Home » Ozone » Level 3 » Question 3

Air Pollution Ozone

3. How are we exposed to Ozone (O3)?

  • 3.1 Critical sources of Ozone responsible for health effects
  • 3.2 Relationship between ambient levels and personal exposure to Ozone
  • 3.3 Short-term exposure to high peak levels or exposure in hot spots of Ozone

3.1 Critical sources of Ozone responsible for health effects

The source document for this Digest states:


Ozone is a secondary pollutant produced by photochemical activity in the presence of precursors. The working group felt that it was beyond its core competence to give a detailed description of ozone formation and dispersion patterns.


Ozone is formed in the troposphere by photochemical reactions in the presence of precursor pollutants such as NOx and volatile organic compounds. Where there is an abundance of NO, O3 is “scavenged” and as a result its concentrations are often low in busy urban centres and higher in suburban areas. O3 is also subject to long-range atmospheric transport and may be considered as a trans-boundary problem.

Source & ©: WHO Regional Office for Europe  "Health Aspects of Air Pollution" (2003), Chapter 6 Ozone (O3), Section 6.2 Answer and rationales, Question 10

The same information on
Particulate MatterNitrogen Dioxide

3.2 Relationship between ambient levels and personal exposure to Ozone

The source document for this Digest states:

Can the differences influence the results of studies?


Personal exposure measurements are not well correlated with ambient fixed site measurements. To account for that, in some studies, additional information (e.g., activity patterns) was used to improve personal exposure estimates based on fixed site measurements. Being a highly reactive gas, O3 concentrations indoors are generally lower (less than 50%) than those in ambient air. There are very few indoor sources in most homes (such as xerographic copiers, electrostatic air cleaners). Outdoor O3 levels vary across city areas because O3 is scavenged in the presence of NO. Early morning and late night exposures outdoors are lower because of the diurnal cycle of ambient O3. Thus, for O3, cumulative daily or long-term average exposures are largely determined by exposures occurring outdoors in the afternoon. The studied effects of exposure misclassification are in the direction of underestimation of O3 exposure effects and may conceal real effects.


The spatial variability of ozone levels may be low within large areas. This is obviously an obstacle in designing epidemiological studies built on differences in exposure of different communities, but favours the use of fixed site monitors to characterize exposure levels for large populations, both in studies with spatial and temporal contrast. However, there are gradients within cities, due to the reaction of ozone with NO emitted from traffic and other combustion sources. There may even be a substantial variation between neighbouring residential areas, as measured by front-door samples (286). In addition, there is a strong diurnal variation, with the highest levels usually in the afternoon. Further, ozone levels are commonly much lower indoors than outdoors. Short-term personal exposure measurements are thus not well correlated with ambient fixed site measurements (286). The use of outdoor ozone concentration from fixed site monitors, as a measure of short-term ozone exposure in epidemiological studies, may, therefore, result in misclassification error, both in studies with temporal or spatial contrasts.

However, the temporal correlation was in one study found to vary among subjects, due to the activity pattern, geographical variables, home variables such as ventilation and the distance from the monitoring station and traffic (287). In spite of the poor temporal correlation on the individual level, in the largest follow-up study on O3 exposure, the differences in average levels between communities were similar when outdoor measurements or personal measurements were used, but only during the ozone season, which is warm. The reason for this is probably that people spend more time outdoors and that the differences between outdoor and indoor levels are smaller, due to open windows. This finding is relevant for studies on long-term effects since – during the warm season – the outdoor measurement provides a valid estimate of the spatial variation provided time spent by subjects in the different areas was measured (288). It has also been shown that (128, 248, 288) having air conditioning decreases the personal O3 exposure level, and also its correlation with outdoor measurements. Most of these random misclassification effects cause true effects to be interpreted as less strong (100). It is, however, possible that the exposure errors are correlated to the exposure level, which would lead to a positive or negative bias. Systematic errors may also occur in studies of urban areas where the ozone levels are substantially lower in the city centres (spatial error). A few epidemiological studies have explicitly assessed the consequences of the poor correlation between personal exposure and the commonly used ozone levels measured at fixed sites. The misclassification error was found to bias the effect estimates towards the null hypothesis (289, 290).

