Factors Related to the Development of Bilingual Vocabularies in Young Children with Autism Spectrum Disorders

Background: Two studies (Hambly & Fombonne, 2011; Leadbitter, Hudry, & the Preschool Autism Communication Trial consortium, 2009) have found that children with Autism Spectrum Disorders (ASD) do not experience additional language delays when exposed to two languages. A third study has shown that the language of bilingual children with ASDs is not delayed compared to matched monolinguals with ASD (Petersen, Marinova-Todd, & Mirenda, 2011). The factors associated with becoming bilingual during early childhood have never been studied in children with ASDs.

Objectives: Identify factors related to the development of bilingual language in young children with ASDs.

Methods: 33 children (mean age=60 months) with ASD, bilingual exposure history, and a minimum vocabulary size of 50 words in their dominant spoken language (L1) were studied using parent report data. 10 children had no words in a second language (NO-B). The remaining 23 children were divided using the median L2 vocabulary size (69 words) on the MacArthur-Bates Communication Development Inventories (MBCDI, various languages) to assign children to the low vs. high bilingual vocabulary groups: 11 children had L2 vocabularies from 2-69 words (LOW-B) and 12 children had  70 or more words (HIGH-B). The three groups were compared using one-way Analysis of Variance to look for differences in exposure history, dominant language abilities, social abilities, and family demographics. Differences with a p value of <.10 were explored further with post-hoc testing. The correlations between L2 vocabulary and key factors were explored in the 23 children with bilingual vocabularies to look for relationships between exposure and language variables and L2 vocabulary size.

Results: Recent exposure history (p=.09) was significantly different only between the LOW-B (6% L2 exposure) and the HIGH-B (27% L2 exposure) groups (p=.04). Statistically significant differences were also present on language measures: HIGH-B scores were higher than both the LOW-B and NO-B scores on the Vineland Adaptive Behavior Scale-II (VABS-II) Expressive score (p=.01) and the MBCDI (p=.01). Data from the 23 bilingual vocabulary users showed strong correlations between higher L2 vocabularies and 1) less exposure to L1 (r= -.83, p=.000),  2) higher VABS-II Expressive score (r=.60, p=.005), and 3) less-impaired Social Responsiveness Scale scores (r= -.46, p=.031).

Conclusions: High levels of current L2 exposure (e.g., >25% exposure to an L2) combined with more advanced L1 language (e.g., large L1 vocabularies and expressive language scores in the high average range) are likely factors in the development of larger bilingual vocabularies. It is not clear from this data why some children with large amounts of exposure did not develop L2 vocabularies, but it is possible that differences in the type or quality of exposure (e.g., L2 at home vs. at daycare/school, the presence or absence of indirect L2 exposure in the home) or in the approach to teaching an L2 could explain why some children start acquiring small bilingual vocabularies while others do not.