脳卒中後の体幹機能評価の統合的構造解析：マルチスケール因子分析とラスク分析に基づく新しい評価モデル

To investigate the components of trunk function using multiple existing trunk function assessments and construct a new evaluation model.

A prospective observational cohort study with a 1-week follow-up.

A single-center tertiary care hospital.

In total, 200 patients with early-onset stroke (mean age, 75.9 ± 12 years; 57% male) were consecutively recruited. Key inclusion criteria were a diagnosis of acute cerebral infarction or hemorrhage and the ability to mobilize within 48 hours of onset.

Not applicable.

The pre-specified primary outcome measure was trunk function, which was measured 1.2 ± 2.5 days post-stroke using the Verheyden and Fujiwara Trunk Impairment Scales (TIS-V and TIS-F), Functional Assessment for Control of Trunk (FACT), and Trunk Control Test (TCT). Exploratory factor analysis (EFA) was used to classify functional factors, followed by Rasch analysis to confirm unidimensionality and item difficulty within each factor. Assessors were not blinded to participants' clinical information.

EFA identified four functional factors: static sitting, basic movements, and dynamic sitting tasks (More Challenging and Less Challenging). Rasch analysis confirmed unidimensionality for all factors after excluding four items from the Less Challenging dynamic sitting factor. The resulting factors and item difficulties revealed a hierarchical pattern, suggesting that trunk function recovery progresses from static controlled to complex dynamic movements.

To our knowledge, this is the first study to clarify the factor structure and hierarchical organization of trunk function after stroke, providing a framework to reorganize evaluation tools and improve personalized physical therapy. Further research is needed to develop a short-form version to enhance clinical utility.