Associations between social and built environments and cardiovascular disease mortality across 342 Latin American cities: results from the SALURBAL study.

Background: Latin American cities have heterogeneous mortality and urban environment profiles. We aim to describe the variability in cardiovascular disease (CVD) mortality across and within cities, and within and between Latin American countries, and investigate the associations between urban environmental features and CVD mortality in these areas. Methods: Using harmonised data from the Salud Urbana en América Latina (i.e. Urban Health in Latin America) project, we investigated associations between urban features and CVD mortality in 342 Latin American cities encompassing 1252 sub-city units in nine Latin American countries. We obtained CVD mortality data from vital statistics and population denominators from national census bureaus. We evaluated the social environment through an educational attainment index from national censuses. Built environment indicators covered urban form, landscape, street design, transportation, and greenness. We applied multilevel negative binomial models with random intercepts for sub-city units nested within cities to describe the associations of social and built environmental features with CVD mortality adjusted by age and sex in each country. Results: CVD death rates varied across sub-cities within and between Latin American countries, with greater variability in age-adjusted CVD death rates between sub-cities within cities (60%) than between cities within countries (21%) and between countries (19%). In fully adjusted models, higher educational attainment in sub-cities (rate ratio (RR) = 0.95; 95% confidence interval (CI) = 0.94-0.96) and higher landscape fragmentation in cities (RR = 0.97; 95% CI = 0.95-0.99) were associated with lower CVD mortality. Conversely, higher street connectivity in sub-cities (RR = 1.02; 95% CI = 1.01-1.04) was associated with higher CVD mortality. No clear associations were observed for sub-city greenness, population density, and the presence of mass transport infrastructure in cities. Conclusions: Our findings emphasise the impact of urban inequalities on cardiovascular health, while also identifying education as an important and modifiable urban environmental feature that could be amenable to public health policies aimed at addressing disparities in CVD mortality within Latin American cities.