How to avert Big Food bias in science? Researchers strategize to prevent conflict of interest

To ensure scientific integrity, research findings must be considered ‘credible’ by any number of parties, including peers, practitioners, policymakers, and the public.

Yet when industry sponsors nutritional research, findings can result in bias. This raises ‘significant’ professional, public, and media concern, according to researchers in the UK and Australia.

As Dr Katherine Cullerton from The University of Queensland’ School of Public Health Scientists explained, researchers have long been divided on the best way to manage industry involvement in diet and health research.

“While some scientists feel that the food industry should never be involved, some take the view that achieving healthier diets among whole populations will require actions by the food industry – and for these actions to be effective it requires new research and access to food industry data.”​

Further, where government funding for research is limited, more academics turn to industry for financial support. “This is a particular challenge in low income countries which represent important, emerging markets for food companies,” ​Dr Cullerton told FoodNavigator.

And finally, some industry players want to influence food and public health policy in their favour, she continued. To do so, they may choose to fund certain areas of research for many reasons, one of which could be to ‘deflect attention’ away from food products that are associated with poor health comes.

Without an international consensus on how to manage or prevent conflicts of interests, all of these factors make it challenging for health scientists to know what is ‘appropriate engagement’, and what is not.

So Dr Cullerton, together with researchers from The University of Cambridge, set out to determine agreed principles to guide interactions between population health researchers and the food industry.

Consensus with some confusion​

The study sought to build consensus on this issue by seeking the views of population health researchers and research stakeholders – such as funders, policy offers and journals – internationally, Dr Cullerton explained.

The team conducted a two-stage online Delphi study for 100 researchers across 28 countries. They also ran an online survey for 86 stakeholders in 26 countries.

Participants offered their ‘levels of agreement’ for 56 principles derived from a systematic review, before their comments were analysed using qualitative content analysis.

“Our study showed there was consensus on many of the principles designed to prevent or manage conflicts of interest, however researchers were divided on which companies it is acceptable to receive funding from and the types of interactions with these companies,”​ explained the lead study author.

Indeed, high levels of agreement on principles were achieved for both groups (researchers 68% and stakeholders 65%), with the highest levels of agreement concerning research methods and governance.

Principles that required ‘value-based’ decision making, however, were more contentious. Examples included determining which elements of the commercial sector are acceptable to interact with.

What can be learnt?​

The results provide the basis for developing internationally-agreed guidelines to help population health researchers govern interactions with the food industry, noted the Cambridge-Queensland team.

“There is general agreement between researchers and stakeholders on principles ​concerning research methods and governance.  ​However, the research also highlights that greater understanding of the risks associated with accepting food industry funding or simply interacting with food companies was necessary,”​ Dr Cullerton told this publication.

Some researchers are unaware that they are susceptible to conflicts of interest or that they might be at risk of unconscious biases adversely affecting their science, she continued.

“Ultimately, all of these factors can represent significant reputational risks for researchers.​

“In the final stage of this research, we will be developing internationally-agreed guidance and a toolkit to help researchers better manage the risks resulting from interacting with the food industry.” ​

Source:
Plos One​
'Building consensus on interactions between population health researchers and the food industry: Two-stage, online, international Delphi study and stakeholder survey'
Published online: 22 August 2019
DOI: https://doi.org/10.1371/journal.pone.0221250​ 
Authors: Katherine Cullerton, Jean Adams, Oliver Francis, Nita Forouhi, Martin White