[VALIDATION OF PREDICTIVE EQUATIONS FOR WEIGHT AND HEIGHT USING BODY CIRCUMFERENCES IN MEXICAN ELDERLYS].

Nutricion hospitalaria : organo oficial de la Sociedad Espanola de Nutricion Parenteral y Enteral

PubMedID: 26667750

Osuna-Padilla IA, Borja-Magno AI, Leal-Escobar G, Verdugo-Hernández S. [VALIDATION OF PREDICTIVE EQUATIONS FOR WEIGHT AND HEIGHT USING BODY CIRCUMFERENCES IN MEXICAN ELDERLYS]. Nutr Hosp. 2015;32(6):2898-902.
INTRODUCTION
weight and height measurements are important data for the nutritional assessment of elderly people and the implementation of the nutritional care process. Malnutrition is common in this population, who has high rates of disability that difficult to measurement this variables.

OBJECTIVE
evaluate the validity of predictive equations for weight and height that include body circumferences created for brazilian population, in mexican elderly people.

METHODS
this is a comparative, observational, prospective and cross-sectional study, 61 elderly were evaluated. Body weight, height, half span, calf, arm and abdominal circumferences were determinated. Weight and height were estimated with de predictive equations published by Rabito et al. Bland-Altman analysis and Intraclass Correlation Coefficient were used to assess the levels of agreement between the estimated and the measured values. The level of statistical significance was p < 0.05.

RESULTS
the age mean was 78.7 ± 8.7 and 55.7% were females. The weight mean was 61.9 ± 14.1 kg, height mean was 155.4 ± 9.5 cm and Body Mass Index (BMI) mean corresponded to 25.5 ± 5.1 kg/m. The Bland-Altman plots indicated that the 95% confidence interval (95% IC) limits for the difference between real and estimated weight ranged from -14.3 kg to 8.1 kg, the mean of the difference or systematic error (SE) was -3.1 kg, we observed an statistically significant coefficient of 0.12 (p < 0.03). The 95% IC limits for the difference between real and estimated height ranged from -11.1 to 15.9 cm, the diffe rence mean or SE of 2.4 cm, we observed a coefficient of -0.04 (p = 0.67) . Intraclass Correlation Coefficient of 0.72 (p < 0.00) and 0.88 (p < 0.00) were obtained for weight and height, respectively.

CONCLUSION
the equations developed by Rabito showed a good agreement when compared with the actual weight and height of elderly people. We observed variations in the estimated weight in obesity elderlys.