Examining Expressive Language Benchmarks in Young Minimally Verbal Children

Background:  Despite accessing early intervention, up to 30% of children diagnosed with ASD are classified as nonverbal or minimally verbal at age five (Anderson, 2007). Within the minimally verbal population of children with ASD, there is high variability in both speech and language abilities, without a single identified pattern of characteristics (Tager-Flusberg & Kasari, 2013). Identifying children in this population according to characteristics of their development has important implications for guiding intervention research.

Objectives: The purpose of this study was to classify young minimally verbal children according to expressive language benchmarks proposed by Tager-Flusberg et al. (2009), and to examine the relationship between expressive language measures and autism symptom severity.

Methods: Data were analyzed from pre-intervention assessments of participants in an ongoing intervention study (HRSA grant #R40MC27707). Fifty-three minimally verbal children with autism between 36 and 54 months of age (mean age=32.6 months) participated in this study. A measure of phonology (consonant inventory) was derived from the Profiles of Early Expressive Phonology (Williams & Stoel-Gammon, 2014). Naturalistic language samples were transcribed and coded to derive measures of vocabulary (number of different words) and pragmatics (number of different communicative functions). The percentage of children in the sample who met benchmarks for each expressive language stage (i.e., pre-verbal, first words, word combinations, sentences) was then calculated for each domain (phonology, vocabulary, pragmatics). Additionally, two separate regression models examined the relationship between phonology and vocabulary with autism symptom severity as measured by the Autism Diagnostic Observation Schedule (ADOS; Lord, Rutter, Dilavore, & Risi, 2008).

Results: For the domain of phonology, 43.3% of the sample met criteria for pre-verbal status (i.e., 6-12 months), 13.6% of the sample met criteria first words (12-18 months), 13.6% of the sample met criteria for word combinations (18-30 months), and 29.6% of the sample met criteria for sentences (i.e., 30-48 months). For the domain of vocabulary, 41.5% of the sample met criteria for pre-verbal status, 58.5% of the sample met criteria first words, and none of the participants met criteria for word combinations or sentences. For the domain of pragmatics, 57% of the sample met criteria for pre-verbal status, 43% of the sample met criteria first words, and none of the participants met criteria for word combinations or sentences. Separate regression analyses indicated that vocabulary and phonology were significantly related to ADOS scores, controlling for IQ (vocabulary: Beta=-.167 p=.049; phonology: Beta=-.169 p=.007).

Conclusions:  While the 53 children were dispersed across all developmental phases for phonology, most children remained in early phases of vocabulary and pragmatic development. Although about half of the participants produced speech sounds at the word combination and sentence level, none of the participants met criteria for language use at higher levels. These results are reinforced by findings from the regression analyses, indicating that, on average, children with higher autism symptomology were producing fewer different words and fewer speech sounds. Future research should focus on individualizing interventions to better target specific language weaknesses highlighted in these language profiles for individuals with ASD.