Characteristics of Developmental Regression In Autism Spectrum Disorders

Background:  Autism spectrum disorders (ASD) are characterized by significant social, communicative, and behavioral impairments. A substantial number of individuals with ASD experience developmental regression (DR).  DR is defined as the loss of previously established speech and/or other skills, which had been present for at least 3 months before the age of 36 months.  The prevalence of DR in ASD ranges from 15% to 47%. The relationship of DR to other features of ASD is not clear.  Defining clinical features associated with DR will facilitate identification of a distinct ASD-DR phenotype and enhance both research and clinical efforts.

Objectives:  The primary aim of this study is to define the nature and scope of associated features and behavior problems in children with ASD and DR vs. those with ASD only. We hypothesize that children with ASD and DR will show more behavior problems across several measures. 

Methods:  499 participants between the ages of 4 and 21 years of age with ASD were identified from a larger dataset of individuals participating in a genetic study of ASD.  These participants had a mean age of 9 years (SD=4.4) at study entry and were predominantly male (86%). As part of a standard clinical protocol, all participants were assessed with the ADI-R, Aberrant Behavior Checklist (ABC), Repetitive Behavior Scales (RBS), Social Responsiveness Scale (SRS), and Vineland Adaptive Behavior Scales (VABS or VABS-II). DR was defined as the presence of language or behavioral regression based on a positive score (1 or 2) on any ADI-R regression item. An individual was classified as DR if they were positive for either language regression or behavioral regression. We did not include individuals who experienced regression with illness in either group. The ASD and DR subset (ASD-DR) consisted of 160 individuals (32%); the remaining 339 ASD individuals (68%) did not have DR.  We conducted multivariate analysis of variance adjusting for age at ADI-R and developmental level as measured by the VABS/VABS-II.  

Results:  The two groups did not differ with respect to sex.  The MANOVA, adjusted for age and developmental level, showed that the ASD-DR and ASD only groups differed significantly on the dependent measures (Wilks’ λ=0.944, F(11,477)=2.58, p=0.003). Examination of univariate tests indicated that the ASD-DR group had significantly higher (greater impairment) mean scores on the ABC Lethargy scale (p=0.04) and Inappropriate Speech scale (p=0.02).  The ASD-DR group also had a significantly higher mean RBS total score (p=0.04) indicating more frequent repetitive behaviors.  Finally, the ASD-DR group had more autism symptoms as noted in significantly higher scores for the Reciprocal Social Interaction (p<0.001), Nonverbal Communication (p=0.01), and Repetitive Behavior (p=0.04) domains on the ADI-R.   

Conclusions:  The presence of DR in ASD appears to be associated with increased impairments in both autism symptoms and associated features. The increased prevalence of these symptoms and behavior problems suggests that ASD and DR may represent a meaningful subgroup for further study. Understanding the range of behaviors associated with DR can help elucidate the biological underpinnings of DR in ASD as well as highlight potential therapeutic targets.