Conference Agenda

We study mortality between 2002 and 2018 at municipality level in Italy, i.e. for some 8,000 local administrative areas for which detailed mortality information is missing. We computed standardized mortality ratios (SMRs), relating observed to expected deaths, the latter deriving from the application of national age-specific mortality rates to local population age structures. We verified that the indicator produced reliable results, analyzed its properties, and used it to describe and interpret the specificities of small-area mortality in Italy.

The main results are that territorial heterogeneity in terms of SMR was low but slightly increasing in the period under examination (the coefficient of variation passed from 12.5% to 15%), with an increase of the between-region deviance (although the within-region component remained prevalent, at about 80% of the total). Municipality “fragility” (as measured by an ad-hoc Istat-prepared composite index) and poverty (per capita income) correlated closely with mortality. Spatial regression models, where the territorial component is explicitly introduced to keep spill-over ad omitted variable effects under control, proved necessary, both SAR (Spatial Autoregressive Model) and SEM (Spatial Error Model).

Cardiovascular (CVD) mortality is a key component of life expectancy dynamics. CVD mortality is influenced by risk factors and important heterogeneity in CVD mortality exists. This paper assesses the role of risk factors in educational and regional inequalities in CVD and all-cause mortality in Spain and its regions.

We use multiple cause of death mortality data and assess CVD mortality and CVD mortality with mentions of any of the following risk factors: Diabetes, chronic Kidney disease, Obesity, Lipidemias and Hypertension. We use Poisson models to examine relative inequalities and potential gains in life expectancy by eliminating educational and regional inequalities on risk factors associated to CVD.

We found a positive impact on Spanish life expectancy from eliminating educational (0.2 years for females, 0.14y for males), regional (0.19y, 0.2y), and both axis of inequalities (0.27y, 0.26y) on risk factors associated to CVD, which varies by region.

To conclude, the relative gains from these risk factors are higher than their relative weight on CVD and total deaths. Therefore, these risk factors are an efficient target for health policy with two aims: Reducing all-cause and CVD mortality inequalities and reducing the burden of CVD mortality.

This paper explores the classification of "Deaths of Despair" (DoD), focusing on mortality due to alcohol consumption, drug use, and suicides. This work aims to analyse the magnitude of these causes of death in Italy, a country previously unexplored, and to examine their joint trends to assess whether they can indeed be treated as a single group of causes of death. We employ the Potential Gain in Life Expectancy to evaluate the impact of these causes on overall health in Italy, then investigate the identified time series leveraging Cointegration Analysis. Utilising ISTAT data from 1983 to 2018, aggregated by gender and age groups at the NUTS1 level, the study reveals a decline in DoD mortality in Italy, driven mainly by a reduction in alcohol-related deaths. The magnitude and trajectory of the analysed causes, especially drug-related deaths, offer valuable insights into the unique dynamics of DoD in Italy. Cointegration Analysis indicates that there is no significant dependency structure between these causes or across regions, with only a few exceptions. These findings suggest that in Italy, DoD should not be treated as a homogeneous group but as distinct outcomes of despair that require separate consideration in public health analyses and interventions.

Reducing socioeconomic disparities in mortality is a critical public health concern. Acknowledging the importance of socioeconomic disparities in health is crucial to establish effective systems for monitoring health inequalities. Educational level is the most widely recognized proxy for socioeconomic status. We used the Mortality Inequality Database 2019–2020. The database combines the National Register of Causes of Death, covering all deaths in 2019 and 2020, and the National Base Register, which includes demographic data (i.e., place of residence and education level). We estimated Age-Standardized Mortality Rate Ratios between education levels, by sex, age group, cause of death, and region, and used quantile regression models to analyze regional mortality disparities in Italy, accounting for these variables. The link between education and mortality varies by cause of death, sex, and age. Across the Italian regions, individuals with lower education levels face mortality rates 1.5 to more than 3 times higher than those with higher education. Males experience more pronounced inequalities in most of the regions, though for some causes (e.g., diabetes and COVID-19), females report higher inequalities. The implementation of policies that support education is crucial for promoting health equity and protecting the most vulnerable.

Mortality inequalities are socially patterned and increasing in many populations. Typically, social mortality gradients are measured by gaps in survival between predefined groups across one socioeconomic status dimension, or as a gradient across a population ranked by social position. In interpreting trends, the implicit assumption is that the distribution of mortality-relevant personal characteristics that underlie social positions is static. We know that this is not the case. Populations change. The distribution of health-relevant personal characteristics in (sub)populations rises and falls over age, and change markedly across birth cohorts. How they change is rarely the subject of deep investigation. The shift in the social sciences toward causal inference has led to a marked improvement in estimating the impact of social positions on mortality at the individual level. But, by treating changing characteristics as confounders, we lose sight of how compositional change in social characteristics impact mortality inequalities at the population level. In this paper, we argue that changing social gradients can only be properly understood by quantifying the dynamic compositional change happening among a myriad of social variables within and across cohorts.