Profiling clinical heterogeneity in Autism Spectrum Disorder at time of children's diagnosis: A cluster analysis from the ELENA cohort.
Adding adaptive and behavior scores to DSM-5 criteria sorts newly diagnosed kids with ASD into four useful sub-types.
01Research in Context
What this study did
The team looked at 458 French children who had just learned they have autism.
They added scores for daily-living skills, IQ, language, and behavior problems to the usual autism checklist.
A computer then grouped the kids into clusters that looked most alike.
What they found
Four clear sub-types showed up, not just "mild" or "severe."
One group had high IQ but big behavior issues; another had low IQ and calm moods, and so on.
DSM-5 labels alone would miss these patterns.
How this fits with other research
Rivard et al. (2023) ran the same math on the kids with mixed delays and also found clusters—three instead of four—showing the method works across diagnoses.
Bachrach et al. (2026) later showed that Israeli preschool teams already use IQ and irritability scores to pick special vs. mainstream classes, proving these extra numbers guide real placement choices.
Modabbernia et al. (2016) warned that birth oxygen trouble raises autism risk; Harmony’s four profiles help you see which kids might need closer developmental watches after a rough start.
Why it matters
Stop treating every new ASD client as "level 1, 2, or 3." Add an adaptive test like the VABS and a brief behavior scale at intake. The combo will show you which of the four common profiles the child fits, letting you write goals that match strengths and trouble spots from day one.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pull the last five intake folders and add a Vineland and a brief behavior checklist; look for the four-cluster pattern to adjust goals.
02At a glance
03Original abstract
BACKGROUND: Autism Spectrum Disorder (ASD) encompass a heterogeneous group of neurodevelopmental conditions characterized by deficits in social communication and repetitive behaviors. The rising prevalence of ASD highlights the urgent need for effective diagnostic and intervention strategies. However, the significant clinical, cognitive and etiological heterogeneity within ASD populations poses substantial challenges to these efforts. AIMS: This study aimed to identify distinct ASD subtypes at time of diagnosis within the ELENA cohort by incorporating not only DSM-5 criteria but also measures of adaptive functioning and behavioral problems. MATERIAL AND METHODS: Data from 458 children and adolescents with ASD were analyzed using hierarchical agglomerative clustering. Variables included autistic symptoms, intellectual quotient, adaptive behavior and behavioral problems. Clusters were identified based on these parameters, and post-hoc analyses were conducted to assess statistically significant differences in sex and age among the four clusters using Chi-square test and Student's t-tests. RESULTS: Four distinct clusters were identified from the analysis: (1) High Autistic Symptom Severity with Lowest Behavioral Problems, (2) High Autistic Symptom Severity with High Behavioral Problems, (3) Low Autistic Symptom Severity with Highest Behavioral Problems and (4) Low Autistic Symptom Severity with low behavioral problems, while significant age differences were observed across clusters, no significant sex differences were found. DISCUSSION: These clusters exhibited significant variability in adaptive functioning and behavioral problems, suggesting that DSM-5 criteria alone do not fully capture the complexity of ASD. The findings underscore the importance of incorporating measures of adaptive functioning and behavioral problems into ASD assessments and interventions. Future research should aim to validate these clusters in larger and more diverse populations and explore the integration of genetic and neuroimaging data to further refine the characterization of ASD subtypes. Additionally, longitudinal studies are needed to assess the stability and clinical relevance of these subtypes over time. TRIAL REGISTRATION NUMBER: NCT02625116.
Research in developmental disabilities, 2025 · doi:10.1016/j.ridd.2025.105040