Physical function varies according to characteristics of executive dysfunction in patients with stroke.

Introduction: While executive dysfunction significantly impacts physical function, most clinical assessments rely on total composite scores, potentially overlooking the heterogeneity of specific subcomponent impairments and their unique contributions to motor recovery. This study aimed to identify distinct profiles of executive dysfunction based on the score distribution of Frontal Assessment Battery (FAB) subitems and clarify their specific associations with motor and balance impairments in patients with stroke. Methods: Seventy-eight patients with subacute stroke were participated in this observational study. Assessments were used the FAB, Brunnstrom Recovery Stage (BRS), Berg Balance Scale (BBS), and Timed Up and Go Test (TUGT). Representational Similarity Analysis (RSA) and k-means clustering were applied to FAB subitem scores to identify patient subgroups based on the similarity of their executive profiles. Results: Three distinct clusters were identified: Cluster 1 (Inhibitory Control Impairment, N = 32), Cluster 2 (Motor Programming Impairment, N = 18), and Cluster 3 (Mild, N = 28). Notably, no significant differences were observed among the clusters in total FAB scores or other global cognitive measures (P > 0.05). However, physical function significantly varied; Cluster 3 demonstrated significantly superior performance in BRS, BBS, and TUGT scores compared to Clusters 1 and 2 (P < 0.05), which exhibited lower physical function despite their different executive profiles. Conclusions: This study suggested that the individual differences in executive dysfunction-which cannot be observed in total scores-and motor dysfunction within clusters are associated with different factors.