A GBD 2019 study of health and Sustainable Development Goal gains and forecasts to 2030 in Spain

This study aimed to report mortality, risk factors, and burden of diseases in Spain. The Global Burden of Disease, Injuries, and Risk Factors 2019 estimates the burden due to 369 diseases, injuries, and impairments and 87 risk factors and risk factor combinations. Here, we detail the updated Spain 1990–2019 burden of disease estimates and project certain metrics up to 2030. In 2019, leading causes of death were ischaemic heart disease, stroke, chronic obstructive pulmonary disease, Alzheimer’s disease, and lung cancer. Main causes of disability adjusted life years (DALYs) were ischaemic heart disease, diabetes, lung cancer, low back pain, and stroke. Leading DALYs risk factors included smoking, high body mass index, and high fasting plasma glucose. Spain scored 74/100 among all health-related Sustainable Development Goals (SDGs) indicators, ranking 20 of 195 countries and territories. We forecasted that by 2030, Spain would outpace Japan, the United States, and the European Union. Behavioural risk factors, such as smoking and poor diet, and environmental factors added a significant burden to the Spanish population’s health in 2019. Monitoring these trends, particularly in light of COVID-19, is essential to prioritise interventions that will reduce the future burden of disease to meet population health and SDG commitments.


Official projection of the population pyramids and of life expectancy up to 2100
IHME forecasting and life expectancy estimates are created by modeling the future population in reference and alternative scenarios. These are modeled as functions of fertility, migration, and mortality rates. First, statistical models were developed for completed cohort fertility at 50 years old (CC50). This indicator is used because it is more stable over time compared to the total fertility rate. CCF50 was then modelled as a time-series random walk function taking into account educational attainment as well as contraceptive met need. CC50 and covariates were then used to model age-specific fertility rates up to the year 2100. IHME considered underlying mortality, a risk factor scalar, and an autoregressive integrated moving average (ARIMA) model. Second, IHME modelled net migration as a function of the Socio-demographic Index, crude population growth rate, and deaths from war and natural disasters using the ARIMA model. IHME then developed a reference scenario and alternative scenarios based on contraceptive met need, educational attainment, and Spain's estimated gross domestic product. Past data inputs, model estimation, and forecast data distributions were used to develop forecast uncertainty intervals (UIs). Detailed techniques for forecasting methods are published elsewhere. 1

Population, fertility, mortality, and migration estimates
Data used to produce demographic assessments of the key indicators of fertility, mortality, migration, and population in Spain are derived from the following sources: o These data sources are synthesized and corrected for known biases using the GBD spatiotemporal Gaussian process regression (ST-GPR). ST-GPR then generates age-specific fertility rates for 5-year age groups between ages 15 and 49 years. This was extended to estimate age-specific mortality for groups 10-14 and 50-54, which was then aggregated to create the total fertility rate from 10 and 54. ST-GPR also estimates adult mortality as the probability of death between ages 15 and 60 years based on vital registration, sample registration, and sibling histories. 1000 draw-level estimates were produced by the demographic estimation processes to estimate uncertainty. Using a relational model life table system, IHME estimated HIV-free life tables using estimates of under-5 and adult mortality rates. A Bayesian hierarchical cohort component model analyzing estimated age-specific fertility and mortality rates were used to estimate annual and single-year age estimates of net migration and population. 1000 draw-level estimates were produced by the demographic estimation processes to estimate uncertainty. Detailed methods are described elsewhere. 2

Mortality data
The Cause of Death Ensemble model and spatiotemporal Gaussian process regression were used to calculate cause-specific death rates and cause fractions. To match the total all-cause deaths calculated as part of the GBD population, fertility, and mortality estimates (see section 2 above), IHME adjusted cause-specific deaths. To calculate YLLs, deaths were multiplied by standard life expectancy at each age. To make sure there was consistency between incidence, prevalence, remission, excess mortality, and cause-specific mortality for most causes IHME used DisMod-MR 2.1, a Bayesian meta-regression modelling tool. To calculate YLDs, prevalence estimates were multiplied by disability weights for mutually exclusive sequelae of diseases and injuries.
Results were considered in the context of the Socio-demographic Index (SDI). The SDI is a composite indicator including income, education, and fertility rate in women younger than 25 years. For each metric, IHME developed uncertainty intervals using the 25th and 975th ordered 1000 draw values of the posterior distribution. IHME does not have estimates for 37 years of cause of death data by cause, sex, and age for Spain by the 16 regions. This is because Spain is not one of the locations for which subnational GBD data are produced.
GBD uses vital registration with medical certification of cause of death for cause of death analysis. The GBD cause of death list includes cause of death data obtained using various revisions of the International Classification of Diseases and Injuries (ICD). Deaths that could not be the underlying cause of death (e.g. cardiopulmonary failure) or were inadequately specified (e.g. injury from undetermined intent were reassigned to the most probable underlying causes of death. Deaths were redistributed based on evidence from published studies or expert judgment, or statistical algorithms. 4 Detailed methods for mortality data are published elsewhere.