Educational inequalities in premature mortality by region in the Belgian population in the 2000s

BACKGROUND In Belgium, socio-economic inequalities in mortality have long been described at country-level. As Belgium is a federal state with many responsibilities in health policies being transferred to the regional levels, regional breakdown of health indicators is becoming increasingly relevant for policy-makers, as a tool for planning and evaluation. We analyzed the educational disparities by region for all-cause and cause-specific premature mortality in the Belgian population. METHODS Residents with Belgian nationality at birth registered in the census 2001 aged 25-64 were included, and followed up for 10 years though a linkage with the cause-of-death database. The role of 3 socio-economic variables (education, employment and housing) in explaining the regional mortality difference was explored through a Poisson regression. Age-standardised mortality rates (ASMRs) by educational level (EL), rate differences (RD), rate ratios (RR), and population attributable fractions (PAF) were computed in the 3 regions of Belgium and compared with pairwise regional ratios. The global PAFs were also decomposed into the main causes of death. RESULTS Regional health gaps are observed within each EL, with ASMRs in Brussels and Wallonia exceeding those of Flanders by about 50% in males and 40% in females among Belgian. Individual SE variables only explained up to half of the regional differences. Educational inequalities were also larger in Brussels and Wallonia than in Flanders, with RDs ratios reaching 1.8 and 1.6 for Brussels versus Flanders, and Wallonia versus Flanders respectively; regional ratios in relative inequalities (RRs and PAFs) were smaller. This pattern was observed for all-cause and most specific causes of premature mortality. Ranking the cause-specific PAFs revealed a higher health impact of inequalities in causes combining high mortality rate and relative inequality, with lung cancer and ischemic heart disease on top for all regions and both sexes. The ranking showed few regional differences. CONCLUSIONS For the first time in Belgium, educational inequalities were studied by region. Among the Belgian, educational inequalities were higher in Brussels, followed by Wallonia and Flanders. The region-specific PAF decomposition, leading to a ranking of causes according to their population-level impact on overall inequality, is useful for regional policy-making processes.