Early Communication Correlates of Language Development and Autism Symptomatology in Toddlers at-Risk for ASD

Background: Early nonverbal behaviors, such as initiating and responding to shared attention and requesting objects or actions, show associations with verbal language and autism symptomatology. Understanding the relationship between nonverbal social communication, verbal language, and autism symptomatology has implications for assessing early interventions in toddlers exhibiting symptoms of autism spectrum disorders (ASD).

Objectives: Investigate the relationship between nonverbal and verbal social communication and autism symptomatology in toddlers at-risk for ASD.

Methods: Forty-nine toddlers (mean CA 21 months, range 15-30 months; 8 females) who met a predetermined cutoff on the Screening Tool for Autism in Two-Year-Olds (STAT) and had a clinical presentation consistent with an ASD were enrolled in A Multi-Site Clinical Randomized Trial of the Hanen More than Words Intervention. Toddlers were administered the Early Social Communication Scales (ESCS) and Mullen Scales of Early Learning (MSEL) as part of the initial assessment. The ESCS was coded (Mundy et al., 2003; Yoder & Fey, 2008) to capture toddler-initiated joint attention (IJA), response to joint attention (RJA), toddler-initiated behavioral requesting (IBR), and weighted frequency of infant communication (WFC), a measure of intentional gestural and verbal communication. The MSEL language subscales yielded raw score measures of expressive (MSEL-EL) and receptive (MSEL-RL) language abilities; the STAT indexed autism symptomatology.

Results: ESCS variables—IBR, IJA, WFC, and RJA—were intercorrelated, rs ranged from |.33| to |.48|, ps<.05, with one exception: IBR was not significantly correlated with IJA. Symptomatology and language variables—STAT total score, MSEL-RL, and MSEL-EL—were significantly correlated, rs ranged from |.45| to |.66|, ps<.01. IBR and WFC were associated with STAT total score, MSEL-RL, and MSEL-EL, rs ranged from |.38| to |.60|, ps<.01. RJA was associated with STAT total score, r=-.35, p<.05, and MSEL-RL, r=.33, p<.05. IJA was not significantly correlated with STAT total score or language subscales of the MSEL.

ESCS variables with a significant univariate association with STAT, MSEL-EL, or MSEL-RL were entered as simultaneous predictors in regression equations to identify unique predictors of autism symptomatology and language abilities. No unique predictors of STAT total score were found, although WFC (p=.07, partial R²=.06) approached significance, F=3.75, p<.05, adjusted R²=.19. IBR (p<.01) and WFC (p<.05) each demonstrated unique associations with MSEL-RL, F=8.10, p<.01, adjusted R²=.37, IBR partial R²=.17, WFC partial R²=.06, and with MSEL-EL, F=9.99, p<.01, adjusted R²=.43, IBR partial R²=.24, WFC partial R²=.05

Conclusions: The strong associations between IBR and MSEL-EL and MSEL-RL indicate the significance of requesting behaviors to language abilities in toddlers. The associations between WFC and MSEL-EL and MSEL-RL indicate that intentional gestural and verbal communication is also a significant indicator of language abilities at this age. Both IBR and WFC exhibited significant univariate associations with the STAT, suggesting that deficits in requesting and intentional social communication are significant, if not unique, predictors of autism symptomatology among toddlers displaying autism symptoms. Understanding the associations between verbal and nonverbal communication and autism symptomatology will provide the building blocks for later tests of treatment efficacy in this Multi-Site RCT of the Hanen More than Words Intervention.