Article Text

This article has a correction. Please see:

Download PDFPDF
Ozone, heat and mortality: acute effects in 15 British conurbations
  1. Sam Pattenden1,
  2. Ben Armstrong1,
  3. Ai Milojevic1,
  4. Mathew R Heal2,
  5. Zaid Chalabi1,
  6. Ruth Doherty3,
  7. Ben Barratt4,
  8. R Sari Kovats1,
  9. Paul Wilkinson1
  1. 1London School of Hygiene & Tropical Medicine, London, UK
  2. 2School of Chemistry, University of Edinburgh, Edinburgh, UK
  3. 3School of GeoSciences, University of Edinburgh, Edinburgh, UK
  4. 4King's College, London, UK
  1. Correspondence to Sam Pattenden, PEHRU, London School of Hygiene & Tropical Medicine, Keppel Street, London WC1E 7HT, UK; sam.pattenden{at}


Background Acute associations between mortality and ozone are largely accepted, though recent evidence is less conclusive. Evidence on ozone–heat interaction is sparse. We assess effects of ozone, heat, and their interaction, on mortality in Britain.

Methods Acute effects of summer ozone on mortality were estimated using data from 15 conurbations in England and Wales (May–September, 1993–2003). 2-day means of daily maximum 8-h ozone were entered into case series analyses, controlling for particulate matter with aerodynamic diameter of <10 μm, natural cubic splines of temperature, and other factors. Heat effects were estimated, comparing adjusted mortality rates at 97.5th and 75th percentiles of 2-day mean temperature. A separate model employed interaction terms to assess whether ozone effects increased on ‘hot days’ (where 2-day mean temperature exceeded the whole-year 95th percentile). Other heat metrics, and non-linear ozone effects, were also examined.

Results Adverse ozone and heat effects occurred in nearly all conurbations. The mean mortality rate ratio for heat effect across conurbations was 1.071 (1.050–1.093). The mean ozone rate ratio was 1.003 per 10 μg/m3 ozone increase (95% CI 1.001 to 1.005). On ‘hot days’ the mean ozone effect reached 1.006 (1.002–1.009) per 10 μg/m3, though ozone–heat interaction was significant in London only. On substituting maximum for mean temperature, the overall ozone effect reduced to null, though evidence remained of effects on hot days, particularly in London. An estimated ozone effect threshold was below current guidelines in ‘mean temperature’ models.

Conclusion While heat showed robust effects on summer mortality, estimates for ozone depended upon the modelling of temperature. However, there was some evidence that ozone effects were worse on hot days, whichever temperature measure was used.

  • Ozone
  • mortality
  • heat
  • interaction
  • time series
  • epidemiology
  • climate
  • mortality studies
  • time series study
  • air pollution

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • All authors contributed to the paper.

  • Funding This work was primarily funded through grant NE/E008593/1 in the Environment and Human Health programme of the Natural Environment Research Council.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • Correction
    BMJ Publishing Group Ltd