Contribution of main causes of death to social inequalities in mortality in the whole population of Scania, Sweden.

Maria Rosvall,Juan Merlo,Martin Lindström,John Lynch,Basile Chaix

doi:10.1186/1471-2458-6-79

Abstract

BackgroundTo more efficiently reduce social inequalities in mortality, it is important to establish which causes of death contribute the most to socioeconomic mortality differentials. Few studies have investigated which diseases contribute to existing socioeconomic mortality differences in specific age groups and none were in samples of the whole population, where selection bias is minimized. The aim of the present study was to determine which causes of death contribute the most to social inequalities in mortality in each age group in the whole population of Scania, Sweden.MethodsData from LOMAS (Longitudinal Multilevel Analysis in Skåne) were used to estimate 12-year follow-up mortality rates across levels of socioeconomic position (SEP) and workforce participation in 975,938 men and women aged 0 to 80 years, during 1991–2002.ResultsThe results generally showed increasing absolute mortality differences between those holding manual and non-manual occupations with increasing age, while there were inverted u-shaped associations when using relative inequality measures. Cardiovascular diseases (CVD) contributed to 52% of the male socioeconomic difference in overall mortality, cancer to 18%, external causes to 4% and psychiatric disorders to 3%. The corresponding contributions in women were 55%, 21%, 2% and 3%. Additionally, those outside the workforce (i.e., students, housewives, disability pensioners, and the unemployed) showed a strongly increased risk of future mortality in all age groups compared to those inside the workforce. Even though coronary heart disease (CHD) played a major contributing role to the mortality differences seen, stroke and other types of cardiovascular diseases also made substantial contributions. Furthermore, while the most common types of cancers made substantial contributions to the socioeconomic mortality differences, in some age groups more than half of the differences in cancer mortality could be attributed to rarer cancers.ConclusionCHD made a major contribution to the socioeconomic differences in overall mortality. However, there were also important contributions from diseases with less well understood mechanistic links with SEP such as stroke and less-common cancers. Thus, an increased understanding of the mechanisms connecting SEP with more rare causes of disease might be important to be able to more successfully intervene on socioeconomic differences in health.

Highlights

To more efficiently reduce social inequalities in mortality, it is important to establish which causes of death contribute the most to socioeconomic mortality differentials
Relative differences in mortality by socioeconomic position (SEP)/workforce participation were assessed by age-adjusted hazard rate ratios (HRR) presented separately for each age group and sex
With the exception of chronic obstructive pulmonary disease (COPD)–which was only present in older age groups, with too few cases at younger ages– these were the causes studied in further analyses

Summary

Introduction

To more efficiently reduce social inequalities in mortality, it is important to establish which causes of death contribute the most to socioeconomic mortality differentials. Few studies have investigated which diseases contribute to existing socioeconomic mortality differences in specific age groups and none were in samples of the whole population, where selection bias is minimized. The strength and patterning of social inequality in mortality will differ across age, causes, and over place and time [2,3] It follows that, secondly, to more efficiently reduce social inequalities in mortality, it is important to establish which causes of death contribute the most to socioeconomic mortality differentials. Few studies have investigated which diseases contribute the most to existing socioeconomic mortality differences in specific age groups [9,14] and none were in samples of the whole population, where selection bias is minimized. Since specific causes of death differ by age, the disease entities that contribute most to existing disparities in overall mortality can be expected to differ by age

Objectives

Methods

Results

Discussion

Conclusion