Working Memory Training Does Not Enhance Instrumental Activities of Daily Living Performance Following Stroke

Background/Objectives: Post-stroke cognitive impairment significantly impacts long-term functional outcomes, particularly in instrumental activities of daily living (IADLs). Working memory training (WMT) has emerged as a potential cognitive rehabilitation strategy; however, its transfer to real-world functionality remains unclear. This study evaluated whether adaptive computerized WMT enhances IADL performance compared to a non-adaptive control condition in chronic stroke survivors. Methods: A single-blind, randomized controlled trial was conducted with 50 adults aged 50–79 years, ≥12 months post-ischemic stroke, diagnosed with mild neurocognitive disorder. Participants were randomized to adaptive WMT or non-adaptive cognitive training, each completing 25 home-based sessions over 12 weeks via a standardized online platform. Primary outcomes included the Lawton and Brody IADL Scale and the Working Memory Questionnaire (WMQ); secondary outcomes included the Working Memory Index (WMI) from the WAIS-IV. Analyses included frequentist and Bayesian methods. Results: Both groups showed significant pre–post improvements in IADL independence and WMI (p < .05; BF₁₀ > 10), with no significant between-group differences on overall IADL outcomes. The adaptive WMT group demonstrated specific gains in WMQ-Storing (p = 0.033; BF₁₀ = 3.83), while the control group improved in WMQ-Attention and IADL-Assistance Required (p = 0.004–0.035; BF₁₀ > 6). Bayesian ANOVA indicated that these effects were primarily driven by the interventions, with minimal influence from depressive symptoms or global cognition. Conclusions: Adaptive WMT yielded domain-specific cognitive benefits but did not enhance IADL performance beyond non-adaptive training. These findings highlight the limited far-transfer of WMT and the importance of designing ecologically valid, multimodal rehabilitation strategies post-stroke.