Factors Influencing Obesity Levels in Mexico, Peru, and Colombia

Abstract

The prevalence of obesity has reached alarming proportions on a global scale, exhibiting a marked upward trajectory in developed countries and developing countries. The objective of this research paper is to analyze the principal factors influencing obesity in individuals between the ages of 14 and 61 in Mexico, Peru, and Colombia, utilizing an obesity risk dataset. A variety of socioeconomic, behavioral, and genetic factors were considered, including variables such as age, caloric intake, physical activity, and family history of overweight. Exploratory data analysis (EDA) was conducted using visualization techniques to elucidate the interrelationships between these variables and obesity levels. Subsequently, a logistic regression model was constructed to estimate the probability of different obesity levels based on the predictor variables. The model was evaluated using accuracy, a confusion matrix, and a receiver operating characteristic (ROC) curve analysis. The results demonstrated that the model exhibited a robust capacity to differentiate between obesity levels but encountered notable challenges in distinguishing between adjacent categories (e.g., "Overweight Level I" and "Overweight Level II"). This study offers invaluable insights into the complex multifactorial nature of obesity and provides guidance for public health interventions in Latin American countries to address the escalating obesity epidemic.