TCS Daily

Asthma and Air Pollution

By Joel Schwartz - September 26, 2005 12:00 AM

The prevalence of asthma rose by about 75 percent overall between 1980 and 1996, and by nearly a factor of two in children up to 17 years of age.[1] Prevalence seems to have leveled off since then.[2] Roughly 6 to 7 percent of American children currently have asthma.

What caused this large rise in asthma over the last few decades? One hypothesis is that people no longer get as many infections and other immune challenges in childhood, and that this somehow makes them more susceptible to allergic diseases such as asthma.[3] Obesity is also associated with a greater risk of developing asthma and the prevalence of obesity has been increasing.[4] Increasing indoor exposures to allergens may also play a role, as people spend more time indoors and in buildings with better insulation from outside air.[5] These hypotheses are the subject of continuing research.


However, among the potential causes of asthma, none has received as much popular attention as air pollution. Each year, dozens of news stories, reports by environmental groups and regulatory agencies, and scientific and medical journal articles claim or imply that air pollution plays a major role in whether a person develops asthma. Are they right? We need to distinguish two questions here. First, can air pollution cause people to develop asthma? Second, could air pollution be responsible for the rise in the prevalence of asthma during the last few decades?


The second question has a straightforward answer. Air pollutionat least the wide range of air pollutants we measureis not a plausible cause of rising asthma. Air pollution of all kinds has been declining at the same time asthma has been rising. Figure 1 shows California data for four pollutantsozone, particulate matter under 10 microns (PM10), nitrogen dioxide (NO2), and carbon monoxide (CO)whose ambient levels are regulated by federal standards. Other regulated pollutants have also been declining, for example, sulfur dioxide, lead, and PM2.5.[6]


Figure 1. Trends in air pollution vs. trends in asthma in California[7]


Notes: The lines are linear regression lines. Ozone, CO, and NO2 are the average of the top 30 daily readings for each year (ozone and CO peak 8-hour, NO2 peak 1-hour) across all monitoring sites for the given pollutant. PM10 (particulate matter under 10 microns in diameter) is the average of the annual-average PM10 readings for all monitoring sites. Only sites with data in every year throughout the time period for each pollutant were included in the analysis. Number of monitoring sites for each pollutant: NO2=57, CO=47, Ozone=68, PM10=29. Pollution declined not only on average, but at almost every individual monitoring site. The start of the time period (which ranges from 1984-1987) for each pollutant was chosen to maximize the number of monitoring sites included, while still overlapping the time period during which asthma prevalence rose. CO is listed in parts per ten million (pptm; divide by 10 in order to get parts per million) so that CO values fall within the same range as other pollutants. Ppb = parts per billion; ug/m3 = micrograms per cubic meter.


California measures other pollutants as well. For example, Figure 2 displays trend data for benzene, 1,3-butadiene, and benzo(a)pyrene. The first two are gases emitted mainly by gasoline engines. The third is an organic particulate emitted by diesel engines. All three have been going down at all sites where they are measured. Other air pollutants that have been declining include the gases acetaldehyde, xylene, perchloroethylene, methylene chloride, trichloroethylene, and styrene, and the particulates dibenz(a,h)anthracene, benzo(b)fluoranthene, and hexavalent chromium.[8]


Figure 2. Trends in levels of traffic-related air pollutants in California, 1990-2003 [9]


Notes: All trends include only monitoring sites that operated continuously from 1990-2003 (11 sites in all cases). 1990 was the first year that California began collecting data on these compounds. For each pollutant, levels declined not only on average, but at every individual site. Left axis: 1,3-butadiene, benzo(a)pyrene. Right axis: benzene. Ppb = parts per billion; ng/m3 = nanograms per cubic meter.

I used data from California, because it has one of the most extensive pollution monitoring networks in the country. But you could just as easily perform the same exercise in any other state and get the same result-declining air pollution, rising asthma.

Even though air pollution isn't responsible for the rise in asthma, could air pollution cause at least some people to develop asthma? The trend data suggest that air pollution is at worst a minor factor. If air pollution had been causing a substantial fraction of all asthma cases in the past, then the large declines in air pollution over the last few decades should have resulted in a large reduction in asthma prevalence, rather than the large increase that actually occurred.

