12

Disclosure—and Its Discontents

WHEN I ACCEPT SPEAKING INVITATIONS, I EXPECT THE HOST to pay travel expenses and an honorarium. If I am invited by a food company, I donate the honorarium to my NYU department or the university’s library. But I do not enjoy talking about any of this. Talking about personal finances is always uncomfortable in our society, but with respect to ties to food companies, disclosure can also have unpleasant consequences: public exposure, ridicule, loss of trust and opportunities, and the like. I can well understand why scientists resist disclosure requirements, especially when they have extensive financial ties to disclose.

I learned about disclosure and its discomforts the hard way. In the mid-1990s, I edited a collection of papers on Mediterranean diets for the American Journal of Clinical Nutrition. The papers appeared as a separate supplement paid for by the International Olive Oil Council (IOOC) through a grant to Oldways Preservation & Exchange Trust, a group devoted to promoting the taste and health benefits of traditional diets and their principal foods and ingredients—olive oil conveniently among them.¹ The IOOC was Oldways’ principal sponsor at the time.

Oldways conferences, typically held in olive-growing countries, brought together food writers, chefs, restaurateurs, and academics to learn about the benefits of Mediterranean diets and, of course, the benefits of olive oil. The food professionals I met at these conferences wanted to know more about the history and role of food in society as well as the science behind diet and health. I remain grateful to Oldways for introducing me to the greater food community, not least because these conversations directly inspired the development of food studies programs at New York University.

With Walter Willett of the Harvard School of Public Health, I cochaired Oldways’s first international conference on diets of the Mediterranean in 1993. Many speakers raised provocative questions about the definition of these diets, their relevance for people living outside the region, the health effects of olive oil and other components, and the effects of following Mediterranean diet patterns on food production systems. I thought the conference talks were exceptionally interesting and deserved publication, and I volunteered to take on editing responsibilities. At the time, it did not occur to me that sponsorship by the IOOC and Oldways would influence the publication’s credibility, or mine. The authors were not paid and neither was I. But the journal’s editor, Norman Kretchmer, set me straight. The IOOC had an economic interest in promoting sales of olive oil. Oldways was a partner in that goal. This supplement had commercial implications.

Journal supplements generate revenue for their publishers (the one I worked on cost about $20,000). Because sponsors pay for them, they can be considered advertising, which is why supplement page numbers are identified by the letter s. In part to protect me, Kretchmer insisted on a rigorous external peer review, refused to publish one paper he thought lacking in academic quality, and demanded full disclosure on the title page: “Sponsor: Oldways Preservation & Exchange Trust through a grant from The International Olive Oil Council, Administrator of the United Nations Olive and Olive Oil Agreement.” Authors did not disclose competing interests; this journal did not require such statements until 2002.

As published, the papers presented academically rigorous arguments for the benefits of Mediterranean diets, but they also raised critical questions—evidently too critical. That 1995 journal supplement ended my relationship with Oldways; I was not invited to another of its conferences for the next twenty years. Ironically, on the basis of subsequent, universally favorable research, the United Nations Educational, Scientific and Cultural Organization (UNESCO) placed the Mediterranean diet on its “Representative List of the Intangible Cultural Heritage of Humanity” in 2013, and the 2015 US dietary guidelines promote this diet as one of three exemplary eating patterns. But lesson learned: when work is funded by industry or an industry-funded group, it needs to promote industry interests unambiguously if the relationship is to continue.

Disclosure Policies

Journals have responded to the all-too-human reluctance to expose financial ties by making disclosure requirements increasingly explicit. Today, most scientific and health-professions journals, including nutrition journals, require authors to say who paid for their research and to declare any financial relationships they might have with relevant funders. Table 12.1 summarizes the ways food companies contribute to both categories. They can pay for all or part of the research expenses or can merely donate products to be tested in a study (vitamin supplements, for example). Financial ties to investigators include a longer list of possibilities. Full funding is likely to be more influential than partial funding, but even small contributions can exert influence, as we have seen.

