But even if the authors had found what they claimed, it would be a very weak study. Comparing one city to another may seem a useful idea, but it can cause serious comparison problems: You have to adjust for every health-related difference between them.
Critics note that Dockery and Pope clearly didn't make adjustments. They didn't control for humidity, they didn't control for temperature, they didn't even consider income differences.
Such factors could throw off the whole study: Steubenville is considerably poorer than Portage. "Poor persons tend to die more quickly during extreme weather conditions than wealthier ones," says Roger McClellan, a former CASAC chairman and president of the Chemical Industry Institute of Toxicology in Research Triangle Park, North Carolina. As part of the National Health Survey, a 1996 study found that persons living below the poverty line, for a variety of reasons including unhealthier lifestyles and less screening, are far more likely to become sick than wealthier persons. Persons covered by Medicaid were four to five times more likely to have emphysema, chronic bronchitis, and congestive heart failure than those whose incomes were above the Medicaid qualification level. Fred Lipfert of the Brookhaven National Laboratory in Upton, New York, has also found a correlation between a sedentary lifestyle and premature death in five of the six cities.
"It's anybody's guess as to why residents of Steubenville have a higher risk of dying than Portage," says Moolgavkar. It may be some form of air pollution, he says. "A lot of indices of pollution were higher in Steubenville. Still, I don't think that's enough to say pollution caused the mortality, much less to say it was fine particles."
Pollution by Proxy
Philadelphia, the City of Brotherly Love, has also proven a lovely place for particulate studies. Schwartz and Dockery found that as particulates went up there, so did deaths. Three years later, in 1995, Neil Roth and his colleague Yuanzhanh Li did a follow-up study and found no connection between particulate increases and deaths. Neither did Moolgavkar and associates, who also concluded in 1995 that "no specific pollutant can be singled out as being responsible for the association between air pollution and mortality." Instead, they wrote in the journal Epidemiology, "The particulate component of air pollution appears to have become the villain because it is a ubiquitous component of air pollution and thus serves as a proxy measure of pollution." (Jonathan Samet, in an accompanying editorial, agreed this could well be the case.)
Yet Browner speaks of the "consistency and coherence" of the studies on particulates. If there's any consistency at all, it appears to be that the Particle Hunter Triumvirate consistently finds that particles are unhealthy, while other researchers consistently fail to find these same effects when they look at the same cities.
When Moolgavkar refers to particulates as a proxy, he is making a point made regularly by skeptics of the particulate paradigm. When particulate levels rise or fall, the levels of other types of pollution tend to rise and fall, too. This is because these changes are generally weather-related. If a city suffers an inversion and stagnant air, particles will go up, but so will lots of other things, such as pollen. Some of those things can be and are independently measured. Others aren't even looked for or can't be looked for because of a lack of monitors. The result is that particles may just be a proxy, or surrogate, for something else in the air. "If one person who dropped dead had consumed sugar laced with strychnine," says Moolgavkar, "and another who also dropped dead had consumed sugar laced with potassium cyanide, would we blame the sugar?"
Many skeptics believe that particulate research is self-fulfilling. In a paper in The Journal of the Air & Waste Management Association, Brookhaven's Fred Lipfert and Ronald Wyzga, senior manager of air quality risk and health assessment at the industry-funded Electric Power Research Institute in Palo Alto, found that focusing on sulfur dioxide or nitrogen dioxide instead of particulates resulted in the same findings for daily deaths. A focus on carbon monoxide showed somewhat larger effects. Are all these pollutants causing health problems? Maybe, but probably not. A possible answer is that all these substances are proxies for weather-related effects.
While air pollution epidemiological studies often control for aspects of weather, air movement is seldom taken into account. Yet this is "the main cause in day-to-day changes in ambient concentrations of pollution," notes Fred Rueter, vice president at the Consad public policy research group in Pittsburgh and an adjunct professor at Carnegie Mellon. Utility and plant exhaust will be pretty much the same for each day; what will change is what happens to those emissions. Will they blow away or build up? Whatever they do, they'll do it together.
Thus levels of particulates (and emissions) are markers for stagnant air. And "when particulates rise," notes Rueter, "so does almost everything else, and that includes indoors as well as outdoors." Indeed, it's long been accepted that pollution is normally worse inside most homes than directly outside. As most of us spend about 85 percent to 90 percent of our time indoors, indoor air is actually more important than outdoor air. Moreover, when air becomes stagnant, concentrations in indoor air become more important because wind-aided ventilation is less efficient. That allows indoor pollutants--cooking by-products, cleaning agents, mold spores, pet dander, tobacco smoke, hair spray, insect droppings, etc.--to increase.
"What I think is the most coherent explanation for all the epidemiological evidence associating outdoor particulate levels to illness," says Consad's Rueter, "is that on the days particulates are high, people indoors are being exposed to allergens producing effects ranging from mortality from asthma and other things to simple reductions in lung function, leading to increased complaints and hospitalizations."
Moolgavkar and his associates wrote that while some have "argued that the association of particulates with mortality is remarkably consistent from city to city, in the presence or absence of other pollutants, and under varying conditions of weather," they have "been unable to identify a single study in which other pollution variables have been adequately controlled." That was in 1995. Does he still feel the same way? Yes, he says. "When you look at the studies, you generally see that only one pollutant is observed at a time. They don't look at the complex mixture; they just focus on particulates. Often times they have the data available, but they just don't use it."
Show Me the Numbers
Given both the controversy and the stakes, Schwartz and Dockery's data should be made public, rather than simply disseminated to persons they select. Yet for years they refused to do this--even to the EPA, though the federal government paid for their research. In early 1994, CASAC wrote to Browner asking that she get "crucial data sets linking exposure to particulate matter and health responses." Several groups filed Freedom of Information Act requests with the EPA to get the Six Cities data. The EPA responded that it didn't have the information.
Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time.
nfl jerseys|11.16.10 @ 2:07AM|#
mcyjfg