By Matthew Maenner
Early identification of autism spectrum disorders (ASD) continues to be an important public health objective. Research has shown that ASD can be reliably identified in children by around 2 years of age, and public health campaigns promote the detection of developmental “early warning signs” that may indicate ASD. Despite these efforts, there is a considerable gap between the age ASD is detected in clinical research, and the age at which children are identified as having ASD in typical community settings. Previous population-based studies have shown that the average age of ASD identification in the community is less than ideal (at 5.7 years), and there is little information about whether these “early warning signs” lead to earlier ASD identification in everyday practice.
Our new study uses data from the CDC Autism and Developmental Disabilities Monitoring (ADDM) Network to answer two questions about how ASD behavioral features are described by community professionals and whether these behaviors are associated with the age of ASD identification. The ADDM Network identified 2,757 8-year-old children that met the surveillance case definition for ASD (based on the DSM-IV-TR criteria) in 2006.
First, we examined the frequency and patterns of diagnostic behaviors that lead to a child meeting the diagnostic criteria for ASD (based on the DSM-IV-TR). There are many different ways to meet the diagnostic criteria for ASD. For example, there are 616 combinations of the 12 behavioral criteria that fulfill the minimum number (6) and pattern needed for “Autistic Disorder” alone. Although there was considerable variability between individuals with ASD, boys and girls had similar patterns of documented behaviors, as did black and white children overall. Among the 2,757 children, the most commonly documented behaviors were impairments in emotional reciprocity (90%), delays in spoken language (89%), and impairments in the ability to hold a conversation (86%). The least frequently documented behaviors were lack of sharing enjoyment or interests (49%) and lack of spontaneous or pretend play (57%).
Our second question was whether particular ASD behaviors (such as those highlighted by the CDC’s “Learn the Signs” campaign) are actually associated with earlier ASD identification in typical community settings. We found that the both the total number and types of diagnostic behaviors in a child’s record were strongly associated with the age that they were identified as having ASD. Children with all 12 behavioral symptoms were diagnosed at a median age of 3.8 years of age, compared to 8.2 years for children with only 7 of the 12 behaviors. Additionally, children with documented impairments in nonverbal communication, pretend play, inflexible routines, or repetitive motor behaviors tended to have an earlier age at ASD identification than children who did have these features in their records. Children with impairments in peer relations, conversational ability, or idiosyncratic speech were more likely to be identified as having ASD at a later age.
These findings give us a clearer understanding of how ASD diagnoses are made in the community, and help inform efforts to maximize early identification and intervention among children with ASDs. It may be more difficult to detect ASD at an early age among children with fewer symptoms, or symptoms that are most apparent at later ages (such as getting along with peers or conversational ability). A recent national telephone survey reported an increase in ASD prevalence among young teenagers, and parents were more likely to describe their later-diagnosed children as having “mild” ASD. It’s possible that increased awareness and intensified screening for ASD could lead to more individuals being identified at both earlier and later ages. Strategies to improve early ASD identification and interventions could benefit by considering the manner in which individuals may meet ASD criteria.