Boosting Working Memory in ADHD: Adaptive Dual N-Back Training Enhances WAIS-IV Performance, but Yields Mixed Corsi Outcomes

<b>Background/Objectives:</b> This study investigates the efficacy of working memory training (WMT) using the dual N-back (DNB) task on cognitive performance in young adults with Attention-Deficit/Hyperactivity Disorder (ADHD). <b>Methods:</b> Over the course of at least 18 daily sessions conducted within one month, 106 participants (33 non-medicated ADHD, 42 medicated ADHD, and 45 controls) were randomly assigned to either a fixed dual 1-back (FD1B) training condition or an adaptive DNB condition, wherein the N-back level increased following successful completion of each trial block. Cognitive performance was assessed pre- and post-intervention using the Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV) Working Memory Index (WMI) and the Corsi Block-Tapping Task. <b>Results:</b> A mixed-design ANOVA revealed significant improvements in DNB performance across all groups, with the adaptive training condition producing larger gains (e.g., a 204.6% improvement in controls, Cohen's d=1.85). WAIS-IV WMI scores-particularly the Digit Span Backward subtest-also improved significantly post-training, with greater effect sizes in the adaptive condition (d=0.46) than in FD1B (d=0.27). Corsi performance showed very modest gains, showing a surprising tendency to be more associated with the FD1B condition than the adaptive condition. Control participants outperformed the medicated ADHD group on WAIS-IV subtests, although no significant differences emerged between medicated and non-medicated ADHD participants. Correlational analyses indicated task-specific training effects, with adaptive training enhancing associations between DNB and Corsi performance in both controls (r=0.60) and medicated ADHD participants (r=0.51). <b>Conclusions:</b> This study demonstrates that dual N-back training improves verbal working memory in young adults with ADHD, specifically in a sample without psychiatric comorbidities. Transfer benefit to visuospatial domains appears limited and may not generalize to adolescents, older adults, or individuals with complex clinical profiles. The results underscore the importance of tailoring training protocols to maximize cognitive outcomes across different domains.