MON=369 vs. BIL=394 words spoken) or in their total conceptual vocabularies (369 vs. 429 words). 

Conclusions: These data do not support a significant relationship between bilingual environments and additional language delay in children with ASDs.  The data, however, do suggest that parents and/or professionals may assume that bilingual environments contribute to a language delay, and thus do not consider an ASD diagnosis even when early concerns are present.