Narratives Abilities In Optimal Outcome Children with a History of Autism Spectrum Disorders

Background: A study is currently following children who have a history of autism spectrum disorder (ASD), but who no longer meet diagnostic criteria for such a disorder. These children have achieved social and language skills within the average range for their ages and receive little or no school support. Several recent studies suggest that this small subset of children, once diagnosed with ASDs, achieve "optimal outcomes" (Sutera et al., 2007; Helt et al., 2008; Kelley, Naigles, & Fein, 2010).

Objectives: The purpose of the current study was to analyze the narrative ability of individuals who have achieved “optimal outcome” (OO) and compare it to the narrative ability of individuals with high-functioning autism (HFA) and typically developing (TD) children to investigate whether OO individuals continue to maintain more subtle pragmatic language deficits.

Methods: The “Tuesday” narrative from the Autism Diagnostic Observation Scale (ADOS) was collected from 45 participants (n = 15 per group), who were matched on age gender, and VIQ [M (SD)= 111.9 (16.4), 105.9 (15.9), and 114.0 (12.4) for OO, HFA, and TD, respectively, p = .26]. Results from n = 15 (mean age= 13.3, range=12.2-14.6; VIQ [M (SD)= 107.2 (18.6), 104.4 (19.1), and 109.6 (9.56) for OO, HFA, and TD respectively, p=.88]), have been transcribed and analyzed by coders naive to diagnosis. Based on prior research (Kelley et al., 2006; Capps et al., 2000) a set of narrative qualities were targeted for analysis:  length of narrative (i.e., mean number of words, utterances); lexical richness (i.e., number of different adjectives, and type-token ratio); production of emotion words and mentalizing terms; stereotyped or idiosyncratic use of language; and dysfluency (i.e., false starts, repetitions, and self-corrections).

Results: The three groups produced narratives of similar length [M(SD) = 26.2 (7.1), 27.0(4.8), 29.0 (8.4) for OO, HFA, and TD, respectively; p = .81]. Controlling for total number of words in the narrative (which differed considerably across individuals, though not across groups), there were striking differences in fluency. Specifically, the HFA group produced significantly more dysfluencies [M(SD) = 17(11)] than either the OO or TD groups [M(SD) = 4.0(4.2) and 4.3(4.3), respectively]. Because of the preliminary sample size, potential trends on other narrative characteristics did not reach significance.

Conclusions: Because pragmatic and discourse features of language are among the most universal and clinically impairing in individuals with HFA, it is important to establish whether these abilities continue to show subtle impairments in individuals with optimal outcomes. Preliminary results indicate that adolescents with HFA produce narrations that are characterized by significant dysfluency; in contrast, individuals with an optimal outcome are indistinguishable from their TD peers. Ongoing analysis with a larger sample will establish the robustness of this finding. Furthermore, analyses will continue to explore whether OO, HFA, and TD individuals exhibit differences in other narrative qualities. In addition to expanding the sample size, analyses will examine the degree to which pragmatic and discourse language abilities (e.g., narrative qualities) are distinct from grammatical and structural language skills (e.g., CELF scores).