Integrating air pollution-health feedback into climate projections: towards endogenous environmental-social links in...

Integrated assessment models (IAMs), often coupling Shared Socioeconomic Pathways (SSPs) and Representative Concentration Pathways (RCPs), simulate how socioeconomic drivers, technology, policy, and environmental processes interact over time. However, these models typically treat socioeconomic drivers as exogenous input, overlooking how environmental outcomes, like air pollution, can in turn affect health and demographics. This limits our understanding of health co-benefits and weakens the basis for climate-health policy integration. Here, we tackle this gap by linking ambient PM_2.5 concentrations from four SSP-RCP scenarios to the cause-specific risk functions and use the resulting risk impacts to adjust the age- and sex-specific demographic projections from the SSPs. This allows for more coherent estimation of how air quality trajectories influence health outcomes across 186 countries and territories through 2050. Our results reveal notable deviations from conventional SSP-based projections. In low-emission scenario (SSP1-1.9), PM_2.5-related deaths over 2020-2050 are overestimated by 8 % (10 million) due to improved air quality. In contrast, deaths are underestimated by 6 % (15 million) in high-emission scenario (SSP3-7.0), where pollution worsens. These differences translate into life expectancy at birth changes of +0.23 and -0.16 years, respectively. The feedback effects are pronounced in Southeast Asian countries with elevated pollution exposure and population vulnerability, exacerbating the Global North-South mortality gaps under SSP3-7.0 while narrowing them in SSP1-1.9/2.6. Our findings underscore the need and potential of incorporating air pollution-health feedback into the integrated modeling frameworks, which would enhance the realism of long-term demographic projections, especially in pollution-prone regions, and support better-aligned climate and public health strategies.