Scholarlink

<p>Obesity has become a pressing public health concern worldwide, and Mexico is among the countries most severely affected. This study analyzes obesity prevalence in adults aged 20 and above across 2446 municipalities using a combination of predictors derived from Geographic Information System (GIS) data, satellite imagery, and socioeconomic information. The variables considered include distance to the United States border, the share of Gross Domestic Product (GDP) generated in urban areas based on nighttime light intensity, the Normalized Difference Vegetation Index (NDVI) as an indicator of green space, poverty, the aging index, educational backwardness, and food insecurity. To account for spatial variation, the analysis applies Multiscale Geographically Weighted Regression (MGWR), which allows the strength and direction of these factors to change across different geographic contexts. This approach contrasts with Ordinary Least Squares (OLS), where all predictors are assumed to have uniform associations nationwide. The findings reveal marked geographic clusters of obesity, with prevalence particularly high near the U.S. border. Moreover, the association of poverty, urban GDP share, NDVI, population aging, education, and food insecurity differ between local and regional levels. These results illustrate the value of MGWR in uncovering spatially uneven patterns of obesity and provide a data-driven basis for geographically targeted public health policies in Mexico.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Exploring the spatial heterogeneity of obesity determinants in Mexican municipalities using multiscale geographically weighted regression

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月22日

查看全文
  • Katerine Restrepo Gómez

摘要

Obesity has become a pressing public health concern worldwide, and Mexico is among the countries most severely affected. This study analyzes obesity prevalence in adults aged 20 and above across 2446 municipalities using a combination of predictors derived from Geographic Information System (GIS) data, satellite imagery, and socioeconomic information. The variables considered include distance to the United States border, the share of Gross Domestic Product (GDP) generated in urban areas based on nighttime light intensity, the Normalized Difference Vegetation Index (NDVI) as an indicator of green space, poverty, the aging index, educational backwardness, and food insecurity. To account for spatial variation, the analysis applies Multiscale Geographically Weighted Regression (MGWR), which allows the strength and direction of these factors to change across different geographic contexts. This approach contrasts with Ordinary Least Squares (OLS), where all predictors are assumed to have uniform associations nationwide. The findings reveal marked geographic clusters of obesity, with prevalence particularly high near the U.S. border. Moreover, the association of poverty, urban GDP share, NDVI, population aging, education, and food insecurity differ between local and regional levels. These results illustrate the value of MGWR in uncovering spatially uneven patterns of obesity and provide a data-driven basis for geographically targeted public health policies in Mexico.