An 18-year study in the US shows that people who live in areas with high ground ozone levels face a 30-per cent greater risk of death due to respiratory problems.

Environmental scientists already knew that increases in ozone levels during periods of heavy pollution caused short-term ill effects such as asthma attacks, increased hospitalisations and deaths from heart attacks. But the new study is the first to show that long-term, low-level exposure to the pollutant can also be lethal.

The research, published in the New England Journal of Medicine, said the issue is not just the effects of high-ozone days but the effects of long-term cumulative exposure, New York University Langone Medical Center said Thursday in a release.

Lead author Michael Jerrett of the University of California, Berkeley, said background levels of ozone have at least doubled since the pre-industrial-revolution. The findings are based on American Cancer Society data that followed 450,000 people from 1982 to 2000, covering 96 metropolitan regions in the US.

Riverside in California had the highest mean daily maximum ozone concentration during the 18-year period with 104 ppb. Los Angeles was a close second, with an estimated 43 per cent increased risk, the report said.

In contrast, San Francisco had the lowest average ozone level (33 ppb) of the 96 regions studied and only a 14-per cent increased risk, probably because of the fog and prevailing winds that reduce ozone formation. The Pacific Northwest also had low levels of ozone, again because of rain and cool weather.

Ozone, a variant form of oxygen, forms a protective layer from the sun's ultraviolet radiation when located in the earth's upper atmosphere. However, that same gas is toxic at ground level where it can be breathed by humans.

Ozone is what is known as a secondary pollutant; it is not formed directly by the burning of fossil fuels. Rather, nitrogen oxides produced by such combustion react in the presence of sunlight to form ozone. It is thus the biggest problem in areas that are sunny and hot, Jerrett said.

As an oxidising agent, ozone reacts with virtually anything it comes into contact with. In particular, it reacts with cells in the lungs, causing inflammation and a variety of other effects that lead to premature aging.

The study found that for every 10 parts-per-billion (ppb) increase in ozone level, there is a four per cent increase in risk of death from respiratory causes, primarily pneumonia and chronic pulmonary obstructive pulmonary disease.

The study analysed the risk of death for both ozone and fine particulate matter, two of the most prevalent components of air pollution. The researchers found that people living in areas with the highest concentrations of ozone, such as the Los Angeles metropolitan area and California's Central Valley, had a 25 to 30 per cent greater annual risk of dying from respiratory diseases compared with people from regions with the lowest levels of the pollutant.

"This is the first time we've been able to connect chronic exposure to ozone, one of the most widespread pollutants in the world, with the risk of death," said Jerrett. "World Health Organisation data indicate that about 240,000 people die each year from respiratory causes in the United States."

''Globally, some 7.7 million people die from respiratory causes, so worldwide the impact of ozone pollution could be very large," he added.

The findings come a year after the US Environmental Protection Agency (EPA) strengthened its National Ambient Air Quality Standards for ground-level ozone from an annual average of 80 ppb to 75 ppb to reflect growing evidence of the harmful health effects of ozone.

A month after the EPA released its new standards, a National Research Council report concluded that premature deaths related to ozone exposure of less than 24 hours are more likely among those with pre-existing diseases, said the university release.

Current standards for ozone pollution cover only eight-hour averages of the gas, but even with that relatively relaxed rule, 345 counties with a total population of more than 100 million people are out of compliance.

The EPA may need to implement an annual standard, said University of Ottawa environmental health scientist Daniel Krewski, one of the paper's authors.

The researchers found no increase in deaths from cardiovascular disease associated with ozone levels - those deaths are caused primarily by the fine particulates present in air pollution.

They also found no increase in overall mortality, suggesting that ozone is causing deaths in people who were probably going to die in another year or two anyway, according to epidemiologist Joel Schwartz of the Harvard School of Public Health, who was not involved in the study.

"We do know that ozone is particularly dangerous for people living with existing asthma or lung disease," Jerrett said. And it didn't matter what someone's weight, income or education was. "It seems to affect a lot of people relatively equally."