Patterns of Changes in Development in Children with Autism Compared to Typically Developing Children

Background: Vocabulary and grammatical development in children with autism has been compared to typically developing children.  For vocabulary acquisition, children with autism appear delayed in their acquisition of vocabulary, but it does increase over time (Dunn & Rapin, 1997).  For grammatical complexity, it appears that at smaller MLUs (i.e. earlier in development) there may be very little difference between autistic and normal children in terms of the relationship between utterance length and grammatical complexity; however, when MLU exceeded 3.0 significant differences arose (Scarborough et al, 1991).  That is, ASD children did not add grammatical complexity to their longer sentences like the typical children. 

Objectives: The current study addressed patterns of change in vocabulary, sentence length, as well as socialization skills, which have long been understood to be a contributing factor in language growth (Mundy, Sigman, & Kasari, 1994).

Methods: 10 boys with Autism Spectrum Disorder (ASD) and 10 typically developing children (TYP) (four boys and six girls) participated in a longitudinal study.  At Visits 1-4, carried out every four months (ASD: 33-45 months old; TYP: 17-29 months old), the mother-child dyads participated in 15-minute free play sessions, which were transcribed and analyzed.  Standardized tests were also administered.  The two groups were matched on language variables at visits 2 and 3.  Four variables were analyzed to inspect patterns of change: number of word types in spontaneous speech, MLU in spontaneous speech, CDI scores, and Vineland Adaptive Behaviors Scales Socialization scores.  Analyses of change over time, group differences, and their interaction were conducted using a Multilevel Model of Change in SPSS.

Results: For the three language variables, significant main effects of both time and group, as well as group by time interactions emerged (ps<.002).  The group differences demonstrated a difference in initial status of the two groups, which confirmed by t-tests, with the ASD group scoring higher.  A significant main effect of time showed that both groups increased their scores on each of the variables over time.  A significant interaction between group and time emerged, with the TYP group having a steeper trajectory on all three variables.   For the VABS socialization scores, no significant group effect emerged.  A significant effect of time emerged (p<.001), demonstrating that both groups increased their socialization scores across visits.  The interaction between time and group was not significant, in that the pattern of change in socialization scores across time was not significantly different for the two groups. 

Conclusions: The language measures revealed significant differences in the pattern of change for the two groups, which was to be expected from prior research.  A difference in the pattern of change in socialization for the two groups was expected due to the close tie between socialization and language.  Some have suggested that problems with the form of the language may stem from difficulties in understanding the function of communication.  The findings of this study may contradict these findings, in that for that this group of ASD children there was a disconnection between the growth of language and socialization skills.