Some of the studies on the long-term effects have tried to reduce spatial or temporal error by incorporating additional information to the outdoor measurements. In the AHSMOG study, individual cumulative exposure was calculated using monthly measurements from air monitoring stations in California, and distance from residence and work to the stations. This interpolation method was found to increase the validity of the exposure estimates (229). One Austrian study also calculated an individual ozone concentration weighting the outdoor measurements by the time spent in the area (234).

Source & ©: WHO Regional Office for Europe  "Health Aspects of Air Pollution" (2003), Chapter 6 Ozone (O3), Section 6.2 Answer and rationales, Question 9

3.3 Short-term exposure to high peak levels or exposure in hot spots of Ozone

The source document for this Digest states:


Adverse health effects have been documented after short-term exposure to peaks, as well as long-term exposure to relatively low concentrations of PM, ozone and NO2. A direct comparison of the health relevance of short term and long-term exposures has been reported for PM, but not for ozone and NO2. For PM, long-term exposure has probably a larger impact on public health than short-term exposure to peak concentrations.

Some studies have documented that subjects living close to busy roads experience more short- term and long-term effects of air pollution than subjects living further away. In urban areas, up to 10% of the population may be living at such “hot spots”. The public health burden of such exposures is therefore significant. Unequal distribution of health risks over the population also raises concerns of environmental justice and equity.


Ozone: Short-term versus long-term
There is ample experimental as well as epidemiological evidence that short-term (one to eight hours) exposure to peak levels of ozone is associated with transient reductions in lung function, with increased reporting of respiratory and eye symptoms, and with increased responsiveness to inhaled allergens. Recent contributions to our knowledge on this include a study among children with asthma (Gent et al., 2003) in which wheeze symptoms were found to increase significantly among maintenance medication users already at 1 hour ozone concentrations above 100 µg/m3 and a California winter study in which asthmatic children were found to experience more symptoms with increased ozone that never exceeded 104 µg/m3 as 1 hour maximum, and 74 µg/m3 as 8 hour maximum (Delfino et al., 2003). Eye, throat and nose irritation were found to increase with 8 hour ozone concentrations never exceeding 121 µg/m3 in asthmatic children studied in France (Just et al., 2002). Earlier work (e.g. Jorres et al., 1996) had already shown that ozone increases allergen responsiveness in subjects with mild asthma or rhinitis. Such discomfort and morbidity effects are different from effects of long-term exposure to ozone which have primarily been associated with reduced lung function (Künzli et al., 1997; Peters et al., 1999), and they are also different from the effects of ozone seen in time series studies, which focus on increased hospital admissions for respiratory and cardiovascular disease and in some studies increased mortality (Thurston et al., 2001).

As documented in the previous report (WHO, 2003), time-series studies find linear or near-linear relationships between day-to-day variations in peak ozone levels and health endpoints, down to low levels of exposure. As there are usually many more days with mildly elevated concentrations than days with very high concentrations, the largest burden on public health may be expected with the many days with mildly elevated concentrations, and not with the few days with very high concentrations.

No analyses have been published to compare the relative public health significance of the short- term and long-term effects of ozone.

Ozone: Hot spots versus background
Being a secondary pollutant, ozone concentrations are usually not significantly higher at specific urban “hot spots”. Higher concentrations can sometimes be detected in plumes downwind of strong emission sources of NOx and/or NMVOC during summertime, when photochemical ozone production is enhanced. On the contrary, ozone levels tend to be lower in polluted urban atmospheres where ozone is depleted due to reaction with freshly emitted NO, often from traffic sources. Because this is due to the presence of pollutants some of which are harmful to health, this observation has no practical public health implications. For most practical purposes, there is no urban “hot spots” issue when it comes to ozone.

Source & ©: WHO Regional Office for Europe  Health Aspects of Air Pollution - answers to follow-up questions from CAFE (2004), Question 4

Themes covered
Publications A-Z