The Development of Joint Attention and Vocalizations in Infants at Heightened Risk for Autism Spectrum Disorder

Background:  Infants at heightened risk (HR) for Autism Spectrum Disorder (ASD) who later receive an ASD diagnosis exhibit delays in both joint attention (JA) and pre-linguistic vocalization development (Morales et. al., 2000; Paul et. al., 2011). In typically developing infants, vocalizations are used early on to establish JA between infants and caregivers (Wu & Gros-Louis, 2014), demonstrating an early-emerging link between these behaviors. Previous studies have focused on JA and vocalizations separately (Toth et. al., 2006), but the simultaneous development of these behaviors, as well as their collective impact on early skill acquisition, remains to be explored. 

Objectives:  This study was designed to examine the development of JA, vocalizations, and their co-occurrence in three groups of HR infants. 

Methods:  50 HR infants (30 males) were observed during the Early Social Communication Scale (ESCS; Mundy et. al., 2003) at 14, 18 and 24 months as part of a larger longitudinal study. At 36 months, each infant was classified into one of three outcome groups: Autism Spectrum Disorder (ASD; n = 9); Language Delay (LD; n = 15), or No Diagnosis (ND; n = 25). Videotaped sessions were coded offline by blind coders. Initiating joint attention (IJA: e.g., eye contact between a toy and experimenter, showing) and initiating behavioral requests (IBR: e.g., reaching, giving, pointing) were coded according to the ESCS manual (Mundy et. al., 2003). Each vocalization was identified and coded as vowel only (VO), syllabic (i.e. containing a consonant), or word. 

Results:  Repeated measures ANOVAs were utilized for all analyses. For JA, a significant main effect of outcome was found only for higher level IJA behaviors (e.g., shows: F(2,47)=14.43, p=.00). For vocalizations, a 3 (age) x 3 (outcome group) x 3 (vocalization type) repeated measures ANOVA revealed a significant 3-way interaction (F(8,188)=8.613, p=.000; see Figure 1). Follow-up ANOVAs indicated an age by vocalization type interaction for the ND (F(4,100)=71.71, p=.000) and LD (F(4,56)=10.611, p=.000) groups, but not for the ASD group. The ASD group consistently produced fewer and lower quality vocalizations at all three ages, showing no increases in production of any vocalization type at any age. Examining the co-occurrence of JA behaviors and vocalizations, a main effect of outcome was revealed for both IJA + vocalization and IBR + vocalization (p’s < .05). Infants later diagnosed with ASD paired vocalizations with JA behaviors significantly less frequently than their ND and LD peers. 

Conclusions:  Infants who later receive an ASD diagnosis produced fewer higher level IJA behaviors, vocalized less, and produced more vowel-only vocalizations then their ND and LD peers. They further demonstrated a specific difficulty coordinating these behaviors. Coordinating JA behaviors with a vocalization likely enhances communicative quality and engages partners in a way that a simple reach or shift in eye gaze may not. These findings highlight the importance of observing how these two types of communicative behaviors develop together in a HR population.