Language Profiles and Outcomes in School-Age Children Diagnosed with Autism Spectrum Disorder as Toddlers

Project Summary/Abstract Late language emergence occurs in approximately 10-20% of children. There is significant variability in language outcomes among late-talking children, further complicated by the heterogeneity of this population, including children with neurodevelopmental disorders such as Autism Spectrum Disorder (ASD). Our group recently found that ASD diagnoses do not persist at school age for over one-third of children diagnosed as toddlers (termed “persistent ASD” and “non-persistent ASD”). Gaps remain in our knowledge about language outcomes and their predictors in children diagnosed with ASD at a young age, particularly those who no longer fulfill ASD criteria. Identifying predictors for late-talking children most at risk for persistent language impairment would strengthen initial counseling recommendations regarding intervention and prognosis. Understanding domains of language impacted in children with persistent and non-persistent ASD could support clinicians in monitoring specific language domains, such as language syntax and morphology, during developmental follow- up visits for children initially diagnosed with ASD. Thus, we will leverage a unique extant database to characterize language profiles and outcomes in 213 children diagnosed with ASD as toddlers (<36 months) and re-evaluated at school-age (5-7 years). Our objectives are to: 1) Examine language outcomes in school-age children diagnosed with ASD as toddlers and identify predictors of language outcomes in these children, and 2) Compare aspects of language as a function of ASD status (persistent vs. non-persistent) in 5-7 year old children diagnosed with ASD as toddlers. Our methods will include a latent class analysis based on available data in the extant BOAT database to identify groups of children based on language function. Latent classes will be formed by variables including functional communication (Vineland-3), expressive and receptive language (PLS-5), verbal IQ (DAS), and speech (parent report of speech disorder). We will also analyze predictors of overall class membership, including baseline cognition, language, and adaptive skills abstracted from an initial clinical evaluation and persistence of ASD. Finally, we will complete subdomain-level analyses of PLS-5 protocols and language sampling analysis using Systematic Analysis of Language Transcripts (SALT) to characterize individual aspects of language/ communication in children who continue to meet diagnostic criteria for ASD vs. those who do not. This project will address an unmet clinical need to better identify predictors of language outcomes in late- talking children diagnosed with ASD, characterize language profiles with a focus on functional impairment, and assess differences in language in children who continue to meet ASD criteria vs. those who do not.