Eye-Tracking Shows Promise as Early Autism Screening Tool for Infants

A 2-Minute Test for Early Clues

Researchers have developed eye-tracking protocols as brief as two minutes that can flag early autism risk markers in infants as young as 12 months old, according to a March 2024 study in Nature. The tests measure how babies visually engage with social scenes — a key area where autistic children often show measurable differences, such as reduced attention to faces or atypical gaze patterns toward social stimuli as detailed in this Cureus review.

Validating the Biomarkers

Large studies are now confirming the reliability of these eye-tracking 'biomarkers' (objective biological measurements). One 2022 Nature study with 585 children aged 7-48 months found that reduced attention to social stimuli and atypical gaze patterns consistently distinguished autistic toddlers with 78% accuracy. These findings align with earlier research showing that nonverbal behavioral markers like social attention differences emerge well before traditional diagnoses. However, researchers caution that these patterns represent correlations rather than definitive causation as noted in this Frontiers review.

A 2025 review highlights unanswered questions about optimal implementation age and how well results predict long-term outcomes.

Why Earlier Detection Matters

Current autism diagnoses typically occur around age 4, but eye-tracking could help identify at-risk children during infancy when early support may have greater impact. As the Nature team notes, these tools are designed to complement — not replace — comprehensive diagnostic evaluations. The goal is earlier access to interventions, not labels. However, the effectiveness of early interventions varies significantly as shown in this Frontiers review of randomized trials.

The Road Ahead

While promising, researchers caution that these tools remain in the validation phase. A 2025 review highlights unanswered questions about optimal implementation age and how well results predict long-term outcomes. Current prototypes show sensitivity ranging from 70-85% and specificity from 75-90% in research settings according to NIH analyses, but performance may vary in diverse populations. Cultural and socioeconomic factors can influence gaze patterns as discussed in SFARI workshops, raising important questions about equitable implementation.