A model-based approach to long-term recovery of limb apraxia after stroke

Abstract

Limb apraxia is a disorder affecting performance of gestures on verbal command (pantomime), on imitation, and/or in tool and action recognition. We aimed to examine recovery on tasks assessing both conceptual and production aspects of limb praxis in left (n = 22) and right (n = 15) stroke patients. Patients were assessed longitudinally on four conceptual tasks (action identification, tool naming by action, tool identification, and tool naming) and five production tasks (pantomime, pantomime by picture, concurrent imitation, delayed imitation, and tool use). They were grouped as impaired or not relative to the performance of a control sample (n = 27) and as acute‐subacute (first assessment within 3 months post stroke) or chronic (over 3 months post stroke). Hierarchical linear modeling was used to analyze the data. Acute–subacute and chronic patients had similar average performance. All tasks, except action identification, showed evidence of recovery in both acute and chronic impaired patients. A faster rate of recovery among acute–subacute patients was observed only in the two pantomime tasks (action identification and tool identification were not compared on this factor).

Keywords:

• Limb apraxia
• Stroke
• Imitation
• Pantomime
• Transitive gestures
• Recovery
• Skilled movement
• Action
• Motor control

Acknowledgments

The study was carried out at Sunnybrook Health Sciences Centre, a University of Toronto affiliated hospital. This study was supported with grant money from the Heart & Stroke Foundation of Ontario issued to Eric A. Roy and Sandra E. Black. Vessela Stamenova's work was supported through scholarships from Natural Sciences and Engineering Research Council (NSERC), University of Toronto Fellowships, Ontario Graduate Scholarship in Science and Technology, and Toronto Rehabilitation Institute Student Scholarships. We would like to take this opportunity to thank Richard Wolfe for his help with the hierarchical linear modeling. Without his statistical expertise, this paper would have not been possible. We would also like to thank William McIlroy for reviewing the paper and giving us feedback. We would also like to thank all the research assistants who have helped collect data over the years: Kira Barbour, Anish Joshi, Quincy Almeida, Jennifer Salter, Anastasia Aranvitidis, and Mark Gravely.