What Do Demographics Have To Do With It? An Oaxaca-Blinder Decomposition of Changes over Time in Inequalities in Alcohol, Narcotics and Tobacco-Related Ill Health in Sweden

The purpose of this study was to document historical trends and socioeconomic inequalities in ill health outcomes related to alcohol consumption, narcotics use and tobacco smoking over the seventeen years prior to the implementation of the Swedish government’s first strategy for alcohol, narcotics, doping and tobacco (ANDT) in 2011. We also sought to explain the changes over time in terms of changes in the population distribution of selected demographic and socioeconomic characteristics. Our two key research questions, for each of alcohol, narcotics and smoking were: 1) How have trends in a) consumption, inpatient care and deaths, and b) income-related inequalities therein developed over time? 2) To what extent can demographic (gender, age, civil status, foreign background), socioeconomic (parental education, own education) and social characteristics (social isolation, proportion of welfare recipients in the municipality) explain the trends in a) levels of consumption, inpatient care and deaths, and b) income-related inequalities therein? For consumption, we investigated the prevalence of heavy drinking and smoking; data on narcotics use were not available. We used International Classification of Diseases (ICD) codes to identify inpatient care and deaths related to alcohol, narcotics and smoking. In our main analyses we used income as a measure of socioeconomic rank. We performed sensitivity analyses to investigate: i) the use of education as an alternative socioeconomic rank, ii) differences between measures of relative and absolute inequality, and iii) sex-differences in the trends over time. We document increasing pro-poor socioeconomic-related inequalities in all of our outcomes except heavy drinking (which was concentrated among higher income individuals, and did not change significantly) during the study period. This reflects an increasing concentration of smoking, and inpatient care and deaths related to alcohol, narcotics and smoking among low income individuals. We are able to explain some of the change over time by demographic and socioeconomic changes (i.e changes in the distribution of our sample by age, foreign background and educational attainment). However, our findings suggest that most of the change observed was due to external factors, such as changing norms and behaviours, and policy or macroeconomic conditions affecting certain groups more than others. In order to achieve the goal of equality in health, ANDT as a policy area must address the increasing concentration of alcohol-, narcotics- and smoking-related outcomes among the poorest and least educated in our society.