Abstract
Background: Fragile X syndrome (FXS) is the most common inherited cause of intellectual disability, and the most common single gene disorder associated with autism. Language impairments in this disorder are well documented, but the nature and extent of syntactic impairments are still unclear.
Aims: To compare the performance of boys with FXS with and without autism spectrum disorder on measures of verb (VM) and noun (NM) morphosyntax with that of typically developing boys of similar non-verbal mental ages.
Methods & Procedures: Conversational samples were obtained from 33 boys with FXS with autism spectrum disorder (FXS-ASD), 35 boys with FXS and no ASD (FXS-O), and 46 typically developing boys (TD). Production of verbal and nominal morphosyntax was assessed separately in these two subdomains. A hierarchical linear model compared morphosyntactic scores in all groups after adjusting for non-verbal cognition, articulatory skill, and caregiver education. The model also tested interactions between group and morphosyntactic subdomain.
Outcomes & Results: Boys with FXS in both groups scored lower than the TD boys on both measures. The FXS-O and the FXS-ASD groups did not differ on either composite measure. All covariates were significantly related to morphosyntactic scores.
Conclusions & Implications: Part of the morphosyntactic impairment in FXS may be attributable to cognitive, environmental, and speech factors. However, it is clear that boys with FXS perform at levels lower than expected from differences in these extra-linguistic factors alone, across both the verb and the noun domains. Clinical interventions should therefore seek to address specific syntactic targets.
Acknowledgements
This research was supported by the National Institute of Child Health and Human Development (Grant Numbers 1 R01 HD038819, 1 R01 HD044935, 1 R03 HD40640, and 5 T32 HD40127). The authors wish to thank Jan Misenheimer for help with data entry; and Kristin Cooley, Anne Harris, Cheryl Malkin, and Sabrina Smiley for assistance with data verification and reliability. The authors also thank Len Abbeduto, Margaret Burchinal, Patricia Cleave, Marc Fey, Gary Martin, Lise Menn, Mabel Rice, Ivan Sag, and Melanie Schuele for comments and suggestions.
Notes
1. To be classified in the FXS-ASD group, each child had to meet the autism spectrum cut-off score for each subtotal: Communication and Social, and the autism spectrum cut-off for the Communication plus Social total. For Module 1, communication cut-off is 2 or higher, social interaction is 4 or higher, and total is 7 or higher. For Module 2, the respective cut-offs are 3, 4, and 8. For Module 3, they are 2, 4, and 7.
2. All primary caregivers were mothers in the sample. See below.
3. All children included had at least 100 utterances. The number of utterances analysed was capped at 100 because (1) equal sample lengths minimize the likelihood of finding spurious differences for token measures of language, and (2) many children did not have many more usable utterances. We believe this is not problematic since it is customary to do so for other measures of syntax.
4. For Module 1, most utterances were produced during free play, a pretend birthday party, a snack break, and interactions using bubbles and a balloon. For Modules 2 or 3, utterances came from make-believe play with toys, a joint interactive play, a book activity, a pretend birthday party, a snack, and bubbles and/or balloon interactions.
5. We did not include the two aspectual markers progressive -ing and perfect (generally -en), because they are usually not included in verb composites (or in tense composites, for obvious reasons).
6. For example, assuming the true percentage being estimated is 50%, with three OCs (the minimum required by CitationEadie et al. 2002) the 90% confidence level interval ranges from 24% to 76%.
7. Simple sentences with transitive verbs, for example, have two noun phrases for one verb phrase. Also, noun phrases can occur in isolation in utterances, but verbs alone are rarer.