19702
Growth Trajectories of Longitudinal Naturalistic Verb Use in ASD: Verb Category Matters
Language is an important source of clinical heterogeneity in autism spectrum disorder (ASD), and delays in this area often motivate parents to seek their first evaluations. Although most children with ASD acquire at least some vocabulary as measured by parent report and standardized tests, language use in naturalistic settings remains minimally understood. Moreover, verb use is particularly understudied: Children with ASD may not demonstrate an overall verb deficit according to traditional parent report measures and standardized tests, but certain types of verbs may be selectively avoided (e.g., mental) or relied on (e.g., action) in actual language use.
Objectives:
Compare patterns of growth in action verbs, general all-purpose (GAP) verbs, and mental verbs in a naturalistic, longitudinal study of typically developing (TD) toddlers and toddlers with ASD.
Methods:
Thirty-three toddlers with ASD were matched to 35 TD toddlers on Expressive and Receptive Language abilities at visit 1 of 6. Children with ASD were classified as high verbal ability (HV; N=14) or low verbal ability (LV; N=18) at visit 6. Language produced by parents and children during six 30-minute semi-structured home-based play sessions (4 months apart) was recorded and transcribed in CHAT format. Verbs produced by children were categorized as action (e.g., run, dance), GAP (e.g., give, want), mental (e.g., think, know). Linear mixed models compared growth trajectories of total verb vocabulary, as well as action, GAP, and mental verbs as raw scores and proportions of total vocabulary in the three groups. Planned comparisons assessed group differences at each time point.
Results:
Growth analyses of total verb vocabulary (tokens and types) showed main effects of visit, group, and group by visit, TD=HV>LV, all ps<.05. Action verbs (raw types and tokens) showed a similar pattern, TD=HV>LV, all ps<.01. In contrast, growth trajectories of action verb proportions (types and tokens) did not differ by group or by group over time, all ps=n.s (Figure 1). Growth patterns for GAP verbs (raw types and tokens) were similar to action verbs, TD=HV>LV, all ps<.05. Trajectories of growth in proportions of GAP verbs revealed a distinct pattern of group effects; TD=HV<LV, with LV greater than either TD or HV (Figure 2). Raw growth in mental verbs and proportion of types/tokens that were mental verbs was highly variable in the HV and LV ASD groups, TD>HV>LV.
Conclusions:
This pattern of results suggests that children with ASD can be meaningfully stratified into low- or high-verbal groups, and that patterns of naturalistic verb use differ by language ability and verb subtype. Action verbs are most robust to language delays, with similar growth in all groups relative to total verb vocabulary. In contrast, LV children, but not HV or TD children, may rely more on GAP verbs than specific semantically rich verbs. Mental verb growth trajectories differed in both the HV and LV groups relative to TD, largely by due to increased variability in children with ASD. These findings extend prior research into a new context (naturalistic) and design (longitudinal), and specify fine-grained patterns of verb use in children with ASD.