A potential objection to this line of reasoning is that air pollution could be a major cause of asthma, but that some other potent asthma-causing factor(s) increased at the same time air pollution declined, masking the asthma benefits from air pollution reductions. But this hypothesis doesn't wash either. Compared to the 1980s, air pollution was far higher during the 1960s and 70s and was far higher still in previous decades. Yet these "industrial doses" of air pollution in the past were not accompanied by high rates of asthma.[10] Furthermore, many developing countries have much higher air pollution than is ever found in the U.S., yet these countries have much lower rates of asthma than the U.S. or other western countries.[11]

Another way to assess the relationship between air pollution and asthma is to follow a group of children over time in areas with different pollution levels and see how many children develop asthma. This was the premise of the Children's Health Study (CHS) in California. Funded by the California Air Resources Board (CARB), the study followed a few thousand children in 12 different California communities throughout the 1990s. The 12 communities ranged from areas along California's Central Coast that have hardly any air pollution, to areas in southern California with the highest pollution levels in the country.

Researchers from the University of Southern California (USC), who ran the CHS, published their asthma results in 2002 in the journal Lancet.[12] The results are surprising. Children living in the six communities with the highest ozone levels were 30 percent less likely to develop asthma than children living in the six lowest-ozone communities.[13] Children living in communities with the highest NO2, PM10, PM2.5, and acid aerosol levels were 20 percent less likely to develop asthma.[14]

The only suggestion that air pollution could cause asthma came when the researchers looked only at those children who played three or more team sports (8 percent of children in the study). In the four highest ozone communities, these children were more than 3 times as likely to develop asthma as similar children in the other eight communities. The sample size was small, so this result should be viewed with caution. Across the 12 communities, 273 children played three or more team sports, and 29 developed asthma.

Yet even if this relationship between ozone and asthma was causal, it has no general implications for Americans. The study was based on ozone levels during 1994-97 in the highest-ozone areas of southern California. No other area of the U.S. has ever had ozone levels as high as these California communities did during 1994-97. Even the study areas themselves now have much lower ozone levels.

Also keep in mind that the apparent increase in asthma with higher ozone applied only to the children who played three or more team sports. Asthma declined with higher ozone exposure for all other children. Thus, if we believe the CHS was uncovering causal relationships between air pollution and asthma, then the take-home message from the study is that higher air pollution reduces the risk of developing asthma.

On Wednesday, the Los Angeles Times reported a new wrinkle in the asthma-air pollution saga, once again based on results from the Children's Health Study.[15] The USC researchers announced that children are more likely to develop asthma if they live closer to freeways.[16] The researchers randomly selected 208 children from the CHS and placed NO2 monitors outside their homes for a few weeks during summer and winter 2000, and also collected data on residential distance from freeways for each child.

Living three-quarters of a mile closer to a freeway was associated with a factor of 1.9 increase in the risk of having asthma, while a 5.7 ppb increase in ambient NO2, 17 percent of the average NO2 level, was associated with a factor of 1.8 increase in asthma.[17] The researchers noted that, based on other studies, the evidence for NO2 itself as a cause of asthma is weak, but suggested that NO2 is acting as a marker for traffic-related pollution in general, especially particulates from diesel exhaust.

So does living closer to a freeway increase your risk of developing asthma, and, if so, is some component of air pollution the actual cause? This latest CHS asthma study suggests air pollution could be having a huge effect. Ironically, that's what makes the study's results so implausible. If a 17 percent change in air pollution levels can cause nearly a factor of two change in the risk of developing asthma, we should have had much higher asthma prevalence back in the 1970s and early 1980s, when most types of air pollution were two or three times higher than current levels. This suggests that asthma cases that the CHS researchers attribute to traffic-related air pollution are actually being caused by something else-either some unmeasured pollutant, or some other factor that is not properly or at all accounted for in their statistical analysis.

