Can we predict ASD in high-risk toddlers?


It has become pretty clear that intensive early intervention can improve outcomes for children with autism, but according to the Centers for Disease Control and Prevention, the age of diagnosis is often after the age of 3 (check it out). What’s a birth-to-three SLP to do? It would be great if we could predict which high risk infants and toddlers were more likely to get a diagnosis using widely available tools, or better yet, using tools we are already using.  This study looked at whether scores on developmental measures at 8 and 14 months could predict diagnosis of ASD at 36 months.

Children at high risk (sibling with ASD) and low-risk (typically developing sibling) were assessed using the Mullen Scales of Early Learning (MSEL) and Vineland Adaptive Behavior Scale (VABS) at 8, 14, 24, and 36 months. Together, these assessments cover several developmental domains, including Communication, Motor, Cognitive, Adaptive, and Social. Those five domains should sound familiar to EI evaluators—they’re listed in Part C of IDEA 2004. Children in the high-risk group were also assessed for autism in the later visits, and all of the results were entered into a program for a statistical magic trick analysis.

So, can we predict ASD diagnosis from early measures of global development in our daily practice? Not quite yet, but we can keep a few key points in mind as we evaluate high-risk infants and toddlers. The authors found a few trends we should consider:

  • Scores at 8 months didn’t predict ASD group any better than chance, but by 14 months the predictions got better, with the VABS Daily Living score (an adaptive measure) being the best predictor
  • High risk infants who ended up with an ASD diagnosis tended to show delays first in the motor domain (8 months) and then the social domain (14 months) (but this resulted in only small differences in VABS and MSEL scores)
  • There were clear differences in VBAS and MSEL scores among groups at 8 and 12 months, and those differences grew by 24 and 36 months (keep that in mind when you evaluate a high risk 1-year-old that “doesn’t look that bad”)

Note: This article is open access (that means FREE)!

Bussu, G., Jones, E.J.H., Charman, M.H., Johnson, J.K., Buitelaarthe, K. (2018). Prediction of Autism at 3 Years from Behavioural and Developmental Measures in High-Risk Infants: A Longitudinal Cross-Domain Classifier Analysis. Journal of Autism and Developmental Disorders. Advance online publication. doi: 10.1007/s10803-018-3509-x.