Emerging Biomarker Approaches May Expand Options for Autism Identification

Expanding Approaches to Autism Identification

Current autism identification typically involves behavioral observations by trained clinicians, a process that can be inconsistent across regions and may overlook individuals whose traits don't match historical stereotypes. Several research teams are exploring biological and behavioral markers that could work alongside existing methods, potentially helping more people access identification services earlier if desired.

Gut Microbiome Patterns

A 2023 Nature study found certain microbial metabolites in stool samples that correlated with autism identification in some children. The researchers emphasize these patterns may only apply to specific subgroups. While one microbiome-based test has received FDA Breakthrough Device designation — a program that accelerates development of promising but unproven technologies — independent validation of its accuracy claims remains ongoing as noted by The Chinese University of Hong Kong.

Gut Microbiome Patterns A 2023 Nature study found certain microbial metabolites in stool samples that correlated with autism identification in some children.

Eye-Tracking Research

A May 2024 JAMA Network Open study (n=146 at a single site) found eye-tracking measurements could distinguish autistic from non-autistic children in controlled settings. The technology tracks differences in how children engage visually with social stimuli. However, as Contemporary Pediatrics notes, translating such research into practical primary care tools requires addressing equipment costs and real-world variability.

Digital Behavioral Analysis

Artificial intelligence is being studied to identify behavioral patterns associated with autism. A 2023 Nature study showed machine learning could analyze brief video clips, while 2021 Science Advances research using healthcare records data suggested digital phenotyping might reduce false positives in screening — though this approach reflects system engagement patterns as much as core traits.

Considerations for Implementation

In the U.S., the average age of autism identification remains around 4 years old, with later identification common in marginalized communities. While some studies suggest early support access may improve outcomes, these findings remain nuanced and shouldn't override individual autonomy. Any new tools must be voluntary, culturally responsive, and developed with autistic community input to avoid harm from over-identification or inappropriate interventions.