Bedside measures of malnutrition and association with mortality in hospitalized adults

Abstract

Background & aims: The impact of malnutrition on the outcomes of hospitalized adults in resource limited settings such as sub-Saharan Africa (SSA) is not fully described. We aimed to determine the association between malnutrition and mortality in adults admitted to hospital in the resource-limited setting of Southwestern Uganda. Methods: We performed a cohort study of adults admitted to the medical ward of Mbarara Regional Referral Hospital. Measures of nutritional status included: 1) body mass index (BMI), 2) the mininutritional assessment short form (MNA-sf), and 3) mid-upper arm circumference (MUAC). Subjects were followed until death or 30 days from admission. We used proportional hazards regression to assess associations between malnutrition and in-hospital and 30-day mortality. Results: We enrolled 318 subjects. The prevalence of malnutrition was 25e59% depending on the measure used. In-hospital and 30-day mortality were 18% and 37% respectively. In the adjusted analysis, subjects with MNA-sf score 0e7 had a 2.7-fold higher risk of in-hospital mortality (95% CI: 1.3e5.9, p ¼ 0.011) than those with a score of 8e14, and subjects with malnutrition determined by MUAC (<20 cm for males, and <19 cm for females) had a 1.8-fold higher risk of in-hospital mortality (95% CI: 0.98e3.4, p ¼ 0.06) than those normally nourished. MNA-sf (HR 1.6, 95% CI: 1.02e2.6, p ¼ 0.039) and MUAC (HR 1.6, 95% CI: 1.0e2.3, p ¼ 0.048) were independently predictive of 30-day mortality. BMI <18.5 was not associated with in-hospital or 30-day mortality. Conclusions: Malnutrition was common and simple measures of nutritional status predicted in-hospital and 30-day mortality. Further research is needed to understand the pathophysiology of malnutrition during acute illness and mitigate its effects.