BackgroundThere is limited evidence from multi-exposure models on the environmental impact on type 2 diabetes (T2D), considering complex interrelations between the exposures. We aimed to identify air pollution, temperature and greenness features associated with T2D prevalence in a large pan-European study, specifically predicting the impact of exposure change under different scenarios.MethodsWe included 10 European population-based cohorts within the EXPANSE project. Information on environmental exposures, covariates and T2D diagnosis at recruitment were obtained for 87,044 individuals, including 5,662 T2D cases. Data on air pollution, greenness, and temperature at residential addresses originated from uniform pan-European models. T2D identification was based on questionnaires, registers or biomarker measurements. Adjusted cross-sectional associations between exposures and T2D were assessed, and all exposure two-way interactions and square terms were a priori included in elastic net regression models to identify the most influential environmental factors and combinations. Results were contextualized within different environmental scenarios.ResultsPM10 and temperature annual mean and variability, as well as their interaction term, were consistently associated with T2D prevalence across strata of sex, body mass index, and urbanicity. Associations with greenness, PM2.5 and PM2.5-average temperature interaction were also observed but less consistently across strata. In the scenario analysis, reducing air pollution was related to lower T2D prevalence, while a beneficial association with increasing greenness was most consistently observed in the highly urbanized subgroup.ConclusionsOur findings support a role of particulate air pollution in T2D development and suggest that beneficial effects of increased greenness primarily occur in highly urbanized areas.