Early Developmental Patterns of Receptive and Expressive Language in Children with ASD

Background: One area of continued debate within autism research is whether patterns of language acquisition in autism differ qualitatively from those seen in typical development. Research has shown that children with ASD have relatively more severe impairments in receptive than expressive language early in life (see Volden and colleagues, 2011 for review), but little is known about how this atypical receptive-expressive profile may change during development.

Objectives: To determine whether the atypical receptive-expressive profile persists from toddlerhood through the preschool years, to identify the factors that predict the extent of the atypical profile, and to quantify the impact that the early atypical profile has on later language outcomes.

Methods: Participants were a large sample of young children with ASD in a longitudinal study (n = 122 at Time 1). An experienced examiner made diagnoses integrating results from the ADOS, ADI-R, and clinical expertise. Autism severity scores were calculated per Gotham and colleagues (2009). Comprehensive evaluations completed at each visit (ages 2½, 3½ and 5½) included the Preschool Language Scale-4, Mullen, and at 5½, Peabody Picture Vocabulary Test-4. Discrepancy scores were calculated for each child by subtracting PLS-4 receptive from expressive age equivalent scores (see Volden et al., 2011).

Results: Descriptively, the percentage of children with relatively higher expressive than receptive skills decreased across development from 81.1% (99/122) at 2½, to 67.6% (75/111) at 3½, to 45.1% (41/91) at 5½. A series of t-tests indicated that on average, expressive language age equivalents were higher than receptive age equivalents at 2½ (t =-9.9, p< .001) and 3½ (t=-4.1, p<.001), but not at 5½ (t =-1.1, p=.26). Using multiple linear regression, we examined predictors (maternal education, nonverbal cognition, and autism severity) of discrepancy scores from 2½ to 3½. Autism severity (t=2.110, p=.038) and nonverbal mental age (t =-3.206, p=.002) at 2½ were predictive of discrepancy scores at 3½ (R² =0.217, F(3, 91) = 8.387, p<.001). Finally, in a simple linear regression, the discrepancy scores at 2½ did not predict language outcomes at age 5½ on the PLS-4 or PPVT (ps>.3), but the discrepancy scores at 3½ did predict language outcomes at age 5½ on the PLS-4 (R²=0.20, F(1,90) = 22.610, p<.001) and PPVT-4 (R²=0.20, F(1,81) = 19.699, p<.001).

Conclusions: Our finding that receptive language lags behind expressive language in young children with ASD at age 2½ and 3½ confirms the results of previous studies. By age 5½, however, there was no group difference in receptive and expressive age equivalent scores, which more closely mirrors patterns in typical development and is consistent with previous findings in older preschool and school-age children (Kjelgaard & Tager-Flusberg, 2001). Higher autism severity and lower nonverbal mental age at age 2½ were related to higher discrepancy scores one year later. We also found that greater discrepancy scores at 3½ were predictive of worse language outcomes two years later, indicating the impact of the atypical profile on language development over time. Future studies should examine the extent of intervention or developmental effects on changes in the receptive-expressive profile.