Modelo de regressão para variáveis antropométricas de alunos do Programa Saúde na Escola

Abstract

Obesity has been increasing in recent decades, becoming one of the public health challenges. Aiming at prevention, promotion and health care in the school environment, the School Health Program (SHP) was created, with nutritional assessment being one of the health actions provided for in the program. The objective of this study was to determine a regression model for anthropometric variables of SHP students, in the city of Unistalda, in the state of Rio Grande do Sul. For this, data were collected from the nutritional assessment of 125 PSE students from five to 18 years old, enrolled in basic education in municipal public schools in the rural area of the municipality. A descriptive analysis was carried out in order to verify the nutritional status of the students, tests to analyze the association between the categorical variables BMI/Age and Height/Age according to sex and also to verify the correlation between the variables BMI vs. age, weight vs. height and age vs. stature. A percentage of 41.6% of excess weight was observed, being 20% obesity. There was no association between the categorical variables. A positive linear relationship was verified in the correlation analysis between the variables. A linear regression model was fitted and, to verify the adequacy of the fit, diagnostic techniques and residual analysis were applied. The gamma distribution with identity binding function showed the best goodness of fit according to the criteria for selection of models considered.