Shared cognitive risk factors underlying rapid automatized naming deficits for the comorbidity of developmental dyslexia and attention deficit hyperactivity disorder: A computational parameter estimation via Bundesen's theory of visual attention

Developmental dyslexia (DD) and attention-deficit/hyperactivity disorder (ADHD) frequently co-occur, yet their shared cognitive mechanisms remain unclear. Rapid automatized naming (RAN) deficits, a hallmark of both disorders, provide a critical window into the overlapping cognitive underpinnings of DD and ADHD. While previous studies usually attributed RAN deficits to generalized processing speed impairments, the multi-stimulus nature of RAN tasks necessitates a refined framework. Guided by Bundesen's Theory of Visual Attention (TVA) model, the present study dissected multiple-element processing into distinct cognitive subcomponents and examined their associations with rapid naming, reading difficulties and attentional deficits. A sample of 218 children aged 6-14 were divided into four groups: DD-only, ADHD-only, DD + ADHD, and typically developing (TD). A common RAN factor was derived via exploratory factor analysis across multiple RAN tasks. Cognitive subcomponents were quantified using a combi-TVA paradigm to computationally disentangle visual processing parameters. Three TVA subcomponents, including processing speed(C),spatial attentional weight (w), andirrelevance inhibition (α) were found to be impaired in both children with DD and those with ADHD. These impaired subcomponents uniquely contributed to RAN deficits. Furthermore, mediation analyses indicated that the deficits in TVA subcomponents appear to transmit their effects onreading and sustained attention through the bottleneck of impaired rapid naming. By adopting a parameter estimation approach, the present study advances the understanding of RAN deficits from a basic cognitive-process perspective. The findings underscore the utility of TVA-based computational modeling for identifying transdiagnostic biomarker and suggest that interventions targeting these shared cognitive bottlenecks could benefit comorbid cases.