João Simões, Alexandra Bernardo, Luísa Lima Gonçalves, José Brito
Air pollution and temperature are important environmental risk factors for mortality, but variation in risks across causes of death, population groups, and co-exposure conditions remains incompletely characterised. This study used nationwide monthly data from Portugal for the period 2010-2022 to estimate cause-specific mortality risks associated with nitrogen dioxide (NO2), fine particulate matter (PM2.5), and temperature, while evaluating sex-age interaction structures and selected co-exposure patterns. Mortality from malignant tumours, circulatory diseases, and respiratory diseases was modelled using Generalised Estimating Equations to account for temporal correlation and interaction effects. Pollutant exposures were defined as concentrations exceeding the 2021 World Health Organization air quality guideline values. Associations varied by cause of death. Respiratory mortality showed the strongest environmental association, with a 1.4% increase in risk per 1 μg/m3 increase in NO2 above the guideline value (RR 1.014, 95% CI 1.009-1.019), and the highest predicted risk was observed among males aged ≥65 years (RR 142.610, 95% CI 142.200-143.020). Circulatory mortality was also associated with NO2 (RR 1.005, 95% CI 1.001-1.008) and showed a submultiplicative sex-age interaction (RR 0.344, 95% CI 0.343-0.344), indicating attenuation of sex differences at older ages. Malignant tumour mortality showed weaker environmental associations but a greater-than-multiplicative sex-age interaction (RR 1.117) and attenuation of the NO2 association at higher PM2.5 levels, although the pollutant-related findings were less robust in sensitivity analyses. These findings indicate that mortality risks are not uniform across causes of death or population groups, supporting the value of locally derived, demographically stratified estimates for environmental health assessment and targeted public health strategies.