Validation of a regression equation to estimate body weight in the Peruvian population from ENDES 2022 and ENDES 2023.

Tirado A.G. · Medicina Interna de Mexico (2025) · DOI: 10.24245/mim.v41i7.10477

Resumen

OBJECTIVE: To develop an equation to estimate body weight using abdominal circumference, height, and age in Peruvian adults. MATERIALS AND METHODS: Analytical study based on data from the Encuesta Demográfica y de Salud Familiar (ENDES 2022). In the population ENDES 2023 external validation was carried out. The variables were: body weight, age, height and abdominal circumference. Multiple linear regression was applied. The coefficient of determination (R2) and error metrics were determined: mean absolute error, root mean square error, and mean relative error. RESULTS: The ENDES 2022 sample was of 30,071 persons and the ENDES 2023 sample was of 31,247.The multiple linear regression model in ENDES 2022 had a coefficient of determination R2 of 0.895, explaining 90% of the variability in body weight. The regression equation had a mean absolute error of 3.50 kg, root mean square error of 4.58 kg and mean relative error of 0.05, indicating high precision. The Spearman correlation was 0.943. When applying the model in ENDES 2023, the R2 coefficient was 0.876, confirming its predictive capacity in an independent sample. CONCLUSIONS: The regression equation based on abdominal perimeter, height and age is a reliable method to estimate body weight in Peruvian adults.

Instituciones: Universidad Científica del Sur