Monitoring of trends in socioeconomic inequalities in mortality

Abstract

Studies from several countries reported that the relative mortality gap between low and high socioeconomic groups widened during the 1970s and 1980s. While this wellknown finding has important policy implications and prompted research on underlying causes, it also calls for more a detailed and accurate monitoring of past and current trends. Objectives. The aim of this paper is to present new estimates of changes in socioeconomic inequalities in mortality between the 1980s and the 1990s in different European countries. The estimates are given with the specific aim to illustrate (i) large variations, both within and between countries, in the pace by which socioeconomic inequalities in mortality changed over time, and (ii) the considerable degree to which the observed trends may be sensitive to data problems and to the methodological choices made. Data and methods. The paper is based on a EU sponsored project on monitoring of socio-economic inequalities in mortality and morbidity. Data were obtained on all-cause mortality by occupational class and educational level in nine western European countries both in the early 1980s and in early 1990s. Trends in mortality were analysed by assessing trends in (a) group-specific standardised mortality rates and (b) summary measures of the magnitude of mortality differences between socioeconomic groups. Results. The weight of evidence from all countries points towards a widening of relative inequalities in mortality between the early 1980s and early 1990s, while the absolute gap remains about the same. However, important variations were observed in the pace of change, both between countries, and within countries (between men and women, and between age groups). In addition, a widening of relative inequalities was found to concur with decreasing life expectancies of the disadvantaged groups in some cases, but increasing life expectancies in many other cases. A second series of analyses illustrate that, even though there may be little doubt that relative inequalities in mortality widened in many countries, data problems can often impede an accurate assessment of the precise rate of change. For example, trends in mortality differences between occupational classes can strongly depend on (a) the social class scheme used and (b) whether or not economically inactive persons are included in the analysis. The use of "unlinked" cross-sectional studies may suffer from subtle but influential biases. Conclusion. Monitoring of trends in inequalities in mortality should go beyond the simple assessment such as "the gap is widening", and monitor in detail the pace of change, both for the national populations at large and for sub-populations such as specific age-sex groups. This monitoring needs to evaluate carefully the potential effects of data problems as well as the choice for specific methods and indicators.