The main purpose of this study was to analyze the geographical variability of mortality inequalities and, in particular, the associations between these variations and socioeconomic and environmental inequalities in small areas of the Barcelona metropolitan area throughout the 1994-2003 period.
There is evidence that the geographical variability of social inequalities in terms of health continues to exist even when it is controlled for by individual risk factors. However, relatively few studies have examined the contribution of exposure to air pollutants to these inequalities.
Given that in the MEDEA Project, the small-area unit was the census section, the study population comprised residents of the Barcelona metropolitan area. The outcome variable was the standardized all-cause and cause-specific mortality rate. The explicative variables were the deprivation index, which summarizes the socioeconomic variables at the census section level, and the estimates of the exposure to air pollutants. We used a spatial analysis based on hierarchical Bayesian models to reduce additional variability by using standardized mortality rates and to assess potential associations between mortality, deprivation, and air pollution.
We found statistically significant associations between deprivation and causes of death related to tobacco and alcohol consumption in men and, in addition to lung cancer, with diet-related causes in women. The coefficients accompanying the air pollution variables were statistically significant only for the Barcelona metropolitan area and in men. We observed that there was a statistically significant, positive interaction between air pollutants and the deprivation index for respiratory mortality and PM10, and for ischemic mortality and NO2, both in the case of men.
Based on our findings, we concluded that deprivation had a statistically significant association with the geographical variation of mortality in the census sections of the Barcelona metropolitan area throughout the 1994-2003 period. In addition, we found that air pollutants more directly related to traffic modified some of these associations.
This study was one of the first worldwide and the first in Spain to use the Besag‑York‑Mollié method in an applied context.