# Neighbourhoods’ social, built, and natural environment characteristics and body mass index in Latin American cities

## TL;DR

Neighbourhood education levels are linked to body mass index in Latin American cities, but the relationship varies by country and gender.

## Contribution

This study provides novel insights into how neighbourhood characteristics influence BMI in diverse Latin American contexts.

## Key findings

- Higher neighbourhood education was associated with lower BMI in Argentina and Chile.
- The relationship between neighbourhood education and BMI varied by gender in Colombia and Mexico.
- Built and natural environment features showed less consistent associations with BMI.

## Abstract

Features of neighbourhoods affect body mass index (BMI) but this has been poorly acknowledged within the highly heterogeneous and unequal contexts of Latin American cities. We evaluated associations between social, built, and natural environment characteristics of neighbourhoods with BMI, and investigated whether these associations were modified by individual socioeconomic position (SEP).

We linked individual data (n = 43 968) from national health surveys to data on neighbourhoods (n = 3428) and cities (n = 165) in Argentina, Chile, Colombia, and Mexico. Linear mixed models were used to estimate associations between neighbourhood education, intersection density, and greenness with BMI, adjusting for individual- and city-level characteristics.

Associations between neighbourhood education and BMI varied by country, in both magnitude and direction. In Argentina and Chile, higher neighbourhood education was associated with lower BMI. This negative association was also observed among women in Colombia and Mexico, although it was weaker. Among men in Colombia and Mexico, however, the association was positive. Associations of neighbourhood intersection density and greenness with BMI were less robust. In general, we did not find strong evidence of effect modification by individual SEP.

Neighbourhood education is associated with BMI beyond individual and city characteristics, although the associations are heterogenous across countries and by gender. Associations with built and natural features were less clear. Our results highlight the relevance of context-specific analysis for planning interventions that are aimed to reduce BMI and its unequal distribution in Latin American cities.