Testimony of Joel Schwartz
Department of Environmental Health
Harvard School of Public Health
Department of Medicine Harvard Medical School
February 5, 1997

I have several points I would like to make today:

1. EPA is not out in front of the science on the proposed particle standard, but rather lags behind a number of governments in Western Europe and International Scientific Bodies. The proposed particle standard was approved by its own scientific review panel. Following those nations in reducing its particle standard will avoid tens of thousands of early deaths in the United States.

2. Substantial evidence exists that fine combustion particles of all types are associated with deaths, hospital admissions, and respiratory illness.

3. Despite claims to the contrary, recent toxicological studies show that animals exposed to combustion particles in controlled conditions exhibit the same effects seen in human epidemiology studies.

4. The focus on combustion particles, rather than on dust, is supported by physiology, toxicology, and epidemiology. This strategy avoids costs to control particles which have been shown not to affect public health.

Today, there is a strong scientific consensus that particulate air pollution at levels below the current EPA standard are associated with substantial increases in mortality and morbidity. Last year, the British Government's scientific panel reviewed the epidemiological studies of the association between particulate air pollution and daily deaths and found they were properly done, consistent across many cities, supported by morbidity studies, and concluded that it would be imprudent not to consider those associations causal. They recommended that the British government set a new particle standard at a level that is only one third of the current US standard\1\.

\1\Expert Panel on Air Quality Standards. Particles, Department of the Environment. London: HSMO, 1995; p.30.

The World Health Organization recently convened a panel of international experts to develop a particle criteria document. They also concluded that there is strong evidence that particulate air pollution below current standards is responsible for increased deaths, hospital admissions, and illnesses, and published dose-response relationships for countries to use in standard setting. These World Health Organization relationships predict that the proposed EPA standards will avoid tens of thousands of early deaths per year. A Swiss Government scientific review panel has likewise recommended a new particle standards be set at a level of one third of the current EPA standard. Despite obfuscatory arguments by industry- supported scientists, the Clear Air Scientific Advisory Panel voted to approve a proposed range of 12 to 20 ęg/m\3\ as an annual average standard for fine particles. The EPA proposal of 15 ęg/m\3\ is in the middle of this range.

The reason why so many scientific bodies have reached this conclusion is the vast scope of literature indicating that particulate air pollution has these effects. For example, studies have shown that increases in daily particle levels are followed by increases in daily deaths in Amsterdam, Athens, Barcelona, Basel, Berlin, Birmingham, Boston, Chicago, Cincinnati, Detroit, Dublin, Erfurt, Eastern Tennessee, London, Los Angeles, Lyon, Madison, Milan, Minneapolis, Mexico City, New York, Philadelphia, Provo, Rotterdam, Santiago, Santa Clara, Steubenville, St. Louis, Sao Paolo, Topeka, Valencia, and Zurich. Recent animal studies have corroborated these findings, showing toxic effects of fine particles, especially in sick animals.

Having failed to convince the scientific community and the established scientific bodies designated to review the merits of their case, industry has launched a lobbying offensive to convince political leaders and the general public with the same arguments that failed to sway a more technically sophisticated audience.

One common argument given is that even if particles are causing tens of thousands of early deaths per year, we should not regulate them because we do not know "which particles" to regulate. Absent that knowledge, we might waste money regulating the wrong source. I will comment on the scientific aspects of this issue. Airborne particles are a complex mixture of particles differing by size, chemical composition, and structure. Fine particles, which are the focus of this regulation, are almost entirely generated by combustion, that is, the burning of fuel or other high temperature processes that generate energy, propel automobiles, or produce products. Each combustion source, in itself, generates a complex mixture of particles. Hence control strategies to reduce exposure to fine particles will never focus on a specific particle, they will focus on sources. The question becomes, then, whether the health effects are due to the types of particles generated by one or only a few of the sources, and whether we might waste resources regulating sources that have little health impact. Fortunately, the vast range of locations where airborne particles have been associated with increased deaths and hospital visits allows us to examine the association in locations where each of the major sources is the predominant source of fine particles. Sulfate particles from coal burning power plants predominate in the Northeastern US and Canada, and many studies have shown associations between those particles and daily deaths and hospital visits. But in Santa Clara CA, a winter time study by Fairley showed that particles that were predominantly wood smoke, with almost no sulfates, were also associated with more deaths, and wood smoke predominates in Spokane and Seattle, where particles were associated with increased hospital visits. In Los Angeles, Sao Paolo, and Mexico City, the predominant source is automobile emissions, in London today it is diesel exhaust, in Erfurt Germany and in Dublin it is coal soot, and airborne particles have been associated with increased deaths per day in all these locations. The epidemiologic data indicates that all of the major sources of airborne particles contribute to the excess deaths imposed on the public by particulate air pollution. In toxicological studies, Godleski of Harvard has shown that exposure to either concentrated air particles from the Boston air (primarily sulfates from coal burning powerplants) or to resuspended fly ash from an oil boiler, killed rats with chronic bronchitis. These effects occurred at particle concentrations that were not extraordinary, but comparable to concentrations seen in US cities. Costa's lab at the US EPA has also shown toxicity using either oil fly ash or concentrated particles from air in Washington DC (sulfates from coal) or a German city (traffic and industrial pollution).

