A Longitudinal Study of the Social-Cognitive Phenotype of ASD and Reading Comprehension Development

Thursday, May 12, 2016: 11:30 AM-1:30 PM
Hall A (Baltimore Convention Center)
N. S. McIntyre1, P. C. Mundy2, M. Solomon3, T. Oswald4, L. E. Swain-Lerro1, M. C. Zajic1, J. B. McCauley4 and H. K. Schiltz5, (1)University of California at Davis MIND Institute, Davis, CA, (2)Education and Psychiatry, University of California at Davis, Sacramento, CA, (3)MIND Institute, Sacramento, CA, (4)University of California at Davis MIND Institute, Sacramento, CA, (5)Human Development, University of California at Davis, Davis, CA

Observations suggest that reading comprehension (RC) impairment is part of the social communication phenotype of higher-functioning school-aged children with ASD (HFASD), showing significant negative associations between ASD symptom intensity, oral language skills, text reading fluency, and RC (e.g. McIntyre et al., 2015; Norbury & Nation, 2011). This includes observations that social cognition, as operationalized by Theory of Mind (ToM) tasks, was a unique predictor of RC even after controlling for oral language and word reading for adolescents with ASD (Ricketts et al., 2013).  


1) To compare changes in RC over a 15-month development period across children with HFASD, ADHD, or Typical Development.  2) To test the hypothesis that social cognition is a specific predictor of RC development among children with HFASD. 


Participants included 8- to 17-year-old children with HFASD (N=70), ADHD (N=30), and typical development (TD, N=40). The groups were matched on age but varied on FIQ (99, 100, 115, respectively). Analyses control for FIQ where appropriate. ASD symptoms were confirmed with the ADOS-2 and presence of ADHD symptoms were confirmed with Conners-3 parent reports. Reading fluency and comprehension were measured with the Gray Oral Reading Test-5 (GORT-5), and a latent ToM score was comprised of scores from the Strange Stories (Happe, 1994) and the Silent Films (Devine & Hughes, 2012) tasks.


A repeated measures ANCOVA revealed little change in RC for any of the diagnostic groups in GORT-5 across 15 months, λ= .99, F(2,139)=0.28, p=.76 η2p =.004.  The HFASD group, however, was significantly lower than both the ADHD and TD groups at Time Points 1 & 2, F(2,143)=7.10,6.10,p’s<.01,η2p=.09 & .08, respectively.

MANCOVA results indicated the HFASD group performed significantly lower than both the ADHD and TD groups on both ToM measures, λ= .88, F(2,139)=0.28,p=.76 η2p =.06.  A regression predicting Time Point 2 RC with Diagnostic Group, Time Point 1 RC, FIQ, language measures, the latent ToM Score, and the Diagnostic Group by latent ToM score interaction term revealed that the latter was significant after considering all other variables in the equation, change in R2=.07,F(1,62)=6.21,p<.02. Follow-up regressions within each of the three diagnostic groups indicated that latent ToM measure explained unique significant portion of the variance in TP2 RC for HFASD Group, but not for ADHD or TD groups (Table 1). Separate regressions revealed the association between the ToM and TP 2 RC in HFASD was significant when both the more verbal (Strange Stories) or less verbal (Silent Films) measures were used the analyses. 


Results indicated that while children with HFASD did not lose reading comprehension skills, neither did they make any progress toward closing the previously-reported gap, placing them at risk for falling further behind as reading materials become increasingly more complex and integral to academic achievement. Furthermore, mental state understanding (ToM) in the HFASD group was impaired relative to the other diagnostic groups and explained unique variance in TP2 RC providing evidence that the social cognitive phenotype of ASD is important to consider when designing reading interventions for children with ASD.