Objective: To investigate the longitudinal stability of the Fugl-Meyer Assessment (FMA) of the upper-extremity item difficulties by using Rasch analysis.
Design: Secondary analysis of existing data from a cohort longitudinal study of stroke recovery.
Setting: University research center.
Participants: A total of 377 people, ages 69.2+/-11.2 years, to whom the assessment was administered at 2 weeks and 6 months poststroke.
Interventions: Not applicable.
Main outcome measures: Differential item function analysis performed by using the Winsteps software program examined whether the item difficulty hierarchical order of a modified 30-item FMA for the upper extremity (reflex items removed) was invariant across 2 testing occasions.
Results: Only 2 items (shoulder flexion to 180 degrees, movement with normal speed) showed large differences in test-retest item difficulty calibration. Item instability had no practical consequences on the longitudinal measurement of person ability.
Conclusions: The 30-item assessment shows a longitudinally stable item difficulty order and is valid for measuring volitional arm motor ability over time.