Structural validity and internal consistency of a hypothesized factor structure of the Fugl-Meyer Assessment of the upper extremity

Background Many studies have confirmed the psychometric properties of Fugl-Meyer Assessment of the upper extremity (FMA-UE). Although several modified versions of the FMA-UE form exist, their structural validity has not been fully established. Objectives To assess the structural validity and internal consistency of the original, short, and hypothesized FMA-UE forms. Methods In this cross-sectional, multicenter, observational study, the factor structure was assessed using confirmatory factor analysis (CFA) to evaluate the adequate model of each hypothetical FMA-UE form (original, 30-item, 27-item, and 6-item forms). The internal consistency of each FMA-UE form and subscale was assessed using Cronbach's alpha after factor structure evaluation. Results We recruited 363 patients with first-episode stroke (median age = 70.0, median days = 75.0). The results of the original form models were not estimated by CFA. Of all FMA-UE forms, the 30-item form lacked three reflex items (4-factor, 30-item model) and the 27-item form lacked three reflex and three coordination items (3-factor, 27-item, second-order model); these forms demonstrated an adequate model fitness (root mean square error of approximation = 0.056/0.059, comparative fit index = 0.995/0.996, Tucker-Lewis index = 0.995/0.995). The 6-item form demonstrated a poor model fit. All FMA-UE forms and subscales showed a high internal consistency (Cronbach's alpha>0.91). Conclusions Both 30- and 27-item FMA-UE forms showed a good factor structure; therefore, these forms are eligible for use in clinical practice. However, future studies should define the factor structure of the 6-item form.