For healthcare professionals only

Head To Head

Should we welcome food industry funding of public health research?

BMJ 2016; 353 doi: https://doi.org/10.1136/bmj.i2161 (Published 20 April 2016) Cite this as: BMJ 2016;353:i2161
  1. Paul Aveyard, professor of behavioural medicine1,
  2. Derek Yach, executive director2,
  3. Anna B Gilmore, professor of public health3,
  4. Simon Capewell, professor of public health and policy4
  1. 1Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford OX2 6GG
  2. 2Vitality Institute, New York, USA
  3. 3UK Centre for Tobacco and Alcohol Studies, University of Bath, Bath, UK
  4. 4Department of Public Health and Policy, University of Liverpool, UK
  1. Correspondence to: P Aveyard paul.aveyard{at}phc.ox.ac.uk, A B Gilmore A.gilmore{at}bath.ac.uk

Researchers should accept research grants from the food industry, write Paul Aveyard and Derek Yach, but Anna B Gilmore and Simon Capewell say that it biases science

Yes—Paul Aveyard, Derek Yach

The food industry consists of farmers, manufacturers, wholesalers, retailers, distributors, and the catering industry. If it disappeared tomorrow, most people in the developed world would die within months. It is a major employer; more people are employed in the food industry in the UK, for example, than any other manufacturing sector.1 For these reasons, government policies seek to support the industry. From this perspective, it would be absurd for health policy researchers to shun collaborating with the food industry.

Of course, our aims are not always allied. Many elements of the industry promote and sell food that undermines health. When faced with effective public health actions to curtail consumption of unhealthy products it sometimes fights against the cause of public health. Therefore not all cooperation is appropriate. For example, although industry’s views are critical to developing public policy, its presence when such policy is decided is inappropriate.

But working with the food industry inevitably involves accepting its funding, in kind at least. Take the FLICC study,2 in which a supermarket searched its database to find regular consumers of processed food, often high in saturated fat and salt, and asked them to participate in the trial. The trial is testing an intervention to promote the motivation and capacity of shoppers to make better use of front-of-pack “traffic light” nutrient labels and uses loyalty card data to determine the nutritional profile of purchases. The retailer paid the costs of its staff’s time used for discussion with researchers and to collect and process loyalty card data. Would the study be better if the researchers had paid the supermarket for their time and information? We doubt anyone would seriously argue so, bearing in mind this would remove funding from other worthy research.

Reputational advantage

Why might the supermarket co-fund this study? It is because supermarket managers share the goals and values of the researchers. There is no reason they should prefer to sell less healthy food over healthier food. It may offer reputational advantage to help shoppers to make healthier choices. Only the most cynical would argue that collaborating in this way is pretending to do the right thing and merely deflects radical and unpalatable options, such as regulation.

In some cases, the interests of the food industry align so strongly with those of health researchers that industry is willing to pay the entire costs of the study. Basic research on better ways to reduce salt, sugar, and fat; intervention research on the benefits of micronutrients to health; or research on the effect of discounting healthy food on dietary patterns are examples. Leading manufacturers are investing billions of dollars to improve the nutritional quality of their products—well in excess of public research investment.

Strong safeguards

Independent researchers should be responsible for design, conduct, and analysis of the research and not the company. This implies research organisations accept direct payments from the food industry. Such payments require strong safeguards to avoid bias and the appearance of bias: researchers should have no commercial interest in the product; payments should be made to the organisation not the researchers and should reflect the cost of the research to avoid researchers feeling beholden to the company; the analysis should be done by statisticians independent of the investigators who designed and conducted the study; and researchers should publish the results regardless of the outcome. There are excellent examples of best practice in industry funded food research3 4 and other contested academic areas.5 6 7 The alternatives are that the research is not done, that it is done by the company itself, or that the public pays.

In many cases food industry and health goals clearly align and co-funding in kind or through direct payment from industry is appropriate. Examples of bad practice do not invalidate findings from appropriate collaborations. Providing safeguards are in place, no reasonable person should doubt the integrity of collaborative research.

