How a Trump Administration Proposal Could Worsen Public Health
Since Donald Trump took office, he has looked to weaken environmental regulations, with hazardous consequences: during the past two years, American air quality has worsened, after decades of steady improvement. Now, the Trump Administration has proposed a new measure that would limit the research that the Environmental Protection Agency can use when regulating public health. A draft proposal, obtained this week by the Times, calls for the E.P.A. to base regulations only on studies whose raw data is made public. The problem is that most studies of the health effects of environmental hazards are based on individual medical histories and involve confidentiality agreements. The new measure would also be retroactive, potentially undercutting many existing regulations. (My colleague Carolyn Kormann wrote last year about an earlier version of the proposal.)
To discuss the measure’s potential effects, I recently spoke by phone with Douglas Dockery, a professor of environmental epidemiology at the T. H. Chan School of Public Health, at Harvard. Dockery was the lead author of the landmark Six Cities study, from 1993, which gathered data on thousands of Americans and found a link between life expectancy and air quality. If the draft proposal moves forward, the study, which has been the basis for public-health regulations for more than two decades, could become inadmissible. During my conversation with Dockery, which has been edited for length and clarity, we discussed how the study was conducted, the importance of confidentiality agreements in scientific research, and Dockery’s fears about what the proposed E.P.A. policy could mean for Americans’ health.
What was your reaction to this latest policy change?
There is a long history of this. The earliest recollection, for me, was going to testify in the Senate, shortly after the original study was published, and walking up to Capitol Hill, and coming across these guys in white jackets with signs saying, “Show us the data.” I didn’t know what that was about, but they were from the Citizens for a Sound Economy.
Which was an industry front group of some sort?
Yes, exactly. [The group, funded principally by the Koch brothers, went out of existence in 2004.] So this has been an issue for twenty-five or thirty years. When the E.P.A. proposed the standard, in 1997, there were calls at that time for release of the data.
O.K., let’s take a step back. Tell me about this study and how it came about.
The study was started in the mid-nineteen-seventies to examine the effects of expected changes in air quality in the United States as a result of depending more on domestic fossil fuels—that is, coal—for power production. It was funded by the National Institute of Environmental Health Sciences and the Electric Power Research Institute, and we got some support from the E.P.A.
How did you conduct the study?
There were many aspects to the study. We recruited children in elementary school and were examining their growth and how their lungs grew over time. We also had a random sample of adults in each of the communities, and we asked them about their occupational history, their smoking history, and many other factors, and measured their height and weight and lung function, and then basically just followed them for the next fifteen years or so, until they died. We then looked at how long they lived, and showed that those who lived in dirtier communities were dying earlier than those who lived in cleaner communities, after adjusting for age and sex and smoking and occupation and obesity and all the other factors that we had measured.
Why do you need confidentiality agreements?
We are asking these people for a lot of individual information and personal characteristics. This was all done by face-to-face interviews with them, and going back and seeing them every couple of years, or writing to them. A lot of personal contact was required. Now you do these studies by trolling databases that are available, but this was depending on having the trust of the individuals in the study to provide you the information, knowing that you weren’t going to be using it for other purposes.
So the study comes out in 1993, and then, later, you go to testify. What happened?
When the study came out, there wasn’t any reaction initially. It was only several years later, when the E.P.A. proposed to create a standard for fine particles, after the American Cancer Society study came out [which corroborated the Six Cities study], that the industry started demanding access to the original data, to see the individual measurements for people in the study. We said that we cannot do that based on the confidentiality agreements we had signed with individuals, and the other confidentiality agreements we had signed, actually, with each of the states and the federal government for access to the mortality data.
We were looking for a way to get around that and provide access to that data while still respecting the confidentiality of the participants, and so we turned to the Health Effects Institute, which was an organization funded by the E.P.A. and the automobile industry, and asked them, as a body that had a reputation as being independent and try to bridge government and industry, to organize an independent review. And they agreed to do that.
They had an open call for independent investigators and had a series of public meetings to propose alternative approaches to looking at the data. They hired a group from the University of Ottawa that won that competition, and we turned over all our data to them, and they were able to look at all the original records after signing all the appropriate confidentiality agreements and setting up secure systems. This took a year or more. And they came out with a report that we didn’t have any input into. They said that the data was of good quality and they could replicate what we did. When they tried different approaches and controlling for different factors and things that had been suggested as alternatives, they could not show that there was any influence of those alternative factors. They published that report and we felt pretty vindicated by that.
Did this take any pressure off from industry groups?
Yes, to a large extent. But, on the other hand, there was, in the omnibus funding bill that was passed in 1998, an amendment that became known as the Shelby Amendment, sponsored by Senator Richard Shelby, that said that universities or any nonprofit doing work funded by the federal government would be subject to a Freedom of Information request. This was a direct result of the report coming out and us not releasing the data publicly at that time. It was specifically targeted at universities and hospitals and any nonprofits. It did not apply to any commercial investigations or studies published by industry.
That basically suggested that anyone could go in and request any record from a university. There was a big pushback by the scientific community on that. Previously, universities and nonprofits were not required to provide information based on Freedom of Information requests. The upshot was that, after all this, it was determined that a study that was used in regulatory rule-making would be accessible by a FOIA request. So that is where we stand right now. And, in any studies we do, we cannot tell our participants that we can maintain the confidentiality of their data, because it is potentially subject to a FOIA request if used in regulatory rule-making.
So if the Shelby Amendment is on the books, why did the E.P.A. need to make this change?
There clearly is an assault on epidemiology, making it harder to do studies and making their influence less apparent. The problem is that, if you have a new chemical or a new agent, it is assumed to be safe until proven hazardous, and the only way we have of determining that is in retrospect, by looking at the effect of these contaminants in the population using epidemiology. This is just one more approach to try and prevent the use of epidemiology in evaluating the health effects and potential policy implications of them.
O.K., but what does this change do that the Shelby Amendment didn’t?
The Shelby Amendment said that the data had to be available under FOIA. This rule is saying that the E.P.A. cannot even consider data unless it is in the public domain. The other question is what the implication is for studies not funded here in the U.S. There is lots of really important work being done in E.U. countries, where they have much stricter rules relating to confidentiality data. Part of this would also be to restrict the use of data from sources outside of the United States.
Have you heard any good arguments from industry people in favor of these rules?
We have worked very closely with industry in many aspects of this. The Six Cities study was funded to a significant extent by the Electric Power Research Institute. To the extent that industry is interested in trying to provide the best science and stable, predictable regulations, I think it has been a very positive association. But there are others who have been pushing it in another direction.
So then why is this happening? Is this a case, as with auto emissions, where the Administration is pushing even further than industry wants?
Actually, I am not as concerned as much about air pollution as I am about other things that depend on epidemiology.
So you are saying because there is so much data on air pollution, you are worried less about that than other areas?
I was going to ask about your biggest fear.
It’s that the E.P.A. would not be able to look at epidemiology to identify new hazards. It’s going to give free rein to a whole range of additional hazards in the environment.
It’s a lot of the new chemicals. We are adding thousands of new chemicals into consumer products and into the environment. One of the most recent examples are the PFAS [per- and polyfluoroalkyl substances] that have been introduced, and we are starting to see suggestions in the epidemiology that there are health effects associated with those—things that were thought to be safe, but are having long-term health effects that are being recognized.