Air quality indicators: short term health effects

Sector Air quality, health
Description Estimated number of preterm deaths due to ozone short-term exposure.
End User Health authorities, environmental authorities, general public
Calculation method

Population data have been obtained for each city, region or country. For Stockholm national data for 2012, with a spatial resolution of 100×100 m2, have been obtained from Swedish statistics. For Bologna and Amsterdam/Rotterdam,  a 1 * 1 km2 population grid disaggregated data has been applied (Gallego 2010).

The data on baseline mortality are from national official sources, for Stockholm from Swedish statistics, for Bologna from the Bologna province statistics and for Amsterdam from Centraal Bureau voor de Statistiek.

Baseline mortality for all ages for the city or region is used in combination with population exposure data for the city according to the HIA tool AirQ developed by WHO (2004), where the attributed mortality is calculated as
∆Y = (Y0 * P) * (eβ*X – 1),
where Y0 is the baseline rate; P the number of exposed persons; β the exposure-response relationship (relative risk) and X the estimated mean exposure (with impact/above any assumed threshold). Calculations will build on the WHO HRAPIE recommendation assuming a 0.3% increase (95% CI 1.4 – 4.3) per 10 µg m-3 increase in daily maximum 8-hour ozone and with a cutoff at 35 ppb (70 µg m-3) (SOMO35).

ID Title Period Statistical processing Unit Threshold Comment
mortO3y Annual deaths due to O3
short-term exposure
yearly See above deaths per year    
mortO3ynorm Annual deaths per 100,000
inhabitants due to O3 shortterm
yearly Assuming 100 000 inh in each grid deaths per year/100,000
Provenance Theese indicators are based on output from the MATCH model
Validation The downscaling made by MATCH in Urban SIS has been validated against observations in Urban SIS deliverable 5.2, where an overview is given in Table 4.
Calculation caveats Spatial representation:
Other caveats:
Could be compared to:
Could be used with:

The WHO REVIHAAP project argues that despite the many respiratory outcomes associated with O3, mainly adverse health outcomes with known baseline rates are suited for health impact assessments. Evidence from time-series studies of short-term exposure to O3 suggest that health impact assessment calculations can be undertaken for a range of end-points, including all-age, all-cause mortality (WHO 2013a). There is still scientific debate whether the effects on mortality of long-term exposure to O3 are well enough documented to be included in health impact assessments.

Multi-pollutant models in the largest European study of short-term exposure (APHEA2) reported short-term exposure increases of total mortality by approx. 0.3% per 10 µg m-3 using the daily 8-h or 1-h maximum, in a linear manner without a significant threshold (Gryparis et al. 2004). A WHO meta-analysis for the AQ guidelines (2003) reported a relative risk of 0.3% per 10 µg m-3 increase with the 95% (CI 0.1–0.4%) which we see as a robust exposure–response assumption to apply.

WHO REVIHHAP conclude that the epidemiological evidence supports calculations that use all-year coefficients for daily maximum 8-h O3 (scaled from the 1-h measures reported in the literature), including adjustment for PM10. It is also recommended that health impact calculations for short-term exposures assume linear concentration–response relationships. Since the epidemiological evidence on linearity does not extend down to zero, appropriate cut-off points for health impact assessments are therefore recommended: at 10 ppb (20 µg m-3) for daily maximum 8-h O3 and at 35 ppb (70 µg m-3), for consistency with previous work using SOMO35 data (WHO 2013).

Given the uncertainties in the effects of long-term exposure to O3 (see the REVIHAAP report) it was suggested that health impact assessments for long-term exposure and respiratory and cardiopulmonary mortality are undertaken as a sensitivity scenario. It is recommended the coefficients from single pollutant models from the American Cancer Society cohort study (Jerrett et al. 2009) are used, assuming an association exists within the range of O3concentrations studied.

The WHO HRAPIE Project recommended use of a meta-coefficient from The APHENA Study (results from 32 European cities) of a 0.3% increase (95% CI 1.4 – 4.3) per 10 µg m-3 increase in daily maximum 8-h O3 (Katsouyanni et al. 2009) and cutoff at 35 ppb (SOMO35). 

Experience user  

Gallego FJ 2010: A population density grid of the European Union. Population and Environment. 31:6, 460-473.

Gryparis A et al. 2004: Acute effects of ozone on mortality from the ‘air pollution and health: a European approach’ project. American Journal of Respiratory and Critical Care Medicine, 170:10, 1080–1087.

Jerrett M et al. 2009: Long-term ozone exposure and mortality. The New England Journal of Medicine, 360:11, 1085–1095.

Katsouyanni K et al. 2009: Air pollution and health: a European and North American approach (APHENA). Boston, Health Effects Institute, Research Report 142.

WHO 2004: Tools for health impact assessment of air quality: the AirQ 2.2 software.

WHO 2013a: Review of evidence on health aspects of air pollution – REVIHAAP Project Technical Report. Copernicus Climate Change Service Urban SIS D4.3 Indicators for urban assessments, C3S_441 Lot3 Urban SIS, D4.3 22 Copenhagen.

WHO 2013b: Health risks of air pollution in Europe – HRAPIE. Recommendations for concentration-response functions for cost-benefit analysis of particulate matter, ozone and nitrogen dioxide. Copenhagen

WHO 2016. European Health For All Database.