Objective To construct an air health index (AHI) based on the assessment of effect of co-exposure to air pollution and non-optimum ambient temperature on non-accidental mortality in Huai′an, China and to assess its health risk prediction ability.

Methods The daily average data of the six major atmospheric pollutants (fine particulate matter, inhalable particulate matter, sulfur dioxide, nitrogen dioxide, carbon monoxide, and daily maximum 8 h average ozone concentration), air quality index (AQI), and meteorological factors (average ambient temperature and relative humidity), as well as the monitoring data of daily mortality in Huai′an from 2017 to 2021 were collected, and these data were collated into a time-series database. The data from 2017 to 2019 were used to establish the exposure-response relationship of atmospheric pollutants and non-optimum ambient temperature to non-accidental mortality. Based on the obtained exposure-response coefficients, the daily excess risk (ER) due to co-exposure to air pollution and non-optimum ambient temperature was calculated and exponentiated to a 0-10+ value for construction of the AHI. In order to effectively evaluate the health risk prediction ability of the AHI, an air quality health index (AQHI) model that considered solely the health risk of air pollution was also constructed. The health risk prediction ability of the AHI was evaluated using the data from 2020 to 2021. By pooling the constructed AHI and AQHI and the time series data from 2020 to 2021, the associations between the AHI, AQHI, and AQI and multiple health outcomes in 2020—2021, as well as the coefficient of determination (R²) and generalized cross validation (GCV) values of the model were quantitatively analyzed and compared.

Results The ER for the effect of each inter quartile range (IQR) increase in the AHI on non-accidental, cardiovascular, and respiratory mortality was higher than the corresponding values for the AQHI and AQI. The values of ER (95% confidence interval CI) for the effect of each IQR increase in the AHI on non-accidental, cardiovascular, and respiratory mortality were 7.394 (5.979, 8.828), 10.289 (8.605, 12.000), and 9.848 (7.200, 12.561), respectively. The values of ER (95%CI) for the effect of each IQR increase in the AQHI on non-accidental, cardiovascular, and respiratory mortality were 6.179 (4.822, 7.553), 5.827 (4.222, 7.456), and 7.791 (5.010, 10.645), respectively. The values of ER (95%CI) for the effect of each IQR increase in the AQI on non-accidental, cardiovascular, and respiratory mortality were 5.697 (5.653, 5.742), 5.493 (5.440, 5.546), and 4.963 (4.903, 5.034), respectively. The model fit goodness-of-fit metrics (R² and GCV values) for the AHI, AQHI, and AQI had been found to be almost identical.

Conclusion The AHI constructed based on the combined health effects of air pollution and ambient temperature in Huai′an has a good health risk prediction ability.