Advancing Recognition: New Research Explores Biomarkers for Earlier Autism Identification

Expanding Screening Approaches

Autism identification has historically relied on behavioral observations, which can delay recognition—particularly for children from marginalized communities. New research explores how biomarkers might supplement existing methods. A JAMA study found eye-tracking biomarkers could help differentiate autistic children in primary care settings, though the sample size was limited to 146 participants.

Microbial and Digital Indicators

Some approaches examine biological signals rather than behavior. Research in Nature identified elevated microbially-derived metabolites (chemical byproducts of gut bacteria) in some autistic children's urine and stool samples. Separately, a Nature Medicine study showed AI analysis of home videos could flag early behavioral patterns with moderate accuracy, though the algorithm requires testing across more diverse populations.

A JAMA study found eye-tracking biomarkers could help differentiate autistic children in primary care settings, though the sample size was limited to 146 participants.

The Value of Timely Identification

Earlier recognition matters because support services can be most effective during early developmental windows. A Wiley review compiled evidence for potential presymptomatic biomarkers (biological signals before behavioral signs emerge), though these remain experimental. Early access to communication supports and sensory accommodations—not treatment—constitutes the primary benefit.

Current Limitations and Considerations

While promising, these approaches face significant hurdles:

• The microbiome findings apply only to a subset of autistic individuals and require replication (Frontiers in Neuroscience)
• Commercial stool tests like those promoted by CUHK lack peer-reviewed validation data
• Eye-tracking and AI tools need testing across socioeconomic, racial, and geographic groups
• No biomarkers establish causation—they may reflect co-occurring conditions or environmental factors

Ethical considerations also warrant attention. Over-reliance on biomarkers risks false positives/negatives given autism's heterogeneity. Algorithmic bias in AI tools remains a concern, as noted in Frontiers in Psychiatry.