Desenvolvimento e validação de equações para estimativa do índice de massa corporal sem o uso do peso corporal em adultos

Abstract

INTRODUCTION: The Body Mass Index (BMI) is an international standard widely used and suggested by the World Health Organization for different populations, for estimating and classifying body composition. In poor communities, without resources to purchase and use a scale to check body weight, in situations of instability or irregularity on the ground, the use of this implement is necessary as an alternative method for estimating BMI. OBJECTIVE: To develop and validate equations for estimating BMI without using body weight. METHODS: The study was conducted in four sports training centers located in the city of Fortaleza, Ceará, and in campaigns carried out for the prevention of kidney diseases from 2019 to 2023. The sample consisted of adults aged 18 to 59 years of age of both sexes. As data collection for the development and validation of the equations, the variables waist circumference, height, hip circumference, age, and weight were used. The research participants were divided into male and female groups and into regression and validation subgroups, by sex. A 95% confidence interval was adopted, reflecting a p-value <0.05. Student's test for repeated measures was used to verify differences between samples, Pearson's correlation to analyze correlation between two variables, Stepwise Regression to identify which independent explanatory variables to include in a linear regression model, Interclass Correlation Coefficient to assess data reproducibility, and Weighted Kappa Coefficient to assess agreement or reliability between measurement methods in categorical data. RESULTS: The pilot study sample consisted of 810 adults, 63% (576) of whom were women. The male and female groups had a mean age of 33.72 + 9.44 years and 33.84 + 8.41 years, respectively. In both sexes, there was no statistical difference between paired comparisons in the regression and validation subgroups p > 0.05. After regression analysis and eligibility criteria, four equations were proposed to estimate BMI, namely EM2 and EM3 for males and EF2 and EF3 for females. For the four proposed equations, very strong and positive correlations were observed (r > 0.90; p < 0.05). In the regression analysis, R2 determination coefficient values between 0.861 and 0.901 (p < 0.000) were found. For the Interclass Correlation Coefficient, concordance values of 0.961 and 0.972 (p < 0.05) were found; finally, in the weighted Kappa test, substantial concordance values of 0.658 and 0.711 were found for both sexes (p < 0.05). CONCLUSION: the adoption of the proposed equations for estimating body mass index in adults, without using body weight, is safe and effective for measuring this body measurement in this population, and is indicated when it is not possible to measure the weight of these individuals.