Objective To construct an air quality health index (AQHI) for Taiyuan City, China based on the impact of air pollution on non-accidental mortality, compare AQHI with the air quality index (AQI), and evaluate their abilities to predict health risks.

Methods Data on air pollutant concentrations, daily mortality, and meteorological factors in Taiyuan City from 2017 to 2021 were collected. A generalized additive model was used to analyze the exposure-response relationships between air pollutants, meteorological factors, and non-accidental mortality risk. AQHI was constructed using both standard method and principal component analysis. AQHI was compared with AQI to evaluate their abilities to predict health risks.

Results With the strongest lag effect, for every 10 μg/m³ increase in the concentrations of PM_2.5, SO₂, NO₂, and O₃, non-accidental mortality increased by 0.56% (95% confidence interval CI: 0.26%-0.87%), 1.43% (95%CI: 0.97%-1.91%), 0.91% (95%CI: 0.27%-1.54%), and 0.34% (95%CI: 0.06%-0.63%), respectively. The strongest lag effects of the first and second principal components of air pollutants on non-accidental mortality were 0.95% (95%CI: 0.51%-1.38%) and 0.48% (95%CI 0.24%-0.72%), respectively. Compared with AQI, the predictive ability of AQHI constructed using both method was improved, with the AQHI based on the standard method better reflecting the impact of air pollutants on non-accidental mortality in local residents.

Conclusion The atmospheric pollutants PM_2.5, SO₂, NO₂, and O₃ in Taiyuan City can affect the risk of non-accidental mortality, with SO₂ showing the strongest lag effect on non-accidental mortality across different populations. Compared to the AQHI constructed using principal component analysis, the AQHI built using the standard method can better predict the impact of air pollution on the risk of non-accidental mortality.