Does Low Body Mass Index Predict the Hospital Mortality of Adult Western or Asian Patients?

Background: Some strategies for screening and assessment of malnutrition include a low but variable body mass index (BMI) cutoff, while others do not. The purpose of this systematic review was to investigate published data for Western and Asian hospital samples to determine how the prevalence of low BMI is associated with increased hospital mortality.

Method: A PubMed search of the past 10 years (2006-2016) was conducted with the terms "BMI," "malnutrition," "adult," "outcome," and "hospital" or "ICU" for articles published in English. Studies that examined BMI levels among Western or Asian populations were included. Forest plots were constructed to determine the odds of hospital mortality in low versus normal BMI groups.

Results: Twenty studies met inclusion criteria. The prevalence of BMI <18.5 kg/m was greater in Asian (15%-20%) than Western (2%-7%) patient groups. In Western populations, BMI <18.5 kg/m was so rare that most studies lacked power to evaluate outcomes. Hospital mortality among ICU patients was greater for patients with BMI <18.5 than those with BMI of 18.5-24.9 kg/m (Western: odds ratio, 1.42 [95% CI, 1.33-1.50]; Asian: odds ratio = 1.78 [95% CI, 1.7-1.86]).

Recommendations: BMI <18.5 kg/m is a possible screening variable for malnutrition. Since low BMI was associated with increased mortality in Western and Asian patient groups, we suggest that all populations at risk based on low BMI undergo a full nutrition assessment with a validated method.