48 Copyright © 2003 The American Society of Neurorehabilitation Constraint-induced movement therapy (CI), a standard-

ized intensive rehabilitation intervention, was given to patients a year or more following stroke. The goal was to determine if CI was more effective than a less-intensive control intervention in changing motor function and/or brain physiology and to gain insight into the mechanisms underlying this recovery process. Subjects were recruited and randomized more than 1 year after a single subcor-tical infarction. Clinical assessments performed before and after the intervention and at 6 months postinterven-tion included the Wolf Motor Function Test (WMFT), the Motor Activity Log (MAL), and the Assessment of Motor and Process Skills (AMPS). Transcranial magnetic stimu-lation was used to map the motor cortex. Positron emis-sion tomography was used to measure changes in motor task–related activation due to the intervention. MAL increased by 1.08 after CI therapy and decreased by 0.01 after control therapy. The difference between groups was significant (P < 0.001). Changes in WMFT and AMPS were not significantly different between groups. Cerebral activation during a motor task decreased significantly, and motor map size increased in the affected hemisphere motor cortex in CI patients but not in control patients. Both changes may reflect improved ability of upper motor neurons to produce movement. Key Words: Positron emission tomography—Transcranial magnetic stimulation—Cerebral infarction—Subcortical