Using alternative or direct anthropometric measurements to assess risk for malnutrition in nursing homes.

Introduction. The aim of the study is to assess the applicability of alternative versus direct anthropometric measurements for evaluating the risk of malnutrition in elderly persons living in Nursing Homes (NHs) using the Malnutrition Universal Screening Tool (MUST). Materials and Methods. Cross-sectional survey conducted in 67 NHs in Tuscany, Italy. For 641 elderly residents there have been detected: weight, standing height (SH), knee height (KH), ulna length (UL), middle-upper-arm-circumference (MUAC). Correlations between the different methods for calculating the BMI (using direct or alternative measurements) were evaluated by the intraclass correlation coefficient and the Bland-Altman method; agreement in the allocation of participants to the same risk category was assessed by squared weighted kappa statistic and indicators of internal relative validity. Results. Considering the BMI calculated using KH, the intraclass correlation coefficient was 0.839 (0.815-0.861), while considering those calculated by UL was 0.890 (0.872-0.905). The limits of agreement were ±6.13 kg/m2 using KH and ±4.66 kg/m2 using UL. Considering the BMI calculated using SH, 79.9% of the subjects were at low risk, 8.1% at medium risk and 12.2% at high risk of malnutrition. The agreement between this classification and that obtained using the BMI calculated by alternative measurements was “fair-good”. Conclusion. When is not possible to determine the risk category by using SH, we suggest using the alternative measurements (primarily UL, due to its highest sensitivity) to predict the height and to compare these evaluations to those obtained by using MUAC to predict the BMI.