Nutritional risk, body mass index and rehabilitation outcome in cognitively impaired vs. cognitively normal patients

IntroductionThe role of low body weight and body mass index in the rehabilitation of cognitively impaired (CI) patients is unclear.Materials and methodsIn a prospective cohort study in a rehabilitation unit for elderly patients recovering from acute illness, we explored functional outcomes in patients with a BMI < 20 kg/m2 and Malnutrition Universal Screening Tool (MUST) ≥ 2 as a marker of high risk of malnutrition, in cognitively intact and impaired patients. One hundred and fifteen patients (mean age 84.7 years, range 69–98, 70 females) were followed up. All received an individually tailored rehabilitation programme as standard. The Barthel Activity of Daily Living (BADL) score was performed on admission and discharge and the primary outcome was improvement in BADL.ResultsPatients with an improved BADL had a higher mean MMSE (20.7v17.7; P = 0.02) and BMI (24.1v 20.9; P = 0.006) and lower mean MUST (0.45v1.11; P = 0.002). A 2-way ANOVA showed significant variance and cognitively normal patients with a BMI ≥ 20 showed the greatest improvement in BADL (P = 0.03). CI patients who improved had a higher BMI (23.8v21.2; P = 0.02) and lower MUST (0.43v1.09; P = 0.004). Patients with a BMI ≥ 20 (19.3v12.8; P = 0.31) and MUST < 2 (18.4 v15.9; P = 0.64) showed a greater mean improvement in BADL after rehabilitation. The total number of CI patients showing improvement in BADL was significantly higher in those with a BMI of ≥ 20 kg/m2 [55/75 (73.3%) v10/22 (45.5%); P = 0.02].ConclusionPatients with cognitive impairment and a BMI < 20 kg/m2 or MUST ≥ 2 on admission to rehabilitation are less likely to show improvement in BADL with rehabilitation when compared to cognitively impaired patients with a MUST < 2 or BMI ≥ 20 kg/m2.