Say What?: Toddlers' Vocabulary Growth Trajectories Differ By Word Features

Background: Infants and toddlers later diagnosed with ASD often experience significant language delays, and concern about language is one of the first reasons parents seek evaluation. It is unknown whether these delays affect certain vocabulary types more than others, which might lead providers to target specific word types for intervention. Recent research suggests that words labeling concrete and image-able referents like ‘cup’ are easier for typical children to learn, whereas words for abstract concepts like ‘later’ are not learned as readily. Given prior research suggesting that children with ASD tend towards a concrete and perceptually driven worldview, we hypothesize that difficulty learning abstract words may partially account for previously observed overall depressions in language development early in life in ASD.

Objectives: To test whether vocabulary growth from 12-24 months differs according to word abstractness in high-risk siblings that went on to develop ASD, and high-risk and low-risk siblings that did not develop ASD.

Methods: Our sample included 365 high-risk (HR) and low-risk (LR) children with the MacArthur-Bates Communicative Development Inventory (CDI; Words and Gestures) completed by a parent at 12, 18, and/or 24 months. Children were assessed for ASD at 24 months using ADOS and DSM criteria, and categorized as HR-ASD (n=57), HR-non-ASD (n=203), and LR-non-ASD (n=105). Words from the CDI were grouped into “concrete” (e.g., words for food and drink; toys) or “abstract” (e.g., words for time; pronouns) categories. We calculated total words understood as a measure of receptive vocabulary and total words understood+said as a measure of expressive vocabulary. Linear mixed models were used to compare growth trajectories.

Results: There was a main effect of Group in overall receptive vocabulary from 12-24 months, F=3.62, p=.03, LR-non-ASD>HR-non-ASD>HR-ASD, and no GroupxTime interaction. There was no main effect of Group in overall expressive vocabulary, but there was a GroupxTime interaction, F=6.03, p=.003, LR-non-ASD>HR-non-ASD>HR-ASD. Similarly, concrete expressive vocabulary (e.g., toys) did not differ by Group, and there was no GroupxTime interaction, all ps=n.s. (Figure 1). In contrast, growth rates in abstract expressive vocabulary (e.g., words for time) differed by Group, F=10.15, p<.001, and there was a GroupxTime interaction, F=41.77, p<.001, LR-non-ASD>HR-non-ASD>HR-ASD (Figure 2).

Conclusions: This pattern of results suggests that concrete words are spared in the expressive vocabularies of HR-ASD, while abstract word vocabularies are impaired. Furthermore, abstract word learning may be difficult even for high-risk siblings that do not meet diagnostic criteria for autism. On-going analyses include parsing concrete vocabulary into social and non-social categories, including effects of mental age at each time point, correlating semantic category growth with autism symptoms, matching on overall vocabulary to account for this effect, and examining possible language disorders in the high-risk sibling group.