Not Quite Transparent

Superficial Transparency in Nutrition Research

The food industry wants to sell you food. And the industry frequently uses nutrition research to do it. For many good reasons, scientific journals require disclosures of conflict of interest. But in a new JAMA viewpoint, John Ioannidis and John Trepanowski submit that these routine disclosures are not adequate for nutrition research. Superficial transparency is not good enough, they say.

Ioannidis and Trepanowski argue for much broader disclosures in nutrition research because it is unique. Personal biases can influence nutrition research because of the intense intense social, cultural, and religious significance of food. On top of that, drawing conclusions about cause and effect is notoriously difficult in nutrition. Thus they say:

When the data are not clear, opinions and conflicts of interest both financial and nonfinancial may influence research articles, editorials, guidelines, and laws.

So they propose expanded disclosures in four areas.

Financial Conflicts Beyond Industry

Because of intense public interest in nutrition, best-selling books on the subject are quite common. Scientific research credentials can enhance the odds for a title’s success. Bold claims in such books can go well beyond the weight of scientific evidence.

Likewise, nonprofit organizations need public visibility and donor support to thrive. Universities have big PR operations to burnish their images, which in turn helps them raise money. Pumping out puffery about cinnamon and weight loss, for example, recently earned lots of press for researchers at the University of Michigan.

Allegiance Bias and Favored Theories

Persisting with difficult research is more likely when researchers believe in what they’re doing. Some researchers have become convinced in the harm that sugar causes, even calling it “toxic.” Such convictions sometimes prove to be true, as in the case of trans fats. Sometimes they do not, as in the case of MSG.

Personal convictions should certainly not be disqualifying. But they are relevant as potential sources of bias.

Dietary Commitments

“Eat your own cooking.” This business aphorism has relevance for nutrition research. People who are highly committed to particular dietary practices – e.g. veganism – may be motivated to conduct research on it. Again, such commitment should be a relevant disclosure.

Advocacy and Activism

ConscienHealth takes pride in advocacy and activism. And it’s usually not a secret when someone is a vigorous advocate. But in the context of disclosures, advocacy commitments are relevant, say Ioannidis and Trepanowski. In fact, they point to an inherent conflict between advocacy and the need for objectivity in the scientific method.

In sum, we agree with what Ioannidis and Trepanowski are recommending. Balancing conciseness with completeness is a challenge. Burying readers in voluminous disclosures will not help. “No relevant disclosures” is a frequent form of disclosure that is even less helpful.

Humans are inherently biased creatures.

Click here for the viewpoint in JAMA and here for a narrower perspective from Vox.

Not Quite Transparent, photograph © PROTom Wachtel / flickr

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.


December 12, 2017

One Response to “Superficial Transparency in Nutrition Research”

  1. December 13, 2017 at 2:19 pm, Allen Browne said:

    I think we should start with “Humans are inherently biased creatures” and go on from there.