Abstract

Autism Spectrum Disorder (ASD) and Attention-Deficit/Hyperactivity Disorder (ADHD) show considerable clinical overlap, yet categorical diagnostic boundaries can obscure shared physiological vulnerabilities. To characterize these traits dimensionally and categorically, we analyzed multimodal eye-tracking synchrony in a large-scale transdiagnostic pediatric cohort from the Healthy Brain Network during naturalistic viewing (N = 2,036). Using a novel two-dimensional framework that simultaneously captures the strength and spatial geometry of gaze alignment across subjects - alongside a measure of shared pupillary responses - we quantified how closely children's visual attention and physiological arousal matched those of their typically developing peers across four distinct media contexts. When autistic and ADHD traits were modeled concurrently, their effects on physiological alignment proved context-specific: autistic traits were uniquely associated with reductions in visual attention and arousal synchrony during a complex social narrative, whereas ADHD traits independently predicted reduced synchrony during fast-paced, visually demanding media. Examining the spatial direction of gaze misalignment further revealed that each child tended to scan the scene along a distinct spatial axis, inconsistent with a uniform group-level processing shift. Categorical models additionally revealed a non-additive interaction in comorbidity: rather than compounding physiological deficits, the attentional flexibility characteristic of ADHD appeared to partially offset the restricted spatial attention associated with isolated ASD, resulting in comparatively less severe gaze decoupling in the comorbid group. These findings support the utility of a Research Domain Criteria (RDoC) framework, positioning model-free physiological synchrony as a candidate transdiagnostic biomarker and suggesting that comorbid ASD and ADHD reflect competing rather than simply additive neurocognitive strategies. ### Competing Interest Statement The authors have declared no competing interest. National Institutes of Health, R15MH125332, 5R01MH131335, 1R01AG085665 New Jersey Department of Health, CAUT25BRP005