Menu
Menu
Menu
Menu
Background: Mental health is a public health issue for European young people, with great heterogeneity in resource allocation. Representative population-based studies are needed. The Global Burden of Disease (GBD) Study 2019 provides internationally comparable information on trends in the health status of populations and changes in the leading causes of disease burden over time. Methods: Prevalence, incidence, Years Lived with Disability (YLDs) and Years of Life Lost (YLLs) from mental disorders (MDs), substance use disorders (SUDs) and self-harm were estimated for young people aged 10-24 years in 31 European countries. Rates per 100,000 population, percentage changes in 1990-2019, 95% Uncertainty Intervals (UIs), and correlations with Sociodemographic Index (SDI), were estimated. Findings: In 2019, rates per 100,000 population were 16,983 (95% UI 12,823 – 21,630) for MDs, 3,891 (3,020 – 4,905) for SUDs, and 89·1 (63·8 – 123·1) for self-harm. In terms of disability, anxiety contributed to 647·3 (432–912·3) YLDs, while in terms of premature death, self-harm contributed to 319·6 (248·9–412·8) YLLs, per 100,000 population. Over the 30 years studied, YLDs increased in eating disorders (14·9%;9·4-20·1) and drug use disorders (16·9%;8·9-26·3), and decreased in idiopathic developmental intellectual disability (–29·1%;23·8-38·5). YLLs decreased in self-harm (–27·9%;38·3-18·7). Variations were found by sex, age-group and country. The burden of SUDs and self-harm was higher in countries with lower SDI, MDs were associated with SUDs. Interpretation: Mental health conditions represent an important burden among young people living in Europe. National policies should strengthen mental health, with a specific focus on young people. Funding: The Bill and Melinda Gates Foundation
The Global Burden of Disease Study produces annual estimates on prevalence, incidence and mortality for 369 diseases and injuries. Each update incorporates new data and methodological improvements to provide stakeholders with the most up-to-date information for resource allocation decisions and are compliant with the Guidelines for Accurate and Transparent Health Estimates Reporting.14 The present study employed estimates from GBD 2019, which are available on the Global Health Data Exchange (GHDx).15 These estimates supersede those from previous rounds of GBD, since the estimates for the whole time series are updated on the basis of addition of new data and change in methods, where appropriate, at each iteration of the GBD study.3 Methods for the generation of GBD 2019 estimates are described in detail elsewhere,3 while the methodology for estimating the burden due to mental health conditions is briefly summarised here. We provided results from 1990 to 2019 for 31 European countries: the 28 EU countries (the UK was still part of EU), plus Iceland, Norway and Switzerland, as part of Schengen area. The catchment population comprised young people between 10 and 24 years old,16 with a total number of 85 million subjects in 2019. The estimates were based on data on incidence and prevalence identified through systematic searches of published and unpublished documents, survey microdata, administrative records of health encounters, registries, and disease surveillance system that are catalogued in the Global Health Data Exchange website (http://ghdx.healthdata.org). We summarized these data sources for the 31 countries of interest, related to MDs, SUDs and self-harm in the 10-24 age range, in Appendix (Overview on data coverage). We included the following measures of disease burden: prevalence (MDs and SUDs), incidence (self-harm,) YLDs, and YLLs. YLDs are years lived with disability (in which the disability equates to a fraction of a year lived in full health) and are the product of the prevalence and the disability weight of that condition. YLLs are years of life lost due to premature death, calculated as the difference between the corresponding standard life expectancy for that person’s age and sex, and the age of actual death. Disability-adjusted life years (DALYs) are the sum of YLDs and YLLs. DALYs were used only to provide the fraction of YLDs and YLLs for each disorder. Prevalence was derived from estimates of point prevalence for all MDs and SUDs, with the exception of bipolar disorders, where one-year prevalence was applied.17 We used prevalence estimates for all conditions which usually last more than six months. This involved also SUDs, even if a small degree of them also contributed also to premature deaths. We used incidence estimates for self-harm, since the great majority of the burden due to self-harm was represented by YLLs due to fatal self-harm.18 Prevalence and incidence were modelled using DisMod-MR 2.1, a Bayesian meta-regression tool. Epidemiological data from different sources were pooled by DisMod-MR 2.1 with the goal of producing internally consistent