EPA has proposed, with CASAC approval, to focus on the fine particles due primarily to combustion, rather than windblown dust, in tightening the particle standard. Industry critics have also challenged this decision. But EPA's focus on true "pollution" and not dust is supported by a study following the Mount St. Helens eruption, which showed very high concentrations (10,000 ęg/m\3\) of dust had little health effect. A Centers for Disease Control study of a dust storm in southeastern Washington State found little impact from an episode where particle concentrations exceeded 1000 ęg/m\3\. In contrast, an episode of combustion derived fine particles at half those concentrations was associated with a substantial increase in daily deaths, hospital admissions, and ambulance calls in West Germany in 1985. The great air pollution episodes of the mid century (London in 1952, Donora Pa in 1948, and the Meuse Valley in Belgium in 1930) were all episodes of combustion related fine particles that occurred in stagnant air conditions which would result in low dust levels. And the increased deaths from all of these episodes are widely agreed to have been causal.

We also know that it is only the fine particles that can penetrate deep into the lung past our primary respiratory defense mechanisms. This is important because the studies of daily deaths and particulate air pollution show a much larger percent increase in pneumonia deaths than of all deaths. Pneumonia is a disease of the lower lung, to which fine particles but not coarse particles, penetrate. The increase in heart disease deaths also seems more plausibly related to particles that penetrate in the breathing region of the lung which is closely connected to the heart.

The animal data also clearly point to the fine particles as much more toxic than the coarse particles. When Dreher and coworkers at EPA placed fine particles and coarse particles collected from the air in Washington, DC in the lungs of animals, they found substantial toxicity from the fine particles, but little from the coarse particles. The same laboratory has shown that fine combustion particles can induce life-threatening heart arrhythmia's in animals with chronic lung disease. Osornio-Vargas and colleagues at the National Institute of Health assessed the toxicity to lung cells of particles sampled from different areas of Mexico City. The particles from the northern part of the city, which were primarily from combustion were much more toxic than the particles from the south, which included much more dust.

Costa and coworkers have shown that the toxicity of airborne particles is related to the concentration of soluble metals on their surface, and that coarse particles have much lower concentrations of soluble metals than fine combustion particles. This may explain the differences in the toxicological data.

As noted above, Godleski has exposed rats to concentrated fine particles from Boston. The exposure averaged less than 100 ęg/m\3\ over a three day period, but peaked at 288 ęg/m\3\. While healthy rats were not affected 37% of the bronchitic rats died following this modest exposure. These results agree with the epidemiology studies which show the greatest increases in deaths occur in people with chronic lung disease. Even lower concentrations of fine particles from Boston air were associated with changes in electrocardiograms in healthy dogs. These electrocardiogram changes are know risk factors for sudden deaths. Again, the epidemiology studies have shown that these deaths are particularly affected by airborne particles.

Epidemiology studies also support the conclusion that the fine combustion particles, and not the dust, are responsible for the observed health effects. In 1994 Thurston and coworkers at New York University reported that coarse particles were not associated with hospital admissions for respiratory disease in Toronto, but that fine particles were. We reported the same results for daily deaths in six US cities last year. Hence the epidemiology is quite consistent with the toxicology, and what we know about the penetration of particles of different sizes into the lung.

Industry has argued that these epidemiologic results all derive from measurement error. These arguments about measurement error have been made consistently by Dr. Lipfert throughout the process of writing the EPA criteria document, and its review by CASAC. They represent nothing new, were available to CASAC when it approved moving to a fine particle standard, and are contradicted by the animal studies. The latest version of this argument suggests that we lose some of the coarse particles of our current monitors, but the amount varies from day to day. This will reduce the correlation of coarse particles with mortality. However, fine particles are also lost by current monitoring techniques. Because the monitors collect particles for 24 hours and then measure them, volatile chemicals on the fine particles often escape off the filters before the 24 hour period is up. Hence both measures are subject to measurement error, and it is not clear which is larger.

Other studies have looked at the effects of long term average exposure to fine versus coarse particles. This is important, because averaging over many measurements averages out the measurement error. For example, Figure 1 shows data from the 24 City Study, which was published last year. The percent of children (aged 8-12) with abnormal lung function in each town (after controlling for age, sex, height, and weight) is plotted against the mean concentration of fine particles in that town. A strong trend is seen, with the percent of children with abnormal lung function increasing threefold as you go from the less polluted to the most polluted communities. Figure 2 shows the same percentages plotted against the concentration of coarse dust particles. No evidence of any association is seen.

In summary, and international scientific consensus has emerged on the adverse effects of combustion related particles. This has resulted in widespread efforts to tighten airborne particle standards throughout the western world. The EPA proposal is not an attempt to push the limits, it follows conclusions by other scientific review bodies and governments.