By now, most professional journals follow the disclosure guidelines of the International Committee of Medical Journal Editors (ICMJE). This group provides a model disclosure form, which asks authors to say whether they or their institutions received payment or services from a third party at any time for any aspect of the submitted work, specified as “including but not limited to grants, data monitoring board, study design, manuscript preparation, statistical analysis, etc.” With respect to competing interests, the ICMJE questions cover the preceding three years and ask about patents or other “relationships or activities that readers could perceive to have influenced, or that give the appearance of potentially influencing,” what appears in the submitted work.²

TABLE 12.1. Funding by Food, Beverage, or Supplement Companies: A Taxonomy

Research support

Complete support

Partial support

Donation of products or materials

Participation in some or all of the research process

Researchers’ financial ties

Salary

Stock ownership

Patent ownership

Consulting

Service on advisory committee or board

Service as expert witness or advocate

Honoraria for speaking, writing, or participation

Travel funds, lodging, meals

Journals have considerable leeway in applying these guidelines. I have written occasional commentaries for JAMA Internal Medicine, which, like other journals in the JAMA network, demands especially diligent adherence to disclosure policies.³ In 2016, an editor asked that I verify the accuracy of my disclosure; its policy requires authors to list all “potential conflicts of interest, including relevant financial interests, activities, relationships, and affiliations” as well as “all financial and material support for the research and the work.” The JAMA policy is specific about potential conflicts: “employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript.”⁴

The American Society for Nutrition (ASN) has uniform disclosure policies for authors and editors of all its publications. Its guidelines apply not only to authors but also to their close relatives—spouses, children, siblings, in-laws—and are also highly specific about the ways in which potential conflicts might occur:

Serving as an officer, director, member, owner, trustee, or employee of an organization… or as an expert witness, advisor, consultant, or public advocate (with or without compensation) on behalf of an organization with a financial interest in the outcome;

Receiving support, including grants, contracts or subcontracts, fellowships, consulting agreements, or gifts (e.g., chemicals, experimental diets, trips)… during the time the research was conducted, or over the past 3 years;

Being employed; having rights to patent applications, patents, sales, licensing, or royalty agreements; serving on an advisory board or speakers’ panel; or owning shares in a company or organization that may gain or lose financially.⁵

Even though ASN has strict disclosure guidelines, its journals publish many studies sponsored by food companies. How is this possible? The guidelines make it clear that editors do not see it as their job to prevent or manage authors’ conflicts of interest: “The disclosure of a potential conflict of interest does not necessarily exclude an article from consideration for publication; the goal of disclosure is transparency.” Transparency is good. Management—taking action to minimize the hazards of the disclosed financial ties—would be better but would require editors (some with financial ties of their own) to deal with awkward personal and professional issues.

ASN applies its guidelines to peer reviewers as well as editors. Its requirements for supplement guest editors are even more stringent: “Supplement coordinators must disclose all compensation from the sponsor for editorial services on manuscripts published in the supplement publication and/or for attending, speaking at, or organizing a meeting or symposium, including reimbursements for travel expenses.” Had these guidelines been in place when I edited the supplement on the Mediterranean diet, I and all the other authors would have had to disclose our participation in the Oldways conferences and our paid travel and meals. Our nonfinancial level of commitment to Mediterranean diet practices should have been evident from our articles.

Avoiding embarrassing disclosures is one reason for not getting involved in financial relationships with food companies in the first place.⁶ My current conflict-of-interest declaration says, “Professor Nestle’s income from New York University’s retirement plan and a small stipend support her research, manuscript preparation, and website at FoodPolitics.com. She also earns royalties from books and honoraria from lectures to university and non-profit groups about matters relevant to this publication.”

I had to go into much greater detail on the federal disclosure form I filled out as an external reviewer of an early draft of the 2015 dietary guidelines. Completing this form was a tortuous exercise involving hours of looking up federal tax returns and reviewing my annual NYU faculty activities reports to try to recall all my consultant fees, lecture honoraria, payments for articles, book royalties, and travel and hotel reimbursements during the required three-year period. The form also demanded a statement of intellectual interests. For this, I explained that dietary guidelines are one of my major research and professional interests, that I had been a member of the 1995 Dietary Guidelines Advisory Committee, and that I wrote about the history, development, and politics of dietary guidelines in the 1988 Surgeon General’s Report on Nutrition and Health, in articles published in the 1990s, and in my books Food Politics and What to Eat. I also disclosed that I had been writing about the 2015 dietary guidelines process on my blog and was frequently interviewed and quoted by reporters about that process. My views were no secret. I was appointed anyway, presumably because the agencies wanted my input.

Disclosure Discontents

With respect to disclosure demands, I try to do the best I can with what feels even to me as intrusively personal. I am guessing that most other nutrition professionals also try to do the best they can. But much evidence exists that some investigators are so uncomfortable or offended by disclosure requirements that they forget to comply or do so reluctantly, with overkill, or with sarcasm.

