Abstract
OBJECTIVE: First, to investigate whether the 6 aspects of executive functions assessed by the Frontal Assessment Battery have different components and, if so, to extract those components using principal component analysis. Second, to identify patient groups based on their characteristics using cluster analysis. DESIGN: A cross-sectional study. SUBJECTS: Seventy-eight patients with stroke. METHODS: The Frontal Assessment Battery, Mini-Mental State Examination, Trail Making Test, and Stroop Color Word Test were performed within 5 days. RESULTS: Based on principal component analysis, the Frontal Assessment Battery was classified into cognitive control (subscales 1-3, 5) and behavioural control (subscales 4 and 6). Two clusters that reflect these components (cluster 1, n = 68; cluster 2, n = 10) were identified. The between-group comparison showed that compared with cluster 1, cluster 2 had lower scores on Frontal Assessment Battery subscales 4 and 6, the Frontal Assessment Battery total scores, and other executive functions scores. The Mini-Mental State Examination scores had no significant differences. CONCLUSIONS: The Frontal Assessment Battery can be classified into 2 components, and the impairment of Frontal Assessment Battery subscales 4 and 6 identified a specific group of patients with stroke with severe executive dysfunction.