Population-specific modifications of the short-form Mini Nutritional Assessment and Malnutrition Universal Screening Tool for elderly Taiwanese.

BACKGROUND The Mini Nutritional Assessment (MNA) and the Malnutrition Universal Screening Tool (MUST) are two widely used malnutrition screening scales. OBJECTIVE The study was to compare the grading ability of the two scales, and to determine whether adoption of population-specific anthropometric cut-points could improve the grading ability of these tools and whether calf circumference (CC) could be an acceptable alternative to BMI in these scales. DESIGN Purposive sampling. SETTINGS Outpatients receiving annual physical examination at an area hospital. PARTICIPANTS Community-living elderly who were 65 years or older, able to communicate orally, without acute health conditions and willing to sign a study-consent. METHODS Subjects were measured for anthropometrics and blood biochemical indicators, and interviewed for personal data and answers to the MNA and MUST. The risk of malnutrition was evaluated with the short-form MNA (MNA-SF) and the MUST, each in three versions (the original, Taiwan version-1 (T1) which adopted population-specific anthropometric cut-points, and Taiwan version-2 (T2) which replaced BMI with CC). Long-form (LF) MNA versions served as references. RESULTS Results showed that (a) in both scales, patterns of nutritional status rated with the original versions were different from those rated with respective modified versions but ratings made with two modified versions were the same, (b) the T2 versions showed the best grading ability based on agreement with the reference (MNA-LF), and (c) MNA-SF versions rated greater proportions of subjects at risk of malnutrition than the respective MUST versions. CONCLUSIONS (a) Adoption of population-specific anthropometric cut-points improves the grading ability of the MNA-SF and the MUST in community-living Taiwanese, (b) CC is an acceptable alternative to BMI for both MNA-SF and MUST, and (c) nutritional assessment tools should be as much population or ethnically specific as possible to account for cultural and anthropometric differences across populations.