Children with Autism Spectrum Disorder (ASD) with or without a History of Developmental Regression: Differences in Developmental Levels, Adaptive Functioning, and Behavior and Emotional Problems

Background: Previous research has documented the occurrence of developmental regression (i.e., the loss of previously-acquired language or social skills) in some children with Autism Spectrum Disorder (ASD). However, there is limited research on the developmental outcomes of children with ASD and a history of regression.

Objectives: We assessed the proportions of language and social regression in preschool children with ASD and examined differences in developmental levels, adaptive functioning, and behavior and emotional problems in children with ASD with or without regression.

Methods: We used data from the Study to Explore Early Development, a US multi-site community-based case-control study. In this analysis, we included 671 children born between 2003 and 2006 who met the SEED case definition for ASD based on the Autism Diagnostic Observation Schedule and the Autism Diagnostic Interview-Revised (ADI-R). Based on caregivers’ ADI-R responses, children were assigned to one of four groups: (1) no regression, (2) language regression only, (3) social regression only, and (4) language+social regression. We used the Mullen Scales of Early Learning (MSEL) to evaluate developmental levels (i.e., language, motor, and visual reception abilities); Vineland Adaptive Behavior Scales-II (VABS-II) to assess adaptive functioning, and Child Behavior Checklist-Preschool (CBCL-P) to examine behavior and emotional problems. Descriptive statistics and ANOVA tests were computed to examine differences in these areas according to regression status.

Results: 8.94% (n=60) and 8.64% (n=58) of mothers reported child language regression and social regression, respectively and 8.35% (n=56) reported regression in both domains.

Differences in Developmental Levels: We found significant differences between the four regression groups in MSEL age equivalents (all p-values <0.05). Specifically, children in the no regression group had higher age equivalence than children in the language only group in visual reception, fine motor, receptive language, and expressive language domains. Also, the no regression group had higher age equivalence than the social only group in the fine motor domain, and higher age equivalence than the language+social regression group in the receptive and expressive language domains. Similarly, the no regression group had higher early learning composite standard scores than the language only regression group.

Differences in Adaptive Functioning: Significant differences were also found between groups in the VABS-II communication domain only, with children in the no regression group having higher VABS-II communication standard scores than children in the language only and language+social regression groups (p=.012).

Differences in Behavior and Emotional Problems: We found significant differences between groups in CBCL-P internalizing behavior t-scores (p=.002), but not CBCL-P externalizing behavior t-scores. Children in the no regression group had lower t-scores in the internalizing problems domain than children in the language only and social only groups.

Conclusions: These findings confirm the occurrence of regression in some children with ASD, and suggest that children with loss of skills might have lower developmental levels and adaptive functioning as well as increased behavior and emotional problems. Therefore, children with a history of regression are likely to need more support and interventions to target their specific needs.