Emerging Biomarkers May Complement Autism Identification Methods

Researchers are exploring objective measures that could complement current autism identification methods, which rely primarily on behavioral observation. Two approaches—analyzing eye-tracking patterns and gut microbiome signatures—are showing promise in recent studies for recognizing neurodevelopmental differences earlier.

Eye-Tracking as a Potential Screening Tool

A 2024 study in JAMA Network Open found that eye-tracking patterns differed between autistic and non-autistic children when viewing social stimuli. These differences, detectable as early as 12-18 months, align with known variations in social attention among autistic individuals. The technology remains specialized but is becoming more accessible for research purposes. Importantly, this study was conducted in specialty clinics—not general pediatric settings—and the equipment used isn't typically available in primary care offices.

These differences, detectable as early as 12-18 months, align with known variations in social attention among autistic individuals.

Microbiome Patterns as Possible Indicators

Separate research published in Nature Mental Health identified distinct gut microbiome signatures in some autistic children's biological samples. These microbial metabolites showed associations with certain presentations of autism, leading to FDA Breakthrough Device designation for a related screening tool. This designation means the approach addresses an unmet need and warrants expedited review—not that its clinical validity is proven. Researchers stress these signatures likely apply only to subsets of autistic individuals, and the method requires validation across diverse populations.

Considerations for Earlier Recognition

Early identification—often before age 2.5—is associated with improved access to supportive services during key developmental periods. Current methods depend on behavioral observation, which can be subjective and may miss subtler presentations. Objective measures could help address this, particularly for groups facing diagnostic disparities. However, experts caution in a 2024 Science Direct review that most biomarker research remains preliminary. Microbiome analysis especially needs broader validation, and while AI-assisted tools are being explored, real-world implementation faces multiple hurdles.