Composite indices for assessing cumulative environmental exposure: A methodological comparison at the municipal level in France

Capturing cumulative environmental exposure is a key challenge in environmental health and exposome research. Composite indices are increasingly used to integrate multiple environmental stressors, yet their construction relies on methodological choices that may substantially influence exposure assessment and territorial prioritization. We compared four commonly used aggregation methods for constructing composite indices of environmental exposure and assessed how these choices affect the characterization and ranking of the 34,816 municipalities of mainland France. Using a harmonized dataset of 30 environmental indicators covering air pollutants, pesticides, drinking water contamination, industrial emissions, extreme temperatures, radon exposure, and surrounding greenness, we constructed three continuous composite indices (unweighted sum, rank-based aggregation, and factor analysis of mixed data [FAMD]) and one clustering-based classification (k-prototypes). We assessed agreement between methods using distributional analyses, spatial patterns, and correlation metrics, and compared their ability to identify highly exposed municipalities. The three continuous indices showed moderate agreement (Spearman ρ = 0.63 to 0.78), but differed markedly in their distributions and contrast between municipalities. The rank-based index produced a more even spread of values, whereas the FAMD-based index yielded a compressed distribution. The clustering approach identified five distinct exposure profiles with clear spatial organization. The identification and prioritization of highly exposed municipalities varied across methods, although major urban areas were consistently classified among the most exposed. This study provides a comparative framework to guide the use of composite environmental indices for environmental health surveillance and research. Methodological choices in composite index construction are not neutral and can influence both the representation of cumulative environmental exposure and the prioritization of territories. Continuous indices and clustering approaches capture complementary dimensions of exposure and should be selected according to epidemiological or public health objectives. Our study identifies in the literature the most commonly used methods for constructing composite indices of cumulative environmental exposure. Based on the results of the literature review, we compare four index construction methods using the same dataset and compare them systematically. Our results indicate that, despite an overall consistency among the constructed indicators, the choice of method can substantially influence the results in terms of distributions. The construction methods identified in the literature serve distinct but complementary purposes and should be selected according to the objectives of the analysis. We provide guidance on the choice of method.