Effects of Socio-economic Inequalities in Childhood Respiratory Infections

Project Information

ESICRI

Grant agreement ID: 101106261

DOI

10.3030/101106261

Project closed

EC signature date 17 July 2023

Start date 1 April 2024

End date 31 March 2026

Funded under

• Marie Skłodowska-Curie Actions (MSCA)

Total cost

No data

EU contribution

€ 230 774,40

Coordinated by

KOBENHAVNS UNIVERSITET
Denmark

Periodic Reporting for period 1 - ESICRI (Effects of Socio-economic Inequalities in Childhood Respiratory Infections)

Reporting period: 2024-04-01 to 2026-03-31

Summary of the context and overall objectives of the project

Respiratory infections (RIs) are the most common health problem in childhood, yet substantial inequalities in their burden persist across Europe. Children from disadvantaged socio-economic circumstances experience higher risks of severe RI outcomes, including hospitalisation, but the causal pathways underlying these inequalities remain poorly understood. Addressing this knowledge gap is essential for designing effective and equitable public health interventions. The Effects of Socio-economic Inequalities in Childhood Respiratory Infections (ESICRI) project aims to quantify and explain socio-economic inequalities in childhood RIs across Europe using the EU Child Cohort Network (EUCCN), which includes harmonised data on more than 200,000 children from 16 birth cohorts. Leveraging recent advances in federated analysis through DataSHIELD, ESICRI will conduct privacy-preserving multi-cohort analyses to investigate how social conditions and early-life risk factors jointly shape infection risk. The project has two main research objectives: (1) to characterise social and geographic patterns in childhood respiratory infections across Europe, and (2) to determine whether early-life environmental, behavioural, and psychosocial factors mediate the relationship between socio-economic circumstances and RI outcomes. ESICRI will address these objectives through three integrated work packages. First, large-scale federated epidemiological analyses will quantify inequalities in upper and lower respiratory tract infections and test interactions between socio-economic indicators and early-life exposures. Second, the project will develop dsMediation, a novel open-source DataSHIELD package enabling privacy-preserving multi-cohort causal mediation analysis. Third, the new functionality will be used to identify and quantify mediating pathways linking socio-economic disadvantage to childhood RIs. Methodologically, ESICRI goes beyond the state of the art by combining life-course epidemiology, causal inference, and federated analytics in a uniquely large European dataset. The project will implement both one-stage and two-stage individual participant data meta-analyses and apply counterfactual mediation methods within a secure distributed computing environment. Gender differences will be explicitly examined through adjustment and stratified analyses. ESICRI will deliver the first large-scale European evidence on mechanisms driving socio-economic inequalities in childhood respiratory infections and will provide the research community with a reusable open-source tool for federated mediation analysis. The findings could potentially be used to inform European policies aimed at reducing child health inequalities and improving early-life infection prevention strategies.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The project delivered substantial scientific and technical progress across all planned work packages. Using harmonised data from 16 European birth cohorts within the EU Child Cohort Network, I conducted federated analyses to quantify socio‑economic inequalities in childhood respiratory tract infections. These analyses included descriptive exploration, regression modelling, and one‑stage and two‑stage individual‑participant data meta‑analytic models executed through the DataSHIELD platform. The results consistently showed socio‑economic gradients in both upper and lower respiratory tract infections across childhood, with notable heterogeneity across cohorts. This work generated a complete set of scientific findings.
The project also achieved its technological objective through the development of the dsMediation package, a DataSHIELD‑compatible tool enabling privacy‑preserving causal mediation analysis across distributed datasets. The package implements four methodological approaches (simulation, regression, weighting, and imputation) and includes bootstrap procedures for uncertainty estimation. It was extensively tested using synthetic and publicly available datasets to ensure accuracy, stability, and reproducibility. The package is fully functional and has already been adopted in multi‑cohort studies within the EU Child Cohort Network.
Building on these developments, the project advanced the analysis of early‑life risk factors as potential mediators of socio‑economic inequalities in childhood respiratory infections. I identified mediators available across cohorts, developed the analytical framework using Directed Acyclic Graphs, and conducted preliminary mediation analyses using the dsMediation package. These analyses confirmed the feasibility of the planned models and the suitability of the harmonised datasets for mediation analysis.
In parallel, the project delivered all planned scientific training and capacity‑building activities. I gained advanced expertise in mediation analysis, causal inference, and multi‑cohort epidemiology through specialised courses, hands‑on analytical work, and close collaboration with methodological experts during the secondment at the University of Turin. These activities strengthened the technical foundations required to complete the project’s scientific objectives and contributed to the successful development and application of the dsMediation package.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The project generated three main scientific and technical results. First, it produced large‑scale federated evidence on socio‑economic inequalities in childhood respiratory infections across 16 European birth cohorts, offering a detailed understanding of how early‑life socio‑economic conditions relate to infection patterns. Second, it delivered the dsMediation package, a fully functional tool enabling privacy‑preserving causal mediation analysis within the DataSHIELD ecosystem. Third, it established and applied a multi‑cohort analytical framework to investigate early‑life mediators of socio‑economic inequalities in childhood respiratory infections, with final mediation analyses approaching completion.
These results have the potential to advance research on early‑life determinants of infection risk and health inequalities, and to support wider use of federated causal inference methods in multi‑cohort epidemiology. The dsMediation package, in particular, enables analyses that were previously not feasible in distributed data settings and can be applied to a broad range of health outcomes.
To ensure further uptake, the main needs relate to continued methodological development and application. These include extending the dsMediation package to additional mediation approaches, integrating it into broader federated analytical workflows, and applying the methods to other outcomes within cohort consortia. No regulatory, commercialisation, or IPR‑related needs were identified, as the outputs are scientific and methodological and were intentionally released as open‑source resources.