Particulate Air Pollution: A Deadly Mystery.

A Conversation with Ila Cote

From the Environmental Review Newsletter Volume Two Number Nine September 1995

Introduction:

Many medical studies have found an association between mortality and particulate air pollution, but how airborne particles damage health is not well understood. In one such study, Pope et al. (1995) compared air pollution data from 151 U.S. cities to the health records of 552,138 adults who resided in those areas. This study confirmed the previous findings that death rates are higher in areas with greater particulate air pollution.
Based on the early outcomes of this type of research the American Lung Society sued the EPA to force an overdue reevaluation of federal standards for particulate air pollution. The EPA is now in the process of reevaluating the standard.
We spoke with Dr. Ila Cote who is a coordinator for the EPA's research to support this reevaluation. Dr. Cote is assistant laboratory director of EPA's National Health and Environmental Research Laboratory. Dr. Cote received the Ph.D. in medical sciences in 1979 and was subsequently at NYU in the department of environmental medicine. She has worked at EPA for ten years conducting risk assessments and overseeing air pollution research.

ER: Dr. Cote, what are the particles in the air that are apparently making people sick?

IC: It is smaller particles - less than ten micrometers. [A period on the page of the newsletter is about 400 micrometers across. ed.] Larger particles are not readily inhaled; they are just too heavy and they fall out of the air or they are removed by the nose. The national ambient air quality standard is 150 micrograms per cubic meter of air for particles less than ten micrometers in diameter. More specific information about the sizes and chemical composition of particles associated with health effects is very limited.

ER: Where do these particles come from?

IC: Two types of particles are most common: particles from natural fugitive sources, essentially dust, and combustion products such as motor vehicle exhaust or power plant emissions. Additionally these two types of particles can be further subdivided based on chemical composition. It is thought that chemical composition may play a role in causing health effects. But what chemical attributes may be important is unclear. Scientists are asking which components are of greatest concern. The answer to that question would change how you would chose to regulate or lower particle concentrations in the United States. Any kind of regulatory strategy that might be proposed is going to be expensive, so having the answers to what kind of particles you are concerned about is important. The western states are really worried about how dust is considered in the standard. Even now, they go over the current standard every time they have a bad dust storm. And if EPA lowers the standard, then many days out of the year they could be out of compliance. It is difficult to figure out what could be done about that.

ER: How good is the evidence is there that small particles in the air are a health problem?

IC: The epidemiologic - health studies of people in their natural environment - evidence seems to me to be compelling. There have been a number of studies looking originally at mortality, but more recently also at illness, in association with particulate matter concentrations. These studies have been done in a number of cities and by a number of different investigators. They fairly consistently report increased mortality and illness associated with particulate matter concentrations below the current federal standard. There are three kinds of endpoints that are observed in these studies: One type is an increased mortality with short-duration changes in particulate matter exposures - if particulate matter is high on one day, within a few days after that you get an apparent increase in mortality. A second type is mortality associated with long-duration exposures - a city with chronically worse air pollution will have higher death rates than a city with less air pollution. Thirdly, researchers see indices of respiratory illness increase. The two causes of deaths that seem to be elevated are respiratory disease and lung cancer. Several different kinds of respiratory illness have been reported, such as increased asthma attacks, increased pneumonia in the elderly. Many people have continued to look at these epidemiology data to determine if other things could be causing this increase in mortality, and generally, the association has not gone away. I think there is a consensus developing that the increased mortality and morbidity is not an artifact of the analyses; that there is a clear association between adverse health effects and particulate matter or something that particulate matter is acting as a surrogate for. One study I find especially interesting was done in Utah by Arden and Pope. There is major steel plant there that is the major source for the particle matter in the area. Elevated mortality associated with particles was identified in the areas. Then there was a strike and the steel plant was shut down for a year or two, and there was a decrease in mortality. When the strike was over and the plant opened up, mortality rose again. It was a naturally occurring experiment.

ER: Why has the medical community been slow to accept these results?

IC: One reason is that until these recent epidemiologic analyses, people thought particulate matter was a problem we had fixed, and the remaining particle exposures were not of concern. There has not been an intensive effort to further describe the problem since the standard was set back in 1983 or 1984. Another major reason that people have not readily accepted these analyses is there have been considerable animal and other types of human data - clinical and occupational - on particle exposures. The observed effects have not been terribly impressive at lower particle concentrations. Consequently, researchers do not understand, if the epidemiologic data are true, how it works. In other words, they do not understand the possible biologic mechanisms by which particles cause the effects observed in the epidemiologic studies. Hence an important question becomes, why in a more experimental setting do you not see anything, but in epidemiology studies, that have a lot of confounders do you see increased mortality and morbidity? The problem with epidemiology is that you always end up with an association rather than a causal relationship. Perhaps it is not the particulate matter that is causing the health problems, maybe particulate matter changes with some other chemical that is causing the problem. It could be that when air pollution gets bad, particulate matter and something else gets bad, resulting in an apparent but specious association between particles and adverse health effects.

ER: We don't understand how these small particles could cause illness?

