Obesity in ecuadorian adults. An application of Poisson regression and Logistic regression

EasyChair Preprint no. 1569

Abstract

Obesity is a disease which affects around 650 million people worldwide and more than 4.5 million in Ecuador. These figures are alarming because it is recognized as a significant factor in the non-communicable chronic diseases, which appear in all age groups. This work focusses on finding the relations between obesity depicted by Body Mass Index (BMI) and other variables such as gender, province, age group and geographic area for adults between 19 and 59 years old. To achieve this purpose, Logistic and Poisson regression models have been applied, and these results were significant in both models.  After contrast pseudo R square, the best model has been identified as logistic regression due to its best fit and prediction. As a result, it has been obtained that for an adult male who lives in a rural area of Guayas, the possibility of being obese decreases by 78%. Furthermore, it can be affirmed that if an individual who lives in the province of Guayas moves to the province of Pichincha, and the rest of variables remain constant, then the possibility of being obese declines by 31%.

Keyphrases: Logística, Obesidad, Poisson, Regresión

@Booklet{EasyChair:1569,
  author = {Sandra Garcia-Bustos and Nadia Cárdenas-Escobar and Iris Bustamante Peña and Stefany Uguña Salazar and Kevin Peralta Rizzo},
  title = {Obesity in ecuadorian adults.  An application of Poisson regression and Logistic regression},
  howpublished = {EasyChair Preprint no. 1569},

  year = {EasyChair, 2019}}