Public Health Scholar On China’s Research Partnerships With Coca-Cola

Originally published by US-China Today on July 8, 2019. Written by Emma Cockerell and Jason Chwa. 

Critics have long accused food and beverage companies of trying to exonerate their products from blame for obesity by funding organizations that highlight alternative causes. We investigate the supposed model of transparency between Coca-Cola and China’s private and government research institutions and its implications for public health.

In January 2019, an investigative report published by Dr. Susan Greenhalgh unraveled how through the International Life Sciences Institute, Coca-Cola has been quietly influencing the means in which China tackles the growing rates of obesity and brought attention to the manner by which companies can effectively exonerate their products and advance corporate interest in the realms of both profit and health policy. However, conclusions about the intentions of food and beverage companies in funding scientific organisations have been prevented by limited access to industry’s internal documents. Dr. Paulo Serodio’s research focuses on the health effects of corporate influence on health outcomes and public policy and utilizes network science and machine learning tools on innovative text-based datasets. Dr. Serodio has conducted an extensive systematic review on studies that have received funding from Coca-Cola and observed that most research focuses on nutrition and emphasizes the importance of physical activity. Here, US-China Today conducted an email interview with Dr. Serodio to gain more insight into how corporations can influence public policy in countries where such practices undergo less scrutiny and how the scientific and political climate in China allowed for American corporate interests to gain such a significant foothold.

Recently, you commented that you were surprised at the extent to which Coke was able to penetrate the Chinese government. What factors/characteristics makes a country particularly vulnerable to having its public policy influenced by corporations?

The lack of democratic institutions and basic accountability. Without the ability to vote out of office in free and fair elections, [and with] politicians who cater to special interests, then accountability is unachievable.  

In your publication examining Coca-Cola’s research funding, you mapped all scientific research acknowledging funding from Coca-Cola, identified the institutions/authors, and used the company’s disclosure to gauge whether its funded researchers acknowledge the source of funding. In lamest terms, describe the process by which you were able to acquire all of the  data for your study. For example, you mentioned that you utilized the Web of Science database and the conservative eligibility criteria. Were there any barriers/conflicts you faced when collecting data? Knowing that there has been relatively little access to the industry’s internal documents, how were you able to gain access to all of these studies? If any, what were some of the possible flaws/biases in the method by which you retrieved and analyzed the studies?

The exercise of mapping industry funding of scientific research is a very difficult one. To begin with, we still rely on self-reporting. Even though (some) journals have put in place strict guidelines on reporting funding sources and competing interests, we are still relying on the authors themselves to reveal them. And as we pointed out in the paper, they don’t always do it.

Secondly, until recently we had no way of systematically collecting this information: we could conduct thorough literature searches in bibliometric databases by author, institution, journal and even “topic”/”theme”, but not by funding source. This changed when Thomson Reuters’ Web of Science began to parse the funding/competing interest statements in published articles and including them as metadata in its databases. This allowed researchers (with institutional access to Thomson Reuters’ resources) to conduct systematic searches of industry-funded articles.

Collecting the information isn’t difficult. I wrote a script in an open-source programming language that scraped and parsed all the metadata pertaining to articles funded by Coca-Cola from the Web of Science platform. The process is akin to automating what a human would do in collecting and storing this information, but several orders of magnitude faster. The final sample of articles we collected does suffer from some limitations (for example, metadata on funding statements is only available for studies indexed by Web of Science after 2007), but the search was comprehensive enough as not to introduce any bias in the final sample.

I wouldn’t say we faced barriers in the data collection (apart from the obvious one: access to Web of Science, which is behind a paywall), but we certainly did face the following in the analysis stage:

Thomson Reuters has a proprietary algorithm that parses the conflict of interest/funding statements in published articles and automatically identifies the funding source. As with any automated procedures, this leads to some misclassification – for example, it may identify Coca-Cola as the funding agency for the paper, when in fact the COI statement refers to Coca-Cola as having funded a conference that was attended by one of the authors of the paper. We double checked the accuracy of the algorithm by, essentially, doing this classification procedure (identifying articles in which Coca-Cola is referred to as the funding source) ourselves.

In many cases of co-authored articles, the funding statement did not reveal who was the Principal Investigator on the grant awarded by Coca-Cola. This added an extra hurdle in identifying those researchers that had been funded by the company but were missing from their transparency disclosures.

We had strong priors about the network of researchers who had ties to Coca-Cola before conducting the search. These stemmed, on the one hand, from Anahad O’Connors article in the NYT, which unveiled Coca-Cola as the main funding body behind the Global Energy Balance Network and, on the other, from email communications between researchers in public universities and Coca-Cola officials obtained by US Right to Know. A comprehensive analysis of this data revealed a strong, cohesive network of researchers who, unsurprisingly, were the most central players in the data we retrieved from Web of Science.

While the goal of the aforementioned study was to evaluate the extent to which Coca-Cola’s Transparency Lists are truly comprehensive, was there anything in the examined studies that stood out? For example, you mentioned that you observed that the funded research emphasizes the importance of physical activity and the concept of energy balance. Were there any noticeable trends in the studies?  

The trend was certainly a focus on the concept of energy balance, which was key to Coca-Cola’s strategy to position its products in the market without suffering a backlash from the public health community. A strong body of research giving seemingly independent scientific support to the idea of energy balance was of utmost importance because it accomplishes two things: (1) it detracted attention from the growing body of research documenting the negative effects on health of consuming sugary beverages; (2) it provided a way to reverse these negative effects (aka increase energy expenditure via exercise). It was a very effective way of reframing the debate, turning the focus onto personal responsibility arguments, whilst promoting physical activity as the solution to restore the energy balance. We certainly saw these arguments in the literature. It also forced us to take a step back and understand that it wasn’t only about funding research projects, it was a much more intricate, multi-layered PR campaign which involved creating seemingly independent scientific research groups that pushed the energy balance agenda, funding academic conferences worldwide where these topics were given visibility, promoting and funding a variety of important sports-related events, and funding training events for health journalists from the top news outlets in the country where Coca-Cola funded academics would push the energy balance agenda.

Ultimately, these were very resourceful tools that could be used in lobbying efforts to push back on, for example, proposals on a sugar tax. Elements of this multi-layers strategy could be used to convince policymakers that the science isn’t settled, or that promoting physical activity in schools (rather than nutrition education) yields better results in tackling the obesity epidemic, therefore maintaining the status quo that benefited the company.

Coca-Cola has unapologetically revealed its intentions to advance its interests by funding a scientific organization. Particularly, the company has sought to advance its energy balance framework by establishing an ostensibly independent broker with assistance from scientists, such as GEBN. How is this possible? Why are the scientists that are a part of GEBN sympathetic with their goals? Would any other scientific institution have been possible? Are there any other organizations Coca-Cola has specifically targeted in recent news?