Nonverbal Referential Communication and Language in Infants At High Risk for ASD

Background:   Nonverbal referential communication (NVRC) consists of social prompts to a partner regarding an object or event, i.e. initiating joint attention (IJA) and initiating behavioral requests (IBR). Infants at high risk for an autism spectrum disorder (ASD) may exhibit atypical NVRC development, yet there are few validated measures of NVRC, and their associations with concurrent measures of language have not been well documented.

Objectives:   Determine the stability of measures of NVRC between 6, 9, and 12 months, and the associations of NVRC with a measure of verbal communication.

Methods:   High-risk (n=72; older sibling with ASD) and low-risk (n=52; no siblings with ASD) infants were observed at 6, 9, and 12 months. The Early Social Communication Scales (ESCS) and the Play-based Assessment of Referential Communication (PARC) measured nonverbal referential communication via IJA and IBR. The Mullen Scales of Early Learning (MSEL) yielded Receptive (RL) and Expressive (EL) language T-scores at 12 months.

Results:   There was a non-significant trend for high-risk infants to demonstrate less IBR during the ESCS, t(65)= -1.85, p=.07, and lower EL on the MSEL, t(115.16)= -1.89, p=.07, at 12 months.

For high-risk infants, IJA during the ESCS was stable between 6 and 9 months, r(26)=.47, p<.03, and between 9 and 12 months, r(17)=.77, p<.01. For low-risk infants, IJA during the ESCS was stable only between 9 and 12 months, r(19)=.70, p<.01. IJA during the PARC did not show significant stability for high- or low-risk infants. IBR during the ESCS and the PARC did not show significant stability from 6 to 12 months for either group.

IBR levels between the ESCS and the PARC were associated at 12 months for both high-risk, r(30)=.44, p<.03, and low-risk infants, r(29)=.59, p<.01, while 12 month IJA was not significantly associated between measures for either group.

For high-risk infants, 12 month IBR (mean of z-scored ESCS and PARC IBR) was associated with 12 month MSEL language (mean RL and EL), r(30)=.45, p<.02. 12 month IJA (mean of z-scored ESCS and PARC IJA) was associated at trend levels with 12 month MSEL language, r(30)=.34, p=.07. Neither 12 month IBR, r(29)=.00, p=1.00, nor 12 month IJA, r(29)=.10, p=.59, were associated with 12 month MSEL language for low-risk infants.

Conclusions:   IJA during the ESCS was stable between 6 and 12 months for high-risk infants, but was not associated with IJA in the PARC. At 12 months, IBR on the ESCS and the PARC were associated for both high- and low-risk infants, providing cross-validation of the measurement of 12 month IBR. This prompted the creation of a composite measure of 12 month IBR, which was associated with early language in high-risk but not low-risk infants. The associations between NVRC and language at 12 months suggest a potential unitary communication construct of early NVRC and verbal language in high-risk infants. We are continuing to follow these high-risk infants in order to understand whether their early communication abilities predict later language and ASD symptomatology.