Body mass index, nutrition and rehabilitation outcome in elderly patients with cognitive impairment

Abstract

BACKGROUND: Cognitive impairment is known to adversely affect rehabilitation but the role of poor nutrition in the rehabilitation of such patients needs exploring. METHODS: In a UK-based prospective cohort study in a general rehabilitation unit for elderly patients we explored rehabilitation outcomes in 115 patients (mean age 84.7, 70 females). We used BMI and MUST as surrogate markers of poor nutrition. All patients were included in an individually tailored programme comprising one-to-one attention and group sessions. The Barthel Index (BI) was performed on admission and discharge. Patients also had BMI, MUST, Mini Mental State Examination (MMSE), and Carlson Comorbidity Index recorded. RESULTS: Improvement on the BI was significantly associated with cognition (P = 0.018), BMI (P = 0.003), MUST (P = 0.001) and length of stay (P = 0.006). Logistic regression identified MUST (P = 0.039, CI 0.37–0.97) and in a separate analyisis, BMI (P = 0.027, CI 0.04–0.15) as independent variables that influenced rehabilitation outcome. Patients with cognitive impairment and a high MUST or BMI <20 kg/m2 were significantly less likely to show improvement in BI with rehabilitation when compared to cognitively impaired patients with low MUST or BMI ≥20 kg/m2. Cox Regression showed the former group of patients had a slower trajectory of improvement. CONCLUSION: Patients with cognitive impairment, a Low BMI and a high MUST are less likely to show an impovement in Barthel Score with Rehabilitation when compared to cognitively impaired patients with a low MUST or BMI≥20 kg/m2. Nutrition may be a reversible risk factor to improve rehabilitation outcomes in this group of patients.