Integrating social and environmental determinants to measure air pollution in European metropolitan areas

Air pollution is one of the most pressing public health challenges of our time, yet existing indices rarely integrate the biological dimensions of pollutant exposure with the social and environmental contexts that shape them. Exposure to air pollutants is not randomly distributed across populations: it reflects deeper structural inequalities rooted in urbanisation patterns, economic development, and demographic composition. This paper presents a novel biosocial methodology for simultaneously modelling six major air pollutants (PM10, PM2.5, Sulphur Dioxide, Nitrogen Dioxide, Carbon Monoxide, and ground-level Ozone), explicitly incorporating social, demographic, meteorological, and geographical determinants. Key drivers of pollution are identified through a Multivariate Random Forest regression, yielding a parsimonious set of significant predictors including car ownership, latitude, temperature, air pressure, labour market participation, and youth dependency ratios. Structural Equation Models (SEMs) move beyond established single-pollutant measures by modelling the complex interdependencies among pollutants and between pollutants and the social and economic features of 130 European metropolitan areas. In this approach, exogenous variables serve not merely as controls, but as policy-relevant levers linking societal structure to biological exposure risk. The hierarchical SEM allows us to propose a novel Air Pollution Index (API). Sensitivity analyses through Monte Carlo simulation confirm the index’s robustness. A model-based cluster analysis further illustrates how social and geographic inequalities shape differential exposure across European cities, revealing distinct pollution profiles that closely mirror broader patterns of socioeconomic disadvantage. This biosocial framework demonstrates how statistical modelling of biological exposure data, anchored in social determinants, can produce more equitable, actionable, and statistically rigorous tools for environmental health policy.