1 March 2018

FOOD FOR THOUGHT

ANYTHING TO DECLARE? 
Dr Sharma

Conflict of interest is about much more than money says Dr Arya Sharma. In a blog post in Obesity Notes (a shortened version reprinted here with permission) he highlights the conflicts and potential biases nutrition researchers should not hesitate to acknowledge and disclose that John Ionnadis and John Trepanowski raise in JAMA.

First, the authors focus on nutrition research for good reasons. They say: “The totality of an individual’s diet has important effects on health, [while] most nutrients and foods individually have ambiguously tiny (or nonexistent) effects. Substantial reliance on observational data for which causal inference is notoriously difficult also limits the clarifying ability of nutrition science. When the data are not clear, opinions and conflicts of interest both financial and nonfinancial may influence research articles, editorials, guidelines, and laws. Therefore, disclosure policies are an important safeguard to help identify potential bias.”

While the potential for financial conflict in relationship to the food industry is well recognised and there are now well-established disclosure norms, other conflicts, of which there are many, are not routinely acknowledged, let alone, disclosed. For one, there are significant financial conflicts that have nothing to do with taking money from industry. For example: “Many nutrition scientists and experts write books about their opinions and diet preferences. Given the interest of the public in this topic, books about nutrition, diets, and weight loss often appear on best-selling lists, even though most offer little to no evidence to support their frequently bold claims.”

DIET BOOKS

Furthermore: “Financial conflicts of interest can also appear in unexpected places. For example, many not-for-profit nutrition initiatives require considerable donor money to stay solvent. Public visibility through the scientific literature and its reverberation through press releases, other media coverage, and social media magnification can be critical in this regard.”

But conflicts can get even more complicated when they start reflecting researchers’ own personal views and biases: “Allegiance bias and preference for favorite theories are prevalent across science and can affect any field of study. It is almost unavoidable that a scientist eventually will form some opinion that goes beyond the data, and they should. Scientists are likely to defend their work, their own discoveries, and the theories that they proposed or espoused.” While that is certainly true for any area of research, nutrition scientists face an additional challenge.

“Every day they must make numerous choices about what to eat while not allowing those choices to affect their research. Most of them also have been exposed to various dietary norms from their family, culture, or religion. These norms can sometimes be intertwined with core values, absolutist metaphysical beliefs, or both. For instance, could an author who is strongly adherent to some religion conclude that a diet-related prescription of his or her religion is so unhealthy as not to be worthwhile?”

The authors propose that nutrition researchers: “disclose their advocacy or activist work as well as their dietary preferences if any are relevant to what is presented and discussed in their articles. This is even more important for dietary preferences that are specific, circumscribed, and adhered to strongly. For example, readers should know if an author is strongly adherent to a vegan diet, the Atkins diet, a gluten-free diet, a high animal protein diet, specific brands of supplements, and so forth if these dietary choices are discussed in an article. The types of articles in which relevant disclosure should be expected include original research, reviews, and opinion pieces (such as editorials).” Although the article focuses on nutrition research, the authors acknowledge that similar biases may exist in other areas of research.

In my own experience, ideological biases (although well-intended) are pervasive through much of the research and publications on topics ranging from physical activity to public health, where I often see strong recommendations made based on evidence that is not even remotely as robust or rigorous as the evidence that comes from, say a large randomised clinical trials of a new prescription drug.

I certainly agree with the authors’ recommendation that: “As a general rule, if an author’s living example could be reasonably expected to influence how some readers perceive an article, disclosure should be encouraged. Authors who have strong beliefs and make highly committed choices for diet or other behaviors should not hesitate to disclose them. Doing so may help everyone understand who is promoting what and why.”

Dr Arya M. Sharma, MD/PhD, FRCPC is Professor of Medicine & Chair in Obesity Research and Management at the University of Alberta, Edmonton, Canada. He is also the Clinical Co-Chair of the Alberta Health Services Obesity Program.

Read more:
• John Ionnadis and John Trepanowski: Disclosures in Nutrition Research: Why It Is Different.
• Dr Arya Sharma: Conflict Disclosures in Nutrition Research