Association Between Malnutrition and Outcomes in Patients With Severe Ischemic Stroke Undergoing Rehabilitation

To investigate the incremental prognostic significance of malnutrition in patients with severe poststroke disability.

Retrospective cohort study. The patients were recruited from 3 specialized inpatient rehabilitation facilities. Nutritional status was assessed using the Prognostic Nutritional Index (PNI), which is calculated from serum albumin and total lymphocyte count. Scores >38 points reflect normal nutrition status, scores of 35-38 indicate moderate malnutrition, and scores <35 indicate severe malnutrition. The association of PNI categories with outcomes was assessed using multivariable regression analyses.

Inpatient rehabilitation facility.

Patients (N=668) with ischemic stroke admitted to inpatient rehabilitation within 90 days from stroke occurrence and classified as Case-Mix Groups 0108, 0109, and 0110 of the current Medicare case-mix classification system.

Not applicable.

Three outcomes were examined: (1) the combined outcome of transfer to acute care and death within 90 days from admission to rehabilitation; (2) 2-year mortality; and (3) FIM motor effectiveness, calculated as (FIM motor change/maximum FIM motor-admission FIM motor score)×100.

Overall, the median time to rehabilitation admission was 18 days (range, 12-26 days). The prevalence of moderate and severe malnutrition was 12.7% and 11.5%, respectively. Ninety-one patients (13.6%) experienced the combined outcome. After adjusting for independent predictors including sex, atrial fibrillation, dysphagia, FIM cognitive score, and hemoglobin levels, neither moderate (P=.280) nor severe malnutrition (P=.482) were associated with the combined outcome. Similar results were observed when looking at 2-year mortality. Overall, FIM motor effectiveness was 30%±24%. After adjusting for independent predictors, severe malnutrition (β coefficient -0.458±0.216; P=.034) was associated with FIM motor effectiveness.

Approximately 1 in every 9 patients presented severe malnutrition. On top of the independent predictors, severe malnutrition did not provide additional prognostic information concerning risk of the combined outcome or 2-year mortality. Conversely, severe malnutrition was associated with poorer functional outcome as expressed by FIM motor effectiveness.