Could it be that an unmeasured pollutant in diesel exhaust has increased over the last couple of decades? This is unlikely. While we don't have information on trends in ambient levels of diesel exhaust, diesel truck particulate emissions were measured several times between 1975 and 1999 in Pennsylvania tunnels. Emissions per mile declined about 82 percent over the 24-year period.[18] Researchers from UC Berkeley measured particulate emissions from diesel trucks in the Caldecott Tunnel near San Francisco in 1997 and 2004 and noted a 50 percent decline in emissions.[19] According to the California Department of Transportation, total miles traveled by heavy-duty diesel trucks increased 50 percent in California between 1986 and 2001, a rate of increase far too slow to offset the benefits of cleaner trucks.[20] The data in figures 1 and 2 also suggest that diesel exhaust emissions have declined.[21] If diesel particulates were a major cause of asthma, we would have expected much higher asthma rates in the past-just the opposite of what we actually find.

The Berkeley researchers also measured automobile (car, SUV, and pickup truck) emissions in the Caldecott Tunnel several times between 1994 and 2001.[22] They found that per gallon of fuel burned, emissions of volatile organic compounds declined 67 percent, while NOx declined 49 percent. They also measured benzene specifically, which declined 80 percent. Gasoline consumption in California increased only 13 percent over the same period. Thus, even with increases in driving and the popularity of SUVs, total emissions from automobiles have sharply declined. The data in the graphs above also suggest that emissions from gasoline vehicles have been going down.

It is possible that other substances might explain the association of asthma with living near a freeway. Tire or brake dust might be potential culprits. Or perhaps there are asthma-related socioeconomic factors that aren't captured in the statistical analyses, creating the appearance of an asthma-freeway traffic association where no causal relationship actually exists. The study's small sample size might also have caused spurious results. There were only 208 children in the study, and 31 of them had asthma.

If air pollution is causing asthma, it doesn't appear that any of the wide range of substances we measure or regulate could be the culprit. The asthma-air pollution link has medical plausibility, but there doesn't seem to be much support for it in the actual data. The most likely reason for its persistence is its political appeal for environmentalists and regulators. Environmentalists and regulators depend on public fear and outrage over air pollution for their funding, jobs, and power. Claiming that air pollution causes asthma serves their needs by linking air pollution to a serious disease suffered by millions of children.

By pursuing their parochial interests under the guise of public health, environmentalists and regulators are diverting attention and resources from the search for the real causes of and cures for this debilitating disease.

Joel Schwartz is a visiting fellow at the American Enterprise Institute.

[1] L. Akinbami, US Childhood Asthma Prevalence Estimates: The Impact of the 1997 National Health Interview Survey Redesign (Atlanta: Centers for Disease Control, April 17, 2005),; D. M. Mannino, D. M. Noma, L. J. Akinbami et al., "Surveillance for Asthma --- United States, 1980-1999," Morbidity and Mortality Weekly Report 51 (SS01) (2002): 1-13.

[2] Akinbami, US Childhood Asthma Prevalence Estimates: The Impact of the 1997 National Health Interview Survey Redesign. The Centers for Disease Control redesigned its health interview survey in 1997, so data from 1997 onward are not comparable with data up to 1996. However, although the measure of asthma prevalence changed in 1997, data for 1997 to 2003 suggest asthma prevalence has leveled off.

[3] S. C. Redd, "Asthma in the United States: Burden and Current Theories," Environmental Health Perspectives 110 Suppl 4 (2002): 557-60.

[4] Ibid.

[5] R. A. Etzel, "How Environmental Exposures Influence the Development and Exacerbation of Asthma," Pediatrics 112 (2003): 233-9; Redd, "Asthma in the United States: Burden and Current Theories."

