Do you consider yourself a type A personality — ambitious and impatient? If so, you can thank the tobacco industry for that bit of self-knowledge. For decades, the cigarette companies Philip Morris International and R. J. Reynolds supported and promoted studies linking a driven personality to an increased risk of heart disease. The apparent motivation? To raise questions over smoking as a contributor. Subsequent research by scientists without funding from the tobacco industry did not link type A personalities to higher rates of disease or death.
Copious case studies document how industry influence can muddy research on the health impacts of soda, tobacco, fossil fuels and more, but researchers are largely unaware of this. It is time for research funders to integrate this information and consider vested interests as a force in the complex research system.
Sometimes, partnerships and licensing are essential. Few drugs, for example, would reach patients without the pharmaceutical industry investing in late-stage trials. And plenty of scientific instruments, computing technologies and household products have come from industry-supported researchers. Still, the scientific community must be wary. In my view, industry influence over health research is waxing, not waning.
In June, Francis Collins, director of the US National Institutes of Health (NIH), cancelled a US$100-million study on the effects of moderate drinking. An advisory committee concluded that interactions between government staff and representatives of the alcoholic-beverage industry seemed to be intended to bias the study to find beneficial health effects, and that it was not designed to detect potential harms such as cancer or heart failure.
The NIH is now working to clarify processes for private-sector collaboration, with a plan expected by the end of the year. Funders everywhere should take this as motivation to become more aware of how commercial bias enters the research system — and to learn how to offset it.
Some measures to counter bias are in place. Authors are now routinely asked to disclose conflicts of interest in research papers. And to prevent companies from ‘burying’ studies that fail to show patient benefit, many countries demand that clinical trials be registered before they start.
These measures are inadequate, and not just because they are enforced imperfectly. Commercial interests can influence research in ways that are largely unmonitored. Industry funding for research on topics such as food, alcohol and tobacco might not directly bias particular studies, but instead encourage a focus on short-term impacts or healthy people, showing less harm. Another tactic is to provide ‘unrestricted’ grants to researchers whose previous work has aligned with the interests of the industry. In these cases, the researchers may well be acting in good faith, but the industry creates an uneven playing field by providing disproportionate funding to a subset of researchers. Other tactics include possible smokescreens and distractions. For example, in 2017, Philip Morris pledged funds to establish a grant-giving foundation to end smoking, raising all sorts of questions. (The foundation says that, by law, it operates independently of its benefactor.)
Individual funders often have policies or frameworks that are intended to curb industry bias, but these vary widely and are confusing. We need a broad discussion about how industry can support research without skewing it.
Researchers need to reflect on industry influences at several levels, from personal contacts to informational sessions or sponsored research. Informal conversations or interactions at conferences might shape academics’ thinking, as might discussions on funding committees, which often include industry members brought in for their perspective. Specifics matter: reformulating or developing a product might require industrial collaboration, but the most peril comes when industries are involved in evaluating the impact on health of that product or making policy recommendations about it.
Government and charity funders who advocate industry partnerships should develop and perform structured risk analyses, as should universities and academic researchers who pursue such partnerships. As already required with political donations, public companies should disclose funds given to academics in regulatory filings and name those receiving the most money.
Scientists should get savvier: some ostensibly pro-science initiatives have industry involvement. Vested interests then target regulators by promoting ostensibly pro-science policies that have the effect of discounting or excluding unfavourable results. Industry also provides funds to individuals who make favourable recommendations to advisory panels, as documented in the past decade for dozens of physicians involved in drug approvals.
There is some movement in the right direction. The charity Cancer Research UK, along with several other bodies that award grants for cancer studies, does not fund scientists who are also supported by the tobacco industry. The UK Prevention Research Partnership, a £50-million ($64-million) funding initiative involving a number of government and charity bodies, has developed guidelines on industry partnerships and has indicated that it will support research into commercial influences on health. (It has also funded a £40,000 consortium-development grant from myself and colleagues, held by the University of Stirling, UK.)
Questioning the role of vested interests in research will not be popular or easy, but it is essential. Governments and academics: step up.
Nature 560, 283 (2018)