Sub-Threshold Autism Symptomatology at Age 2 Is Predictive of Movement Onto the Autism Spectrum at Age 4

Thursday, May 15, 2014
Atrium Ballroom (Marriott Marquis Atlanta)
E. Moulton1, D. A. Fein2, M. L. Barton2, D. Abrams3 and D. Robins4, (1)Clinical Psychology, University of Connecticut, Storrs-Mansfield, CT, (2)Psychology, University of Connecticut, Storrs, CT, (3)Georgia State University, Atlanta, GA, (4)Psychology, Georgia State University, Atlanta, GA

Symptoms of Autism Spectrum Disorders (ASD) are widely thought to emerge by 12 months in some children. This allows for reliable diagnosis as early as 16 months in these children. Another group of children, however, may not meet symptom criteria until later in the toddler or preschool years. Little is known about this group of children who may receive a non-ASD diagnosis before receiving an ASD diagnosis. Are these children simply misdiagnosed or missed early in toddlerhood? Or do they represent a subset of the ASD population who do not show the full disorder until three or four years of age?


To enhance our understanding of the subset of children who appear to move onto the autism spectrum between ages two and four.


Following positive screening on an autism-specific screening questionnaire (M-CHAT(-R)) between 16 and 30 months, 102 children were evaluated and determined to not meet criteria for an ASD. Of these children, 38% received a diagnosis of Developmental Delay, 22.5% received a diagnosis of Developmental Language Disorder, 23.5% received no diagnosis (no DSM diagnosis was met, but development was not typical), 8.8% were typically developing, and 6.9% received other diagnoses (e.g., Motor Delay). The Autism Diagnostic Observation Schedule (ADOS), Mullen Scales of Early Learning, Vineland Adaptive Behavior Scales, and Childhood Autism Rating Scale (CARS) were completed. Diagnostic determination was made based on the clinical judgment of experienced clinicians.


At follow-up at approximately 48 months, 90 children (88%) remained off the spectrum (NonASD-NonASD) and 12 (12%) received an ASD diagnosis (NonASD–ASD). A Chi-square test for independence indicated no significant association between initial diagnosis and movement onto the spectrum (p=.671). The children who moved onto the spectrum were matched to peers who retained a non-ASD diagnosis in terms of gender, age and diagnosis at age two. Independent groups t-tests were used to compare age two CARS and ADOS ratings of the NonASD-ASD group (N=12) to the NonASD-NonASD group (N=24). No significant group differences were found in ADOS domain scores. The NonASD-ASD group was found to have significantly (p=.031) higher CARS total scores (M=24.50,SD=4.34) than the NonASD-NonASD group (M=21.48,SD=3.21). Specifically, the NonASD-ASD group showed poorer nonverbal communication (p=.032) and more abnormal visual response (p=.007). Importantly, however, symptomatology was well below diagnostic threshold for both groups. Individual ADOS items were investigated in these areas, and the NonASD-ASD group was found to have (p=.044) poorer response to joint attention (M=.90,SD=.57) than the NonASD-NonASD group (M=.30,SD=.99).


Movement from a non-ASD diagnosis to an ASD diagnosis between 24 and 48 months occurs for a small subset of children (12% of children who screened positive on M-CHAT but were not diagnosed with ASD). Initial diagnosis does not appear to predict movement onto the spectrum. Children who move onto the spectrum exhibited greater autism symptomatology at age two than children who remained off the spectrum. This indicates that sub-threshold autism symptoms may help to predict this relatively rare occurrence. Further analyses will investigate additional factors that may help to characterize this group.