No—Anna B Gilmore, Simon Capewell

Corporations are legally required to maximise shareholder profits and therefore have to oppose public health policies that could threaten profits. Unequivocal, longstanding evidence shows that, to achieve this, diverse industries with products that can damage health have worked systematically to subvert the scientific process. The research they fund produces uniquely favourable outcomes.8 9 10 11 Internal documents show how they manipulate evidence in their favour, strategically communicate that evidence to influence public and political opinion, and ultimately minimise regulation and legal liability.8 9 12 13

The food industry is diverse, but there is a clear conflict between public health and companies that produce ultraprocessed food and sugary soft drinks. It is unsurprising, therefore, that similar evidence is now emerging for these companies. Studies they fund are generally biased in their favour.14 15 Previously secret documentation shows they are working to ensure research and researchers they fund deflect attention from their products and unwanted regulatory interventions.16 17 They promote weak or ineffective interventions aimed at individuals rather than upstream population level regulation, and emphasise physical activity and energy balance to the exclusion of diet based drivers of obesity, messaging reinforced by non-governmental organisations fronted by industry funded scientists.16

Four flawed excuses

Much like the case with tobacco, scientists taking money from ultraprocessed food companies typically offer four flawed excuses.

“I am an objective, impartial scientist and cannot be biased”

Wrong. Industry funded research is significantly more likely to produce outcomes favourable to that industry (rates vary from about five14 to 888 times). Sometimes researchers are willing to fabricate data or allow corporate lawyers to hide unfavourable studies.8 But even well meaning scientists are (often subconsciously) biased, even by small gifts.18 19

“I control the data and have freedom to publish, so my research cannot possibly advantage industry”

Wrong. Extensive evidence shows that quality, peer reviewed studies where scientists control the data and are free to publish are vital to corporate strategies.8 13 Industry funds these studies, often by the top researchers in the top institutions, because it knows they are safe (that is, the results will not threaten industry interests), will divert attention to alternative risk factors, enhance industry reputation, and create a cadre of indebted experts.8 13 20 These scientists may be doing quality work but it does not advance science. Instead it puts industry in control of the research agenda. By determining what questions are asked and , crucially, not asked, industry can distort the evidence base to defend itself from attack.13

“Industry funds are crucial for research”

Wrong. Industry funding accounts for a surprisingly small proportion of total research funding, less than one tenth in the UK or US.21

“Disclosure and peer review provide sufficient redress”

These are essential and do help, but are insufficient.20 Experimental studies show disclosure can, paradoxically, exacerbate rather than reduce the biasing effect of conflicts of interest.22 Moreover, conflicts of interest may not be identified as relevant, particularly when industry is deliberately funding distracting research, and even the best peer review can fall short.

What can be done?

We must not allow the food industry and its allies to use these smokescreens (such as limited industry collaboration in independently funded studies) to obfuscate and delay action on the key issue in this debate—the funding of obesity related research by corporations whose interests are threatened by effective anti-obesity interventions. Until funding models change, our ability to counter the obesity epidemic will be seriously constrained. We therefore need more radical funding models that allow corporations to fund research while protecting that research from their influence.

Previous work shows that dedicated manufacturer taxes, licence fees, or legally mandated contributions (with the funds raised then administered independently from industry) are most likely to maximise transparency and minimise conflicts of interest.23 Funding of research through a tax on tobacco and alcohol in California and Thailand shows that this approach is feasible.

Change will not occur until public health researchers refuse to take money from the ultraprocessed food industry. It worked for tobacco; in the early 1990s all bar one UK medical school took tobacco industry funding. That is unthinkable today, and the change has underpinned progress in tobacco control.

Footnotes

  • Competing interests: All authors have read and understood BMJ policy on declaration of interests and declare the following interests: DY has been employed full time by PepsiCo and Discovery and has had paid advisory roles with PepsiCo, Tesco, and Mars on promoting healthy and sustainable diets; he has also had advisory roles with the Institute of Medicine, Wellcome Trust, and World Economic Forum. PA currently holds research grants paid to his university from the MRC, NIHR, NICE, and Cambridge Weight Plan. All these grants have been or are held under the safeguards outlined in the article. He has worked as an unpaid member of guidelines groups for NICE and is a committee member of the Association for Study of Obesity and the UK Society for Behavioural Medicine. ABG currently hold grants from NIHR, Cancer Research UK, and the New Venture Fund and is a member (unpaid) of the UK Health Forum and the Council of Action on Smoking and Health. SC is a trustee (unpaid) of the UK Health Forum, Heart of Mersey, and UK Faculty of Public Health and is a member of ASH Scotland.

  • Provenance and peer review: Commissioned; externally peer reviewed.

References

View Abstract