Autism spectrum disorder and the DSM-5: Diagnostic and symptom change from early childhood to early adolescence

Abstract

Changes to the diagnostic criteria for autism in the DSM-5 raised concerns that some children with DSM-IV Pervasive Developmental Disorders (PDD) would not meet the DSM-5 criteria for Autism Spectrum Disorder (ASD). Research specifically applying the DSM-5 ASD criteria to preschool-aged children is required given the importance of a timely autism diagnosis and early intervention. Few studies have specifically applied the DSM-5 ASD criteria to preschool-aged children. Autism diagnoses and symptoms may change as children develop over time; however, few studies have longitudinally examined patterns of diagnostic and symptom change from early childhood into adolescence. The aims of this thesis were therefore to compare autism diagnoses according to the DSM-IV and DSM-5, to longitudinally examine the diagnostic stability of autism over time using both diagnostic systems; and finally to explore the patterns of DSM-5 ASD symptom stability and change from early childhood to early adolescence. Between 2003 and 2005, 185 young children (aged 20-55 months) with autism and/or developmental delay were assessed as part of a longitudinal study. A subset of this sample (n=50) were then re-assessed in early adolescence (aged 11-15 years). Best-estimate clinical diagnoses were given according to DSM-IV and DSM-5 criteria based on a comprehensive diagnostic assessment, including the Autism Diagnostic Interview-Revised (ADI-R) and Autism Diagnostic Observation Schedule (ADOS). File reviews for the 50 participants were completed to retrospectively apply the DSM-5 ASD criteria in early childhood. To assess the sensitivity and specificity of the DSM-5 ASD criteria, ADOS and ADI-R items were assigned to the DSM-5 criteria and retrospectively applied to the full sample in early childhood (n=185). The findings indicated good DSM-5 sensitivity (.84) at the expense of specificity (.54). Children with PDD who were not classified with DSM-5 ASD (n=20) had significantly lower rates of comorbid behaviour and emotional problems than children with PDD who were classified with DSM-5 ASD (n=106). In children with PDD, gender, language delay, developmental level or adaptive behaviour did not significantly differ between those classified with DSM-5 ASD and those who were not classified with DSM-5 ASD. Best-estimate DSM-IV and DSM-5 diagnoses were then compared for the sample. Of the 43 children with PDD, 12% did not meet the DSM-5 criteria for ASD in early childhood (aged less than 5 years). These children typically had DSM-IV diagnoses of Pervasive Developmental Disorder Not Otherwise Specified (PDD-NOS) and required symptoms on one additional subdomain to meet the DSM-5 criteria. By adolescence, they had developed sufficient symptoms to meet the full DSM-5 criteria for ASD. In early adolescence (aged 11-15 years), 10% of children with PDD did not meet the DSM-5 criteria for ASD based on their current symptoms alone, although they met the diagnostic criteria once past symptoms were taken into account. Consistent with previous research, DSM-IV Autistic Disorder remained stable over time (89%) while all children with PDD-NOS had a change in diagnosis. Stability of DSM-IV PDD (98%) and DSM-5 ASD (89%) were high. Changes in core ASD symptoms between early childhood and early adolescence differed based on the measure used. Clinician-observed (ADOS) symptoms remained stable over time; however, ADI-R Social-Communication symptoms reduced over time while changes relating to restricted and repetitive behaviours and interests (RRBs) differed depending on the RRB subtype. RRB symptoms relating to stereotyped or repetitive motor movements, use of objects or speech reduced over time while RRB symptoms relating to insistence on sameness, inflexible adherence to routines, or ritualised patterns of verbal or nonverbal behaviour increased. Higher developmental level in early childhood was associated with lower ASD symptoms on average during childhood and adolescence; however, it was not associated with symptom change over time. Together, the findings from this thesis contribute to the growing body of research into the impact of the DSM-5 ASD criteria in young children and the characteristics of those who may not meet the diagnostic criteria. The findings indicate that patterns of symptom change may differ based on the assessment tool utilised and that differentiating between ASD symptom subtypes is necessary as broader domain scores may mask differing symptom trajectories. Although autism diagnoses may typically remain stable over time, some children may not yet have sufficient symptoms to meet the DSM-5 criteria for ASD in early childhood. Furthermore, the types of symptoms that require the most clinical attention may change as children develop. These findings highlight the importance of ongoing monitoring of children with autism to ensure that intervention programs are adjusted based on the needs of each child.