estimates of prevalence, incidence, remission, and excess mortality by age, sex, location, and year. Proportions of severity were calculated to reflect the different levels of disability, or sequelae, associated with a determinate disorder, eg, mild, moderate, and severe presentations. Severity proportions, as shown elsewhere,10 were applied to the total prevalent cases estimated by DisMod-MR 2.1 to obtain prevalence estimates for each level of severity. As described in detail in other studies based on GBD 2019,3,19 disability weights by condition were applied to estimate YLDs. These have been calculated through a series of severity splits, which definie the sequelae of a health condition as asymptomatic, mild, moderate, and severe. Disability weights derived from different international surveys, where a scale ranging from perfect health (0) to death (1) was used, adding also population health equivalence questions that compared the lifesaving benefits and the prevention programmes for several health states. The analysis of the surveys served for the relative position of health states to each other, while the population health equivalence questions were used to assess those relative positions as values on a scale ranging from 0 to 1. More information on the sequela-specific health state descriptions and on the disability weights analysis are described elsewhere.10 A simulation method based on simulated populations of individuals by location, age, sex, and year, was used to adjust for comorbidity, since the burden attributable to each cause in GBD was estimated separately. Individuals in each population were exposed to the independent probability of having a combination of different sequelae in GBD 2019. A comorbidity correction was then used to estimate the difference between the average disability weight of individuals experiencing one sequela and the multiplicatively combined disability weights of those experiencing more sequelae. Specific YLDs per location, age, sex, and year applied the average comorbidity correction calculated for each sequela.10 Uncertainty intervals (UIs) were used to describe the point estimates of uncertainty from model specification, stochastic variation, and measurement bias. UIs are based on 1000 draws from the posterior distribution of estimates. The point estimate is defined by the mean of the draws, while the the 95% UIs is represented by the 2·5th and 97·5th percentiles ranked estimates from the drawns. In GBD 2019, diseases and injuries and causes of death, were aggregated in three Level 1 causes (communicable, maternal, neonatal, and nutritional conditions; NCDs; and injuries), 22 Level 2 causes, 174 Level 3 causes, and 301 Level 4 causes.3 In this study, we included Level 2 (MDs and SUDs) and Level 3 causes, as follows: Only for MDs, we also aggregate disorders as follows, to describing prevalence rates among the 31 countries of interest: Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) or the International Classification of Diseases – Tenth revision (ICD-10) criteria were used for definition of cases,19 as they were used by the majority of mental health surveys included in the Appendix. the. SDI is a composite indicator of development status, built as the geometric mean of 0 to 1 indices of total fertility rate in women younger than 25 years, mean education for the population aged 15 years and older, and lag-distributed income per capita.3 We used the SDI for each of the 31 countries of this study. For each country, cause and year, we report count, age rates per 100,000 population for age subgroups, and percentage changes from 1990 to 2019 for estimates of prevalence, incidence, YLDs, and YLLs, with 95% UIs,3. The Institute for Health Metrics and Evaluation (IHME) provided aggregated estimates for all 31 countries combined, since the standard GBD aggregate estimates are for the EU, and exclude the other Schengen area countries (i.e. Iceland, Norway and Switzerland) .. Results are presented by sex and age subgroups (10-14; 15-19 and 20-24 years). YLLs were calculated only for self-harm, eating disorders, alcohol use disorders and substance use disorders since these are the only causes considered causes of death in the WHO/ICD-system (https://www.who.int/standards/classifications). We also reported the percentages of YLDs and YLLs of MDs, SUDs and self-harm in the 10-24 age groups compared to the all-causes GBD in the 31 European countries. In addition, we performed Spearman rank-correlations to study the relation between SDI and prevalence rates of MDs and SUDs, and incidence rates of self-harm. We set P-value <0·05 as the threshold of statistical significance. These analyses were conducted with Stata/BE 17.0 (StataCorp LLC, College Station, USA). The funder of the study had no role in study design, data collection, analysis and interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility to submit for publication.