IC: It is certainly not well understood. Nothing readily leaps to mind where people say, "Oh, it has to work like this." The other thing that is interesting about these epidemiologic analyses, is that exposure to particles does not seem to show a threshold dose. Everybody is familiar with the concept of thresholds: You can be exposed to a small amount of a toxicant and nothing happens to you but then at some higher concentration you begin to respond. In part that is because the body has many ways to deal with or mitigate the effects of toxic chemicals. So we might expect that there would be some concentration of particles that could be identified that seems relatively safe. But in these analyses there does not seem to be any concentration at which particle exposures have no effect.

ER: What can be done to reduce peoples' exposure to airborne particles?

IC: If in fact the epidemiologic data are true - any particle concentration causes problems, it doesn't matter where you live geographically, and the chemical composition doesn't seem to matter - it has huge ramifications. You begin to wonder, Do dust storms cause mortality or illness? Should you not go to the beach? Should you not have sandboxes? Let alone all the things you would normally consider in a regulatory evaluation; such as, how do you regulate car exhaust and power plant emissions.

ER: What is the relationship between particle concentrations and health effects?

IC: Scientists have looked at a variety of cities with different particle concentrations and the relationship seems to be linear: the greater the particulate pollution, the greater the deaths and illnesses.

ER: So you are trying to sort out the science in order to write a regulation?

IC: Yes. I am one of several people who has designed an am helping to implement the EPA's research strategy to develop additional information. We have also shared the EPA strategy document with other agencies, industry and the academic community to focus attention on what we think are key issues. ER: What additional information are you looking for? IC: We are primarily focused in three areas: We are further evaluating epidemiology studies that have occurred to date. We are trying to understand the mechanisms by which particles could cause observed effects. We are beginning to understand the conditions under which the same effects can be induced in animal studies. This is giving us greater confidence in the epidemiology analyses and provide a tool to study causal effects. Another piece of information we are collecting is improved monitoring data coupled to new epidemiologic studies to improve characterizations of exposures.

ER: What are the implications for a state or city if they are out of compliance with EPA standards?

IC: The way the federal ambient air pollution standards work is that the federal government sets a standard which limits the amount of pollution in the air. The standard is both a concentration and a defined frequency that a geographic area can exceed that concentration. It is up to each state to decide how they control pollution. That's why California has much stricter automotive emission controls than other places. That is one of the ways they have chosen to address their problems. In the end, the federal government can withhold funds from states that do not comply with the standard. There are a variety of ways a new particulate matter standard could be crafted, but right now it is not clear what is the best way to do this. Scientists are hoping by sorting out some of the scientific questions so that EPA will have a better idea what the most appropriate changes to the standard may be.

ER: Do you have any idea why particles, as opposed to other chemicals or ozone, would injure lung tissue?

IC: Many chemicals can injure the lung. Particle effects are likely to additional to effects caused by common pollutants like ozone. There are two hypotheses: One is that there is something about the particle itself, no matter what kind of particle is irritating to the lung. Alternatively, the chemical composition of the particle may cause irritation. If we had to speculate on a mechanism, it may be that particles induce lung inflammation which initiates a disease process. But currently we do not have enough data to know what is going on.

ER: Why is EPA performing this reevaluation of the standard if the data is so incomplete?

IC: EPA was sued by the American Lung Association because of EPA's failure to review the standard in a timely fashion. It says in the law that a review of the standard must occur every five years. EPA had not done that. ER: Was that politics or just overwork? IC: In my opinion it was probably a combination of being overworked and concern that the current state of the data would make an informed decision on the standard difficult. But the judge said the law required a five year review cycle, and EPA was put on a court ordered deadline. The document that reviews all of the health data has been drafted and was externally reviewed by the Clean Air Science Advisory Committee in August 1995. Within the next two years, EPA has to make a decision about the adequacy of the standard. Exactly what that decision is going to be is not yet clear.

ER: Even with tobacco smoke or ozone - which we know are harmful - we do not know how they damage the lungs.

IC: We may not know all the details about these agents but particulate matter research is about ten or fifteen years behind the tobacco smoke and ozone work. All the work that has been done with other air pollutants has certainly helped us understand lung disease better and will help us explore particle related effects.

ER: How does particulate pollution compare to the risk from smoking tobacco?

IC: It is difficult to compare. If you just look at lung cancer and some of the data on carcinogenic potency, tobacco smoke is not very potent relative to some of the other combustion products like automobile exhaust, but when you smoke your exposures are very high. Because of this, the risk to somebody who smokes is clearly going to be higher for cancer than the risk from ambient air. Other considerations however, include that everyone is exposed to ambient air and one does not have much choice in the ambient air they breath. We have almost no handle on the relative potencies to cause non-cancer respiratory diseases for different kinds of particles.

ER: The Dockery paper said that the effect of particulates is small compared to tobacco.

IC: I suspect Dockery was referring to what we were just talking about - that exposure from smoking is much higher so that risks are much higher. The ambient air pollution risks, while smaller, are still potentially large, however. In one of the recent Dockery and Pope papers, they estimated the number of deaths associated with ambient air pollution at 80,000 per year.

Literature Cited:

1) Particulate Air Pollution as a Predictor of Mortality in a Prospective Study of U.S. Adults: American Journal of Respiratory and Critical Care Medicine Vol. 151 669-674 (1995)