Socioeconomic Inequalities in Body Mass Index in Barcelona 1986-2016: An Unconditional Quantile Regression Approach

Abstract

We used unconditional quantile regression to reveal heterogeneity in the social gradient at different levels of Body Mass Index (BMI) distribution by educational level, employment situation and neighborhood income for 1986 and 2016 in Barcelona. The contribution of these factors to changes in BMI between the years at distinct percentiles is illustrated using the Oaxaca decomposition method. It is confirmed heterogeneity in BMI values at different percentiles for both years. BMI values were higher at the upper percentiles in unemployed and people living in low-income neighborhoods. Overall, improved educational level between years helped to reduce overweight and obesity. In contrast, retired and unemployed population in low-income neighborhoods account for much of the worsening in the upper percentiles of BMI.