Not everyone discloses, and many disclosures are incomplete. The Integrity in Science project run by CSPI documented frequent nondisclosure in leading medical and scientific journals. More recently, Lisa Bero and her colleagues reported that one-third or more of authors in the studies they examined had undisclosed conflicts and that a similar percentage of published reviews omit statements of funding sources. An analysis of Coca-Cola’s transparency initiative found a smaller, but still significant percentage; 17 percent of its funded researchers did not disclose the funding.⁷

To pick just one example: researchers later involved with Coca-Cola’s GEBN coauthored a study questioning the validity of the National Health and Nutrition Examination Survey (NHANES), a major source of data linking sugary drink consumption to weight gain and its consequences. Their conclusion and its implication: NHANES data are physiologically implausible and should be ignored. The authors provided no funding or conflict-of-interest statements, but word must have gotten out. Two days after publication, the journal ran a correction listing the authors’ financial ties to food companies and this statement: “Funding for the study was provided by an unrestricted research grant from The Coca-Cola Company.”⁸

In my year of collecting industry-funded studies, I posted one with the longest disclosure statement I had ever seen—nearly two full pages of printed text. Nine of the study’s sixteen authors reported not only a breathtaking array of financial ties to food companies but also payments that looked irrelevant: honoraria from the USDA, travel support from professional societies, spouses’ employment, and a sister’s grant for a cookbook.⁹ This seemed like overkill. I wondered if the authors were mocking disclosure requirements, and I wrote the editor of the journal to ask whether its policies required such extensive revelations. He replied, “We give our authors considerable freedom and leeway regarding disclosures. This is easier for an online journal as page length is not a limitation. For the manuscript in question, the authors provided this level of disclosure of their own accord and we accepted it.”¹⁰

I wrote about the statement on my blog.¹¹ I also sent it to Yoni Freedhoff, who followed up with a consultation with Chris MacDonald, an ethicist colleague. MacDonald said, “To a non-scientist, it reads like a bit of a joke. Maybe the authors are tweaking the noses of the editors, and of those who think disclosure of financial ties is an important thing.… Simply dumping information on people is a poor way to satisfy the relevant obligations.”¹² Our posts led two of the authors, John Sievenpiper and David Jenkins, to initiate a conference call with us. They explained that this is what the ICMJE requires and that they did not want to be accused of holding back. I suppose scientists with extensive industry ties lack clarity about where to draw the line, but I thought such ties could have been summarized in a sentence or two.

A year later, Sievenpiper participated in a debate published in the Canadian Journal of Diabetes, taking the position that sugar does not cause chronic disease. His disclosure statement described support from Canadian foundations and professional associations but did not mention his financial ties to the sugar, corn sweetener, and soda companies itemized in that earlier disclosure. The journal published a correction beginning with “The publisher regrets…”¹³ Not following guidelines for disclosure sometimes leads to more serious consequences. My failure to insist on a coauthor’s full disclosure of potential conflicts of interest was the offense most responsible for our having to retract our article, an experience I would rather have avoided.¹⁴

The late David Sackett, considered the “father of evidence-based medicine,” suggested another way to take care of requirements he must have found objectionable. In a post on the BMJ’s website, he wrote, “David Sackett has been wined, dined, supported, transported, and paid to speak by countless pharmaceutical firms for over 40 years, beginning with two research fellowships and interest-free loans that allowed him to stay to finish medical school.”¹⁵

The purpose of disclosure is to alert reviewers or readers to the need to consider potential biases. If authors do not disclose industry affiliations and those affiliations are revealed later, it can look as if the authors are hiding something. The appeal of disclosure requirements is that they recognize conflicts of interest as a problem and suggest a mechanism for dealing with them—avoidance. But like much else about conflicted interests, disclosure issues are complicated. Disclosure can have unintended consequences. Among physicians, it sometimes produces perverse effects. Disclosure has been shown to encourage physicians to provide even more biased advice, to reinforce their feelings of immunity from industry influence, and to justify being influenced on the grounds that people have been warned. Patients sometimes have more—not less—trust in advice that discloses industry funding.¹⁶

But the most serious pitfall of disclosure is that it gives the impression of taking care of the bias problem and eliminating the need to do anything about it. Most American colleges and universities require faculty to file annual conflict-of-interest statements disclosing financial ties to corporate entities (mine certainly does), and some institutions take this requirement seriously. Cornell University’s policy, for example, is noteworthy for its explicit penalties for noncompliance. Researchers who do not submit disclosure forms by a specific date may no longer act as principal investigators, accept grants, receive salary from grants, get salary raises, or, if untenured, continue as a faculty member.¹⁷

