Homogeneous Subgroups of Young Children with Autism Improve Phenotypic Characterization in the Study to Explore Early Development.

The team looked at 707 preschoolers who already had an autism diagnosis.

They ran a latent-class analysis on each child’s full set of scores and traits.

The math sorted the kids into the cleanest, most alike subgroups it could find.

Four clear autism "pictures" came out of the data.

Each picture grouped children with similar language levels, social habits, and repetitive actions.

The new four-type map gives clinicians a tighter frame than the single ASD label.

Barbaro et al. (2017) tracked toddlers and saw 88% keep the diagnosis over time.

Their eye-gaze and vocal markers hint at which kids may later leave the spectrum, while García-Villamisar et al. (2017) show what stable subgroups look like if they stay.

Gizzonio et al. (2015) split preschoolers by gender and found no difference in growth; the latent classes cut across gender lines and give finer detail.

McConachie et al. (2005) warn that Asperger labels are shaky under age 5; using the four phenotype groups instead keeps early profiles descriptive without rushing sub-labels.

You can plug assessment numbers into the four-group model during intake.

If a child lands in the "minimal verbal" class, you know to front-load AAC and mand training.

If the child fits the "social-only gap" class, you might start with peer modeling and social narratives.

The map turns one broad diagnosis into four actionable starting points.