Inequalities in mortality by socioeconomic factors and Roma ethnicity in the two biggest cities in Slovakia: a multilevel analysis

Katarina Rosicova,Johan W Groothoff,Sijmen A Reijneveld,Roy E Stewart,Andrea Madarasova Geckova,Jitse P Van Dijk,Martin Rosic

doi:10.1186/s12939-015-0262-z

Abstract

BackgroundThe socioeconomic and ethnic composition of urban neighbourhoods may affect mortality, but evidence on Central European cities is lacking. The aim of this study was to assess the associations between socioeconomic and ethnic neighbourhood indicators and the mortality of individuals aged 20–64 years old in the two biggest cities of the Slovak Republic.MethodsWe obtained data on the characteristics of neighbourhoods and districts (educational level, unemployment, income and share of Roma) and on individual mortality of residents aged 20–64 years old, for the two largest cities in the Slovak Republic (Bratislava and Kosice) in the period 2003–2005. We performed multilevel Poisson regression analyses adjusted for age and gender on the individual (mortality), neighbourhood (education level and share of Roma in population) and district levels (unemployment and income).ResultsThe proportions of Roma and of low-educated residents were associated with mortality at the neighbourhood level in both cities. Mutually adjusted, only the association with the proportion of Roma remained in the model (risk ratio 1.02; 95 % confidence interval 1.01–1.04). The area indicators – high education, income and unemployment – were not associated with mortality.ConclusionThe proportion of Roma is associated with early mortality in the two biggest cities in the Slovak Republic.

Highlights

The socioeconomic and ethnic composition of urban neighbourhoods may affect mortality, but evidence on Central European cities is lacking
Association of neighbourhood characteristics with mortality Using multilevel Poisson regression we assessed the associations of the various area characteristics with ageand gender-adjusted mortality
Introduction of the various neighbourhood characteristics led to a decrease in the random variance at the area level, i.e., a lower intra class correlations (ICC), showing that these characteristics accounted for part of the clustering

Summary

Introduction

The socioeconomic and ethnic composition of urban neighbourhoods may affect mortality, but evidence on Central European cities is lacking. Inequalities in health between socioeconomic and ethnic groups are among the main challenges for public health worldwide and have been an object of study in recent decades [1, 2] These inequalities exist at many levels – between individuals, neighbourhoods, socioeconomic groups, regions, countries and entire continents. Poorly maintained public places, social unrest, crime and violence are but a few of the ways that urban health inequities affect everyone. These peculiar characteristics of cities probably contribute to inequalities in health [29]

Objectives

Methods

Results

Discussion

Conclusion