Study focus
The article examines spatial variation in COVID-19 fatality risk across the 58 mainland municipalities of Attica, using data covering approximately 2.9 million cases and 7,396 deaths from October 2020 to March 2022.
Its central question is how socioeconomic conditions, housing characteristics, urban structure, and the natural environment are associated with intra-metropolitan differences in health vulnerability.
Methodological approach
The analysis combines two complementary stages:
- Bayesian disease mapping using the BYM2-INLA model to stabilise municipal risk estimates and represent their spatial structure.
- Spatial econometric models to investigate direct and spatially transmitted relationships between mortality and social, housing, environmental, and epidemiological factors.
This framework helps distinguish random variation in small populations from systematic spatial patterns of risk.
Key findings
The findings show that health vulnerability is unevenly distributed across metropolitan space. In particular:
- adverse housing conditions and overcrowding are associated with higher risk,
- epidemic pressure remains a robust correlate of mortality,
- income and socio-spatial inequality intensify vulnerability,
- natural land cover displays a protective association,
- municipalities’ position within the metropolitan network and neighbourhood relationships shape local resilience.
The study argues that public-health and urban-resilience policies should consider not only the characteristics of individual municipalities but also the spatial interdependencies operating across the metropolitan region.
Suggested citation
Psycharis, Y., Pantazis, P., & Doukissas, L. (2026). Beyond the pandemic: Spatial inequalities, health vulnerability, and urban resilience in Attica. Cities, 176, 107161. https://doi.org/10.1016/j.cities.2026.107161