Developmental Phenotypes and Severity Profiles of Autism Spectrum Disorders in Preschool Children

Background:

Revisions proposed to the classification of pervasive developmental disorders (PDD) in the DSM-V amalgamate all subcategories into one Autism Spectrum Disorder (ASD) with severity descriptors¹.  Measuring severity has broad applications to clinical, diagnostic, research, and intervention efforts and in disease prognosis.  Gotham² developed a severity score algorithm using the Autism Diagnostic Observation Schedule–G (ADOS-G), a semi-structured, standardized play assessment with excellent interrater reliability, internal consistency, and test-retest reliability³. Gotham’s algorithm combines 10 social interaction and communication ADOS items into one category (Social Affect), and adds 4 restricted/ repetitive behaviour items, age, and verbal skills to generate a diagnostic severity score.  This approach emphasing clinical phenotype has been validated ⁴, with parent reports of adaptive skills and behaviour also of prognostic value ^5,6,7.

Objectives:

Gotham’s algorithm was used to analyze the relationship between clinical diagnosis and severity score in preschoolers referred for autism assessment.  Scores and diagnosis were correlated with parent measures of adaptive and executive functioning and behaviour to expand understanding of preschool phenotypes of ASD.

Methods:

Retrospective data was collected from 602 charts of children assessed between 2006 and June 2011.  Severity scores were calculated from the ADOS-G, with adaptive functioning, executive functioning (EF), and behavioural assessments including the ABAS-II ⁸, BASC-p ⁹, and BRIEF-P¹⁰ respectively. Statistical analysis was performed using SPSS with final clinical diagnosis determined by developmental behavioural paediatricians and /or child psychiatry.

Results:

Severity scores were plotted according to clinical diagnosis, but no defined cut offs distinguished diagnostic groups other than at the ends of the spectrum (Autistic Disorder (AD) vs. Asperger’s Syndrome (AS)).  Significant differences in adaptive (global, conceptual, adaptability, communication) and behavioural scores (anxiety, internalizing problems, hyperactivity) were found when AS, AD, and Mild Autism (Asperger’s + PDD-NOS) were compared individually to other ASD diagnoses, however no significant difference was found in EF.

Conclusions:

Findings suggest severity scores and behavioural phenotypes better distinguished  diagnoses at the ends of the autism spectrum (AD vs. Asperger’s), with differentiation of ASD and PDD-NOS problematic, supporting the proposed revisions for a single diagnostic category. Use of the ABAS-II, BASC-p, and ADOS-G in phenotypic profiling of preschoolers is recommended.  While executive dysfunction is common in ASD¹¹ it may be difficult to capture by parent report in the preschool years¹². Ongoing data analysis warrented.