Early ASD Symptom Severity Predicts Diagnostic Transition to Global Developmental Delay (GDD)

Background:  Although early diagnosis of Autism Spectrum Disorders (ASD) is generally stable over time, some toddlers diagnosed with ASD no longer meet criteria when they are older. In particular, some initially diagnosed with ASD are later diagnosed instead with Global Developmental Delay (GDD). This may be due to difficulty distinguishing between these disorders at a young age, or due to change in symptom presentation over time. Studies have sought to identify initial differences between individuals with ASD and those with GDD. Results have been variable, though children with ASD tend to have greater impairment in social responsiveness and more restricted, repetitive, and stereotyped behaviors. Despite these group comparisons, research on predictors of diagnostic transition between these disorders is limited.

Objectives:  To identify abilities and symptoms in toddlers with ASD at a 2-year-old evaluation which predict being diagnosed with GDD and no longer meeting ASD criteria by age 4.

Methods:  214 children (174 males, 81%) who had been diagnosed with ASD at age 2 (T1, M=26.44 months, SD=4.51) were re-evaluated near age 4 (T2, M=49.73 months, SD=6.85). The initial evaluation occurred after the child screened positive on the Modified Checklist for Autism in Toddlers (M-CHAT; Robins et al., 1999) or was flagged by a pediatrician for possible ASD. Evaluations included ADOS, Vineland Adaptive Behavior Scales (-II; VABS), and Mullen Scales of Early Learning (MSEL). 

Results:  Of the 214 toddlers given an ASD diagnosis at T1, 15 (7%) no longer met criteria for ASD at T2 and were diagnosed with GDD (ASD-GDD); 199 maintained their ASD diagnosis (ASD-ASD). Logistic regression revealed that although all children met criteria for ASD at T1, those with lower ADOS comparison scores at T1 were more likely to meet criteria for GDD rather than ASD at T2, OR=.74, p=.02. Fewer DSM-IV-TR symptoms at T1 in the social domain also predicted transitioning from ASD to GDD diagnosis, OR=.30, p<.001. However, sex and measures of cognitive and language ability (MSEL, VABS) were not related to this diagnostic transition, p>.05.

Conclusions:  Although all children met criteria for ASD at T1, toddlers who would later be diagnosed with GDD (ASD-GDD) had milder ASD symptom severity at T1 than the ASD-ASD group. These results indicate that those who change diagnosis from ASD to GDD over time may be differentiable from those with stable ASD at initial diagnosis. Furthermore, instability in the ASD-GDD group is due to differences in ASD symptom severity, and not by differences in cognitive and language ability. This has clinical implications for accurate early diagnosis of developmental disorders and their treatment. However, the lack of other group differences, such as in cognitive and language ability, contribute to the difficulty in differentiating between these disorders at a young age. Future research will examine the role of intervention in the diagnostic transition of these toddlers.