One reason why academic institutions, professional journals, and government agencies have adopted disclosure requirements is that they are relatively easy to implement, are only minimally disruptive, and make individuals—not their institutions—responsible for preventing, managing, or eliminating conflicted work. This leads some writers to argue that there is only one effective method for dealing with industry-induced conflicts: ban the funding. Banning, however, ignores the financial benefits of partnerships and the realities of the current research-funding environment. Attractive as the idea might be in theory, it is unlikely to happen in practice. The best that can be done to deal with conflicted interests is to disclose them but also to manage their potential pitfalls. Disclosure is helpful, says Bero, but it is not a panacea.¹⁸

At the moment, disclosure requirements are widespread, but compliance, enforcement, and management are highly inconsistent. Plenty of scientists still view disclosure demands as akin to McCarthyism. Plenty are distressed by efforts to make disclosure even more transparent through new PubMed policies or FOIA requests. PubMed.gov is a searchable database run by the US National Library of Medicine and NIH. Since 1997 the site has provided free access to the abstracts and, sometimes, the full texts of nearly 27 million scientific papers. In 2016, CSPI organized a letter signed by sixty-two interested scientists (including me) to those agencies asking them to list information about funding and competing interests along with the published abstracts. Five US senators also urged such disclosure. A year later, PubMed quietly announced that it would henceforth add this information when publishers provided it, although it would not add the disclosures to abstracts already released.¹⁹

I wished this decision had come sooner. The Journal of Nutrition has included funding information with abstracts for years, but most nutrition journals do not. To find disclosure statements for most of the articles mentioned in this book, I had to log on to NYU’s online library, search for the journal, and find the article I wanted to see—a tedious and sometimes futile process. If NYU did not subscribe to the journal or if the journal’s listings were incomplete, I had to order the articles through interlibrary loan. PubMed’s decision should make this process much easier—if publishers start providing the information with their abstracts (I saw little sign of their doing so by mid-2018). Nevertheless, this was an important decision; it indicated that NIH considers disclosure fundamental to the scientific enterprise. CSPI and the senators issued congratulatory statements.²⁰

FOIA Discontents

Disclosure remains a contentious topic, but FOIA requests for emails from academic researchers are even more contentious. Many states have their own public-records laws, which are variations of the federal FOIA. These allow any citizen to request copies of documents—including emails—produced by government agencies and employees. Researchers who work for state universities are subject to these laws.²¹

Although Congress passed FOIA to promote transparency as a means of countering government corruption, its success in achieving that goal is hotly debated. Critics complain about the frustrations of the FOIA process: government agencies take years to release documents and redact them excessively.²² Are FOIA or open-records requests a way to shine light on otherwise-hidden food-industry influence? Or are they instead methods for harassing scientists whose work or opinions the requester dislikes? As I see it, the answer depends entirely on one’s point of view about how the requests are framed, their purpose, and whose emails are targeted. I have no personal stake here. I retired from a private university exempt from such requests.

But I am impressed by how much we now know about food-industry methods from emails obtained through open-records requests to public universities. Coca-Cola’s strategies for engaging researchers are only one example. On the other hand, if you are a researcher subject to a FOIA request, you might have quite a different view. You might consider FOIA an attack on science. You might think FOIA legitimate in theory but view it as enabling witch hunts, especially when FOIA requests are used as fishing expeditions into private correspondence.²³ That, contrary to expectation, is the view of the Union of Concerned Scientists (UCS), a group that has long advocated for transparency in science. UCS characterizes FOIA as “freedom to bully.” It complains that “well-heeled special interests across the political spectrum… companies, organizations, and activists [who] may disagree with researchers’ findings” are using these requests to harass scientists by going after “all materials on a topic in a university’s possession, including researchers’ draft papers, emails, and even handwritten notes.” Strategies like these, UCS says, “can curb the ability of researchers to pursue their work, chill their speech, and discourage them from tackling contentious topics.”²⁴

UCS’s principal concern is harassment of researchers by industry. It notes that two-thirds of open-records requests come from corporations, and it particularly objects to the chemical and fossil fuel industries’ use of FOIA to harass environmental and health researchers who report harm from their products. UCS has produced a guide to help scientists respond to FOIA attacks.²⁵ Andrew Rosenberg, who directs the Center for Science and Democracy at UCS, insists that the group is not entirely opposed to FOIA requests. His group views inquiries aimed at specific targets as justifiable—those related to Coca-Cola and the GEBN, for example. Rosenberg says UCS objects only to fishing expeditions aimed at harassment, and he argues that FOIA should not be used as the primary means of exposing scientists’ financial ties. Although UCS calls for clear and consistent policies regarding transparency in science, it makes no suggestions for what those might be.

