Patterns of Behavioral and Emotional Problems in Young Children With Developmental Disabilities: Cluster Analysis and Longitudinal Follow-Up.
Cluster your preschool ASD intake data—three natural profiles appear and point you to the right first goals.
01Research in Context
What this study did
The team ran a cluster analysis on preschoolers with autism. They pulled scores from cognitive, language, adaptive, and behavior checklists. Math grouped the kids into clusters that share the same pattern of strengths and trouble spots.
What they found
Three clear subgroups popped out. One cluster had strong thinking but big behavior spikes. Another showed slow language with mild acting-out. The third was low across most areas. The splits were sharp enough to guide different lesson plans.
How this fits with other research
Journal et al. (2024) used the same math on social-communication scores and also got three preschool ASD profiles. Their clusters forecast later IQ and adaptive gains, so the subgroups look stable over time.
Gur et al. (2024) repeated the idea on 5,836 toddler records and found four milestone delay profiles instead of three. The extra group is likely because they added very young kids and medical-record data.
Sparaci et al. (2015) did an early version with PDD-NOS toddlers and again found three clusters. The match across studies shows three is the sweet spot once kids reach preschool age.
Why it matters
You no longer have to treat "autism" as one big blob. Run a quick cluster on your assessment battery and place each preschooler in one of three data-based boxes. Pick targets that fit the box—behavior focus for the high-cog group, language plus self-help for the slow-language group, and intensive across-the-board teaching for the global-delay group. Start Monday by adding the cluster variables to your intake spreadsheet.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Add cognitive, language, adaptive, and behavior scores to one sheet, run k-means with k=3, and let the cluster label pick the first two treatment targets.
02At a glance
03Original abstract
Children with autism spectrum disorder (ASD) present with heterogeneous levels of abilities and deficits. The identification of subgroups within a specific age range could be useful for understanding prognosis and treatment planning. We applied Hierarchical Clustering on Principal Components (HCPC) with a sample of 188 preschoolers with ASD and identified three distinct subgroups based on multiple developmental and behavioral domains. Cluster 1 was characterized by relatively high cognitive, language and adaptive abilities, and relatively low levels of social symptoms, repetitive behaviors, and sensory issues within the sample. Cluster 2 was characterized by similarly high cognitive, language and adaptive abilities compared to Cluster 1, but more severe social deficits as well as repetitive and sensory behaviors. Finally, Cluster 3 was characterized by lower cognitive, language and adaptive abilities, and more severe social, repetitive, and sensory symptoms. These findings provide insights into how considering multiple developmental and behavioral domains and core autism symptoms simultaneously can distinguish subgroups of young children with ASD and provide more comprehensive developmental profiles. Moreover, the unique profile of children in Cluster 2 highlighted the usefulness of including different measures and informants when evaluating the abilities and deficits of preschoolers with ASD and the importance of understanding the relationships among different developmental and behavioral factors in this specific population. Autism Res 2020, 13: 796-809. © 2020 International Society for Autism Research, Wiley Periodicals, Inc. LAY SUMMARY: Children with autism exhibit a range of abilities and deficits in different developmental and behavioral areas, making it difficult to tailor treatment and predict outcomes. We identified three distinct subgroups among 188 preschoolers with autism spectrum disorder distinguished by the combination of measures from multiple developmental and behavioral domains. The findings revealed the importance of comprehensive profiling of the child's abilities and deficits to inform subgrouping within autism.
Journal of autism and developmental disorders, 2026 · doi:10.1002/aur.2263