Chapter 2 22 Gregoraci et al. 19 described the changes in smoking-attributable mortality comparing 14 European countries between 1990 and 2004, revealing that inequalities in smoking remain one of the most important entry points for reducing inequalities in mortality. In this current paper, we extend the research on trends in educational inequalities in smoking-attributable mortality in Europe to the period 2000 to 2020. To gain better insight into how smoking contributed to inequalities in partial life expectancy between age 50 and 80, we investigate the contribution of smoking attributable mortality to changes in partial life expectancy for each educational group as well as to educational inequalities in partial life expectancy. Methods Data sources Total and cause-specific mortality by five-year age group, sex and educational level for the period 2000-2020 were collected and harmonized for countries with at least four data points within this period: Austria, Belgium, Denmark, Estonia, Finland, Italy (Turin), Lithuania, Spain (Barcelona), Sweden and Switzerland. Data were part of a post-census longitudinal mortality follow-up, and for most countries, data was available in five-year periods. Barcelona data is from cross-sectional datasets, where the register of mortality is linked to the municipal population registry each year. Educational level was assessed as the highest level of completed education according to the International Standard Classification of Education (ISCED),20 and categorized into three groups: lower (ISCED 0-2, up to lower secondary education), middle (ISCED 3-4, upper secondary education) and high (ISCED 5-8, tertiary education). Following the Preston-Glei-Wilmoth method, we analysed age 50 and over. Educational information was not available for all country-time points at older ages, therefore we excluded ages 80 and onwards. Analysis First, the time periods for which data were available differed between the countries, hence all-cause and lung cancer mortality rates were smoothed and interpolated using penalized splines.21 As a result, yearly mortality estimates for each country between 2000 and 2020 were obtained by age, sex and education, allowing for comparisons between countries. For Lithuania smoothed results did not fit the observed results at the beginning and end of the observation period, hence the results for 2000-2003 and 2020 were omitted. See Appendix I for a comparison between observed and smoothed results. Partial life expectancies (ages 50-79),
RkJQdWJsaXNoZXIy MTk4NDMw