Construction and validation of an air quality health index based on population-weighted exposure level in Urumqi, China

Objective  To construct an air quality health index (AQHI) that can reflect health hazards based on population-weighted exposure level (PWEL) and the number of daily outpatient visits in Urumqi, China.

Methods  Related data of Urumqi in 2018—2023 were collected, including daily pollutant exposure data (CO, SO₂, NO₂, PM_2.5, PM₁₀, O₃-8 h), air quality index(AQI) meteorological data (mean temperature, mean air pressure, mean relative humidity, and daily mean wind speed), raster data on resident population distribution (1 km×1 km), and the number of daily outpatient visits in designated medical institutions. Daily PWEL and the number of daily outpatient visits were calculated to construct a Poisson generalized additive multi-pollutant model, and daily AQHI was obtained and validated based on exposure-response relationship.

Results  Compared with the mean concentration of the original site, there were increases in the annual PWEL concentrations of all pollutants except SO₂ and O₃-8 h in 2018—2023. The multi-pollutant lag model showed that the moving average lag01 d of NO₂ yielded the most substantial effect, with an ER value of 47.17% (95% confidence interval CI: 38.06%-56.28%). Moreover, the single-day lag of PM_2.5-lag7 showed the most significant impact, with an ER value of 2.25% (95% CI: 0.26%-4.25%), while PM₁₀, SO₂, and O₃ showed no significant impact on the number of daily outpatient visits. PM_2.5-lag7 and NO₂-lag01 were used to construct the AQHI, and the results of time-series cross validation showed a mean absolute error (MAE) of 5.18 and a mean root mean square error (RMSE) of 5.57. Furthermore, daily AQHI was highly positively correlated with AQI (r=0.738, P < 0.01), and the partial correlation analysis showed that AQHI was positively correlated with the number of daily outpatient visits in different types of medical institutions (P < 0.01).

Conclusion  AQHI based on PWEL is more accurate than AQI in predicting the degree of air pollution and can reflect the potential health impacts associated with air pollution.