[6] For example, California has collected PM2.5 data since the late 1980s at several locations around California. Data from continuously operated sites show that PM2.5 declined an average of about 40 percent from 1989 to 2003. California's sulfur dioxide levels declined 55 percent from 1984-2003 (based on 16 sites with continuous data, and using the average of the top 30 24-hour readings each year at each monitoring site (same statistic as for the gases in Figure 1). EPA also measures sulfate, a component of PM that comes from sulfur dioxide emissions. This is mainly an issue in the eastern half of the United States, where coal is a major fuel for generating electricity. Sulfate data collected at dozens of sites in the CASTNET network show that average annual sulfate levels declined about 33 percent between 1989 and 2003 (data downloaded from Lead was essentially eliminated as an air pollutant nationwide when it was removed from gasoline during the 1980s.

[7] Sources: Asthma prevalence trend is from California Department of Health Services, Points of Interest -- Asthma in California (Sacramento: May 9, 2003), Air pollution trends are from California Air Resources Board, 2002 Air Quality Data CD, except for PM10 data for 2002, which were provided by CARB staff.

[8] Data for all of these pollutants was retrieved from California Air Resources Board, 2005 Air Quality Data CD, available at CARB also measures formaldehyde, but reliable data are available only since 1996. Formaldehyde was roughly constant from 1996 to 2003. Formaldehyde levels might have increased during 1994-96 when reformulated fuel containing MTBE was introduced. However, according to CARB, prior to 1996 formaldehyde measurements were made with an unreliable analytical method (see California Air Resources Board, 2005 Air Quality Almanac,, Chapter 5).

[9] Data source: California Air Resources Board, 2005 Air Quality Data CD.

[10] For data on pre-Clean Air Act U.S. air pollution levels and trends, see Chapters 1 and 2 of I. M. Goklany, Clearing the Air: The Real Story of the War on Air Pollution (Washington, DC: Cato, 1999).

[11] For international data, see ISAAC Steering Committee, Lancet 351 (1998): 1225-1232.

[12] R. McConnell, K. T. Berhane, F. Gilliland et al., "Asthma in Exercising Children Exposed to Ozone: A Cohort Study," Lancet 359 (2002): 386-91.

[13] This difference was based on peak daily 1-hour ozone levels. Based on peak daily 8-hour ozone levels, children in the high-ozone communities were 20 percent less likely to develop asthma and the difference was on the edge of statistical significance (95% confidence interval for the relative risk of asthma was 0.6-1.0).

[14] This result was also nearly statistically significant, with a confidence interval of 0.6-1.0 for relative risk.

[15] Deborah Schoch, "Study Links Freeways to Asthma Risk," Los Angeles Times, September 21, 2005,,1,3933469.story?coll=la-headlines-california.

[16] J. Gauderman, E. Avol, F. Lurmann et al., "Childhood Asthma and Exposure to Traffic and Nitrogen Dioxide," Epidemiology 16 (2005).

[17] These are the ranges from the 25th to the 75th percentile for NO2 and distance from a freeway. As you might expect, the two measures of exposure to traffic pollution were correlated, with a coefficient of -0.54. That is, greater distance from a freeway corresponded to lower ambient NO2 levels near the home.

[18] A. W. Gertler, M. abu-Allaban, W. Coulombe et al., "Measurements of Mobile Source Particulate Emissions in a Highway Tunnel," International Journal of Vehicle Design 27 (2002): 86-93.

[19] T. Kirchstetter, D. Hooper, Z. Apte, et al. (2004). Characterization of Particle and Gas Phase Pollutant Emissions from Heavy- and Light-Duty Vehicles in a California Roadway Tunnel. 2004 American Geophysical Union Fall Meeting, San Francisco.

[20] Caltrans, Truck Kilometers of Travel, California State Highway System, 1986-2001 (Sacramento, 2003), This is of course a statewide average, but it provides a rough estimate of the increase in truck miles of travel over time.

[21] Diesel soot is a component of PM2.5 and PM10. The large decline in PM during the last 20 years is suggestive of a decline in diesel PM emissions. Declines in benzo(a)pyrene and other organic particulate species also suggest diesel soot emissions have been going down.

[22] A. J. Kean, R. F. Sawyer, R. A. Harley and G. R. Kendall, Trends in Exhaust Emissions from In-Use California Light-Duty Vehicles, 1994-2001 (Warrendale, PA: Society of Automotive Engineers, 2002),



TCS Daily Archives