Spontaneous Expressive Language Impairments in Clinic Referred Children for a Possible ASD

Background:  Although impairments in spontaneous expressive language are prevalent in ASD and other communication disorders, standardized language measures currently available do not adequately capture these deficits, impeding our attempt to clearly quantify language profiles that are necessary for provisions of treatments. A newly developed language measure, the Observation of Spontaneous Expressive Language (OSEL) is a standardized, semi-structured clinician observation, which allows an assessment in children aged 2-12 years targeting: 1) Syntax, and 2) Pragmatic-Sematic skills under three domains: Initiation of Reciprocal Communication, Narrative Skills, and Unusual Features (Table 1).

Objectives:  To examine spontaneous expressive language profiles during a standardized but natural, play-based setting targeting syntax (structure of language), semantics and pragmatics (social use of language) in clinic-referred children for a possible ASD. 

Methods: Total of 66 children (57 ASD cases) were administered the OSEL (Mean Age=5, SD=2.3). The proportion of children showing delays were examined and the severity of delays were quantified based on language quotients ([age equivalent/chronological age]*100). The associations between the OSEL and autism symptom severity (Autism Diagnostic Observation Schedule [ADOS]) and standardized language scores (Comprehensive Assessment of Spoken Language [CASL] or Preschool Language Scale [PLS]) were examined using Pearson’s Correlation.

Results:  Out of 66 children, 38 (58%; 32 ASD cases) had their syntax language quotients less than 70. Out of 66 children, 55 (83%; 47 ASD) and 48 (72%; 41 ASD) had their language quotients less than 70 for Initiation of Reciprocal Communication and for Narrative Skills respectively. Notable gaps between chronological age and age equivalents (AE) were observed for all domains (e.g., mean Syntax AE=2.8 years (SD=0.7); mean chronological age=5.7 (SD=2.2); mean language quotient=53.6 [SD=16.6]). When verbal IQ and other language measures were examined, only 16-44% of all children who showed significant delays on the OSEL were identified as having delays (scores<2 SDs; Figure 1). A mild correlation was observed between the OSEL Initiation of Reciprocal Communication and ADOS Social Affect severity scores (r=-0.34, p<0.05). Moderate to strong correlations between OSEL and other language measures were observed (r’s 0.3-0.8; p’s<0.05).

Conclusions:  Based on a measure of spontaneous expressive language, more than 80% of clinic-referred children for ASD showed significant delays in pragmatic language. Most children (90%) received a diagnosis of ASD, and their language scores were significantly associated with autism symptom severity. This may explain why these children demonstrated notable difficulties in initiating back-and-forth communication with a social partner and providing accounts of events, which is one of the hallmarks of ASD. About 60% of children also showed notable delays in syntax, suggesting that intervention may need to be tailored to target deficits in grammar in a subset of these children. The results also suggest that difficulties in spontaneous expressive language may be missed by other instruments measuring pre-determined answers that are highly tied to concepts. Moderate correlations between the OSEL and other measures also suggest that the assessment of language in a standardized but natural context has an additive value and should be used complementarily with other measures.