Arguments about use of FOIA create a dilemma. Policies that block corporations from obtaining scientists’ emails also block requests from public health advocates. Paul Thacker, a journalist who advocates for transparency in corporate research funding, views the UCS report as part of a backlash against transparency initiatives. He sees the benefits of transparency as greatly exceeding its costs. Thacker and NYU journalism professor Charles Seife wrote an article critical of UCS’s position: “So long as scientists receive government money, they are subject to government oversight; so long as their work affects the public, journalists and other watchdogs are simply doing their jobs when they seek out possible misconduct and questionable practices that could threaten the public interest.”²⁶ Their article appeared on the PLoS Biologue blog but was so forcefully challenged that PLoS deleted it on the grounds that the criticisms were “not consistent with at least the spirit and intent of our community guidelines.”²⁷

That article and others provide compelling examples of the value of open-records requests for exposing corporations’ engagement with scientists, and scientists’ cooperation with corporate interests. Thacker points out that it is only through emails obtained via FOIA that we know how pharmaceutical companies edit journal articles and prepare testimony for authors, how the National Football League manipulates research on concussions, how Monsanto secretly recruits scientists to advocate for genetically modified foods, and how Coca-Cola attempts to influence high-ranking government officials to minimize concerns about sugars and the role of diet in obesity.²⁸

Such exposure works both ways, of course. I am a strong supporter of organic production practices (they use fewer toxic pesticides and do better at regenerating soil), but emails obtained by the New York Times show that companies producing organic foods have engaged academic researchers to produce favorable studies, publications, and testimony. Thacker concludes that scientists at public universities have no more right to try to hide their emails than any other public officials, and he points out the hypocrisy of researchers saying they love transparency in science—until it affects them personally.²⁹

How you view use of FOIA is most likely to depend on how you feel about its targets. Greenpeace, for example, filed an open-records request to the University of Washington to obtain funding records for a scientist considered to be “a denier of overfishing.” Although the scientist had obtained millions of dollars in research funding from many seafood-industry groups over the years, he did not always disclose this in publications. His supporters viewed the request as undeserved, and the scientist used a defense straight out of COI Bingo: “Greenpeace can’t attack the science because they don’t do science. Instead they attack the messenger.”³⁰

We have seen that FOIA helps expose industry influence that would not otherwise be known. One last example: a directive in a House appropriations bill to exempt checkoff programs from FOIA requests. The bill says the commodity research and promotion boards overseen by the USDA are not federal agencies and are paid for by producers and industry stakeholders, not taxpayers. On this basis, the appropriations committee wanted the USDA to recognize that checkoff boards are not subject to FOIA.³¹

We can only speculate about which lobbying group might have achieved insertion of that directive, but in 2015, the Associated Press used FOIA to obtain emails from one checkoff, the American Egg Board (AEB). These showed that the AEB was engaged in a coordinated, well-funded campaign to undermine the growth of Hampton Creek, a company making an egg-free mayonnaise substitute called Just Mayo. The AEB thought Just Mayo’s threat to egg sales had reached the point of crisis. It attempted to stop Whole Foods Market from selling Just Mayo, pushed Unilever to oppose Hampton Creek, induced the FDA to go after the product’s labeling, threatened the company’s CEO, paid bloggers to discredit the company, and recruited a crisis-management firm to campaign against the company.³²

The AEB publicly stated that it had nothing to do with these actions, but the emails contradicted its denials. Eventually, the AEB was unable to stop Hampton Creek. The USDA investigated, chided the AEB for exceeding its mandate, and required ethics training for its members. The AEB’s actions did cause some debate over whether checkoffs are government or industry programs. As the Congressional Research Service explained in an analysis of the appropriations-bill provision, the USDA currently considers checkoff boards to be subject to FOIA, but checkoff groups say that FOIA requests divert time and funds away from research and promotion. I told a reporter that the House bill was yet another congressional attempt to protect corporate food interests, no matter how undemocratic the results. FOIA was designed to hold government accountable. Commodity groups are happy to have their checkoff programs run by the USDA when convenient, but they want the programs to be considered nongovernmental and exempt from FOIA. They cannot have it both ways.³³

Like much else discussed in this book, disclosure issues are complicated. They create personal and professional discomfort and raise questions that are not easily answered. The most difficult questions have to do with how to prevent conflicts of interest or to manage them once they are disclosed. The remaining chapters address those questions, beginning with the history of early management attempts.