Psychometric properties of the Chinese version of the Arm Activity Measure in people with chronic stroke

Introduction: The Arm Activity Measure was developed to assess active and passive functions of the upper limb in people with unilateral paresis, but a Chinese version is not available and its psychometric properties have not been specifically tested in people with stroke. This study aimed to translate and culturally adapt the Chinese version of the Arm Activity Measure (ArmA-C) and establish its psychometric properties in people with chronic stroke. Methods: The psychometric properties of ArmA-C were determined in 100 people with chronic stroke. Results: The ArmA-C had good test–retest reliability (intraclass correlation coefficients [ICC] = 0.87–0.93; quadratic weighted Kappa coefficients = 0.53–1.00). A floor effect was identified in section A of the ArmA-C. The content validity and internal consistency (Cronbach's alpha coefficients = 0.75–0.95) were good. The construct validity of the ArmA-C was supported by acceptable fit to the two-factor structure model and significant correlations with the Fugl-Meyer Assessment for Upper Extremity score, grip strength, the Wolf Motor Function Test score, the Trail Walking Test completion time, and the Oxford Participation and Activities Questionnaire scores. Conclusions: The ArmA-C is reliable and valid for assessing active and passive functions in people with chronic stroke.