Admission cognitive function independently predicts discharge walking function and functional independence after stroke: a multicenter retrospective cohort study

BACKGROUND AND PURPOSE: Post-stroke cognitive impairment is prevalent, yet its independent contribution to functional recovery during inpatient rehabilitation has not been adequately established in large multicenter cohorts. This study aimed to determine whether admission cognitive function, measured by the Mini-Mental State Examination (MMSE), independently predicts discharge walking function and functional independence at discharge from inpatient stroke rehabilitation. METHODS: This multicenter retrospective cohort study analyzed 577 stroke patients (330 ischemic, 247 hemorrhagic) from four rehabilitation centers in South Korea. Patients were stratified by admission MMSE into normal (≥24, n = 258), mild impairment (18-23, n = 98), and moderate-to-severe impairment (<18, n = 221) groups. Primary outcomes were discharge Functional Ambulation Category (FAC) and Modified Barthel Index (MBI). Multivariable regression analyses adjusted for age, sex, baseline FAC, baseline MBI, and comorbidities. ROC analysis identified optimal MMSE cut-off values. RESULTS: Admission MMSE independently predicted discharge FAC (β = 0.013, p = 0.004) and discharge MBI (β = 0.736, p < 0.001). Patients with moderate-to-severe cognitive impairment achieved significantly lower discharge FAC (1.86 ± 1.52 vs. 3.27 ± 1.34, p < 0.001) and MBI (40.43 ± 28.84 vs. 74.76 ± 19.14, p < 0.001) than cognitively normal patients (Cohen's d = 0.99 and 1.42). An MMSE cut-off of 20 showed acceptable discriminative ability for functional independence (MBI ≥60: AUC = 0.771). The predictive effect was strongest in patients with severe motor impairment (FAC 0-1; ρ = 0.585 for MBI, p < 0.001). CONCLUSIONS: Admission MMSE independently predicts discharge walking function and functional independence at rehabilitation discharge. An MMSE score of 20 may serve as a clinically useful threshold for identifying patients at risk of poor functional outcomes. Routine cognitive screening at admission is recommended.