Word Mapping Using Lexical Stress in 12-Month-Old Infant Siblings of Children with Autism and Its Relationship with Early Language Development

Background: Typically-developing infants are highly sensitive to the rhythmic patterns of their language (Mehler, et al., 1988). This sensitivity may be directly linked to the ability to detect words in long speech streams and the learning of new words (Christophe, Mehler, & Sebastian-Galles, 2001; Curtin, et al., 2005). Indeed English-learning 12-month-olds can map trisyllabic word-object pairings with labels that differ solely in stress (STRONG-weak-weak vs. weak-STRONG-weak); Curtin, 2009). Less is known, however, about the atypical processing of word-stress and how this could affect early word learning. For children with Autism Spectrum Disorder (ASD), the most commonly affected aspect of speech is the production of stress, rhythm, and intonation (McCann & Peppe, 2003). Given that approximately 19% of infant siblings of children with ASD (SIBS-A) will also be diagnosed (Ozonoff et al., 2011), and that an additional 15-20% will develop language impairments (Constantino, Zhang, Frazier, Abbacchi, & Law, 2010), this population offers a unique opportunity to study possible atypical processing of word-stress and its impact on language development.

Objectives: We examined whether 12-month old SIBS-A (n=16) learn labels for objects (when those labels differ solely in word-stress), in the same way that typically-developing (SIBS-TD; n=18) 12-month old siblings do. We also examined whether this early ability was related to word comprehension at 12-months.

Methods: Infants were tested using the ‘Switch task’ (see Werker, et al., 1998) where they were habituated to the trisyllabic word forms ‘BEdoka’ and ‘beDOka’ (STRONG-stress denoted by capital letters) paired with novel objects that were presented on a central monitor. At test, one of the word-object pairings was mis-matched. Longer looking times to this ‘switch’ indicated successful mapping of the labels to the novel objects and the ability to discern minimally different word-forms. To assess early language, parents completed the MacArthur-Bates Communicative Development Inventory (MB-CDI) (Fenson, et al., 1993).

Results: SIBS-TD looked longer to the ‘switch’, which was marginally significant (p=.07). SIBS-A infants also looked longer to the ‘switch’, which was significant (p<.05). In terms of MB-CDI scores at 12-months, SIBS-TD understood significantly more words (SIBS-TD: M=103.72 vs. SIBS-A: M=51.4; p < .05). A Pearson r correlation coefficient, revealed that SIBS-TD infants who looked longer during the ‘switch’ test trial were also more likely to understand more words at 12-months (r=.53, p<.05). For SIBS-A, there was no significant relationship between success in correctly identifying the ‘switch’ with the number of words understood at 12-months (p>.05).

Conclusions: Preliminary results indicate that both SIBS-TD and SIBS-A map word-object pairings that differ solely in stress at 12-months. Importantly, this ability was also related to early word comprehension; however, this was only true for SIBS-TD. The ability of SIBS-A to use lexical stress to map labels to objects was not obviously related to word comprehension skills at 12-months. These findings suggest that although SIBS-A may attend to lexical stress like SIBS-TD, this attention is not necessarily related to advanced language comprehension skills at 12-months as it is in typical development. A divergent path of language development for SIBS-A is therefore likely.