Early Gesture Use in Infants with Autism Spectrum Disorders with and without Regression

Background: Twenty to forty percent of parents report regression, or loss of early communication and/or motor skills, in their children with autism spectrum disorders (ASD).  Recent studies investigating the existence and nature of a regressive phenotype in ASD have validated this phenomenon (Wiggins et al., 2009) and underscored important differences in gesture use prior to regression (Luyster et al., 2005). Werner & Dawson (2005) found similarities in joint attention behaviors in 12 month-old infants with ASD and regression and those with typical development; however, a more complete picture of gesture use in the first year of life has not been outlined. The relationship between early regression in communication and later symptom severity (Baird et al., 2008) highlights the clinical importance of this line of inquiry.

Objectives: To test the hypotheses that: (1) Infants later diagnosed with ASD whose parents reported regression use gestures with the same frequency at 9-12 months as typically developing (TD) infants; and (2) both of these groups use more gestures than infants with later ASD diagnoses without reported regression.

Methods: The study utilized retrospective video analysis, examining home video footage recorded prior to diagnosis. Videos recorded during infancy were obtained from parents of 41 children with ASD and 31 TD children. Researchers collected information from parents about occurrence and age of regression using a questionnaire. For the regression group, data were excluded for children whose parents reported regression prior to or during the 9-12 months period of recording. 
Five-minute samples of randomly selected scenes from 9-12 months footage were compiled for coding of gesture use. Using a checklist, independent coders made three judgments regarding children’s behaviors: (1) Was the behavior among those listed as potential gestures on the checklist? (2) Was there evidence that the child was directing the behavior to another person? (3) Did the behavior serve a communicative function of behavior regulation, joint attention, or social interaction?  A consensus procedure was used for disagreements to ensure optimal data integrity. 

Results: Analyses included descriptive statistics and analysis of variance with appropriate follow-up tests. Screening of data revealed that of the 41 children later diagnosed with ASD, 21 had parent-reported regression, and 20 had parents who reported no regression. The group means for total gesture use in video samples from 9-12 months were: TD = 2.95, ASD with regression = 1.50, ASD without regression = 1.05. The ANOVA showed significant group differences in total gesture use. Follow-up tests showed significant difference between the non-regression and TD groups (p=.03). Total gesture use of the regression group was not significantly different from either the non-regression group or the TD group.

Conclusions: Total gesture use by children with ASD with parent-reported regression was at an intermediate level that did not significantly differ from either the TD group or the non-regression group at 9-12 months of age. The non-regression group was significantly different from the TD group, as hypothesized. Future directions and clinical implications will be discussed.