The Multisensory Attention Assessment Protocol (MAAP): Indices of Attention Predict Diagnostic Status and Symptom Severity In Children with Autism Spectrum Disorders

Background: Children with autism spectrum disorder (ASD) show impairments in social orienting, disengaging attention, and intersensory processing; skills typically well-established during infancy (Ames & Fletcher-Wilson, 2010; Bahrick & Todd, in press, Dawson et al., 2004). These impairments impact subsequent developments in joint attention/communication (Bahrick, 2010; Mundy & Burnette, 2005). A new measure, the Multisensory Attention Assessment Protocol (MAAP), assesses basic indices of attention (disengagement, orienting, maintenance, and intersensory processing) to dynamic audiovisual social and nonsocial events in a single protocol. Preliminary findings (Newell et al., 2007) reveal impairments in disengaging and maintaining attention to social (but not nonsocial) events, and impairments detecting audiovisual synchrony.

Objectives: To develop the MAAP as an instrument for assessing core attention impairments in ASD, we assessed interrelations among three indices of attention. We then investigated which attention indices were most predictive of diagnostic status (TD, ASD) and of joint attention initiation (IJA; index of symptom severity) in ASD.

Methods: Thirteen TD children and 13 children with ASD, 1.5-5 years of age, were roughly matched on Mullen composite mental age (ASD=2.30 yrs; TD=2.89 yrs., p=.18). In the MAAP, trials of a central visual event (13s) were followed 3 s later by two side-by-side peripheral events (10s), with the natural soundtrack synchronized with one of the two events. Blocks (20 trials each) of social (affectively positive speech) and nonsocial events (objects striking a surface) were presented. Intersensory matching (to sound-synchronous events), attention maintenance, and disengagement (latency to shift attention to a peripheral event) were assessed. Finally, children participated in the Early Social Communication Scales (Mundy et al., 2003) and number of IJA instances was observed. 

Results: Significant correlations among the three indices of attention (maintenance, disengagement, and intersensory processing), as well as among attention maintenance, diagnostic status, and IJA were found (Pearson rs>.40,ps<.05). Two structural models were consistent with the data, for social, χ²(2)<.50,p>.79, but not nonsocial events. For both models, higher levels of intersensory matching and lower latencies to disengage predicted higher levels of attention maintenance to social events, which in turn influenced diagnostic status (TD, ASD) and IJA performance (ASD only). Further, each 1% decrease in attention maintenance predicted a 1% increase in the probability of having a diagnosis of ASD. Also, each 1% increase in attention maintenance predicted a .15 increase in the number of IJAs exhibited by ASDs (all ps<.05).

Conclusions: These analyses reflect the first attempt to assess relations among multiple indices of attention, diagnostic status, and IJA performance in ASD. Findings indicate that greater intersensory processing and lower latencies to disengage from a competing stimulus predict increased attention to social speech events. Further, greater attention maintenance is associated with a decreased probability of having a diagnosis of ASD and, in turn, is predictive of increased IJA in ASD (i.e., decreased symptom severity). These findings demonstrate the feasibility of assessing fundamental aspects of attention to naturalistic audiovisual events in a single protocol, and using the child’s performance to predict diagnostic status and symptom severity in ASD.