New Biomarkers for Early Autism Diagnosis Show Promise, But Challenges Remain

Eye-Tracking Gains Ground in Primary Care

A study in JAMA Network Open published last month found that eye-tracking biomarkers could help diagnose autism during routine pediatric visits. The research, involving 1,089 infants, showed that measuring how infants visually engage with social stimuli (like faces versus objects) achieved diagnostic accuracy comparable to specialist evaluations in controlled settings. However, experts caution that real-world implementation faces challenges, including the need for specialized equipment and training, as noted in Contemporary Pediatrics. This builds on earlier work using digital tools to detect subtle behavioral differences, as reported by Nature in 2023, though those studies also highlighted variability across the autism spectrum.

Microbial Clues in Urine

Separately, scientists identified a potential urinary biomarker linked to gut bacteria activity. A Nature study found elevated levels of specific microbially-derived metabolites in a subgroup of autistic children, suggesting this could someday help identify certain autism phenotypes. The metabolites trace to compounds produced by gut bacteria that influence brain development. However, this study involved only 163 children, and researchers emphasize the need for larger, more diverse samples to confirm these findings, as discussed in a review on ScienceDirect.

The Gut-Brain Connection

While not yet peer-reviewed, an AI-powered stool test analyzing gut microbiome patterns recently received FDA Breakthrough designation, as reported by MobiHealthNews. Researchers hypothesize such tools might flag microbial signatures associated with autism, though causal relationships remain unclear. The eye-tracking and metabolite findings, by contrast, have stronger evidentiary support in top-tier journals. A Frontiers in Psychiatry article cautions that while AI tools show promise, their clinical utility and accuracy require rigorous validation.

Why Early Detection Matters

Early diagnosis allows access to supportive therapies during critical developmental windows, as highlighted in Neurology Advisor. Current diagnosis relies on behavioral observation, often delaying identification until age 4 or later. Biomarkers could provide objective screening tools to complement existing methods, especially where specialist access is limited. However, The New York Times notes that early screening must balance benefits against risks of false positives and overdiagnosis.