“Teen’s junk food diet caused boy to go blind.” “Even one diet soda a day may triple your risk for developing dementia.” “Dark chocolate may help prevent some forms of cancer.” These are just a few of the health news headlines (or click bait?) that we consume on a daily basis. But they rarely reflect the actual or nuanced research findings.
Indeed, health news reporters and editors must exercise precision and consideration of public interest when constructing news headlines. Yet, too frequently they (and we) fall into the fallacy of equating correlation with causation. A deeper dive into the first news story above (“Teen’s junk food diet caused boy to go blind”) uncovers the subtle yet distinct difference between “causing blindness” (the news headline) and “can lead to blindness” (language from the actual study). Additionally, these do not equate to “will lead to blindness.”
But health news reporters are not entirely at fault here. The burden is also on the shoulders of those of us who conduct research. In this case, the study is published in the Annals of Internal Medicine and titled: “Blindness Caused by a Junk Food Diet.” The researchers state as their objective “To alert clinicians of the visual complications of a diet restricted to junk food.” Which begs the question, “Is blindness the same as visual complication?” And were the pre-existing vitamin deficiencies in the 14-year-old patient entirely the result of consuming junk food or were there other health factors involved?
The nature of empirical research is complex in terms of presenting findings and then interpreting the significance of those findings. If the researchers themselves are nuanced (or unclear) in the findings and/or the implications of those research findings, then we can’t expect fidelity or additional truth when the research is translated into broadcast media messages for mass public consumption. Or can we?
Unfortunately, the general public cannot readily find the answers to these deeper questions, as this medical journal and many others are restricted by a pay wall (full text subscription only). Thus, we are left with the path of least resistance: Translation by Fox News and other popular media outlets.
Even when we are able to digitally access the full text of a research study, our search for truth can be tricky. Take a closer look at this “special article” (read: a non-scientific study bordering on an editorial) published in the reputable The New England Journal of Medicine. “Myths, presumptions, and facts about obesity” was a collaboration of twenty doctors/doctoral/medical/health professionals with impressive educational credentials. Searching the Internet, the team “identified, reviewed, and classified obesity-related myths and presumptions.” They also “examined facts that are well supported by evidence, with an emphasis on those that have practical implications for public health, policy or clinical recommendations.” Sounds rigorous, right? The conclusions of the study are that “false and scientifically unsupported beliefs about obesity are pervasive in both scientific literature and the popular press.” Nothing too surprising or controversial, right? But look deeper.
The conclusions of the study are tagged on the first page with the single disclosure: (Funded by the National Institutes of Health). This lends credibility since the NIH (part of the U.S. Department of Health and Human Services) is the medical research agency of United States’ “making important discoveries that improve health and save lives.” Not so fast. If we skip to the very end of the research publication (before the References list), we can see (in much smaller, barely legible font) two columns of disclosures of sponsorships and potential conflicts of interest. This is the most important part of the study in being able to interpret its value and significance to the general public. The bias of the funders are inextricably connected to the purpose, findings and outcomes of the study regardless of the disclosure statement. In other words, simply disclosing potential conflicts of interest does not diminish the conflict.
Here are the various sources of funding (both indirect and indirect) and disclosures for potential conflicts of interest for this “special article” presented in the format of a research study:
- Global Dairy Platform
- Kraft Foods
- Knowledge Institute for Beer
- Arena Pharmaceuticals
- Basic Research
- Novo Nordisk
- Danish Brewers Association
- Danish Dairy Association
- International Dairy Foundation
- European Dairy Foundation
- Coca-Cola Foundation
- General Mills Bell Institute of Health and Nutrition
- Jenny Craig
- Ulmer and Berne
- National Cattlemen’s Association
- Mead Johnson Nutrition
- Frontiers Foundation
- Orexigen Therapeutics
- Jason Pharmaceuticals
- Almond Board of California
- United Soybean Board
- Northarvest Bean Growers Association
- McNeil Nutritionals
- Red Bull
- World Sugar Research Organisation
- Archer Daniel’s Midland
- Eli Lilly and Company
Armed with this knowledge above, we read the research findings in a different light. Table 3 (page 451) in the study. lists “Facts about Obesity” defined as such because there exists “sufficient evidence to consider them empirically proved.” Here are some examples listed:
- “Some pharmaceutical agents can help patients achieve clinically meaningful weight loss and maintain the reduction as long as the agents continue to be use.”
- “Provision of meals and use of meal-replacement products promote greater weight loss.”
- “In appropriate patients, bariatric surgery results in long-term weight loss and reductions in the rate of incident diabetes and mortality.”
One can’t help but connect the dots between the sanctioned “facts” and the organizations listed in the disclosure above. The numerous pharmaceutical companies and Jenny Craig, for starters, have a vested interest in designating these propositions as facts, don’t they?
We could play the same “connect the dots” game with the “myths” and “presumptions” listed in the study. The researchers define myths as “beliefs held true despite substantial evidence refuting them.” Table 1 on page 448 lists 7 myths, one of which is “Breast-feeding is protective against obesity.” Might the various dairy councils listed in the disclosure above have a vested interest in refuting this myth to increase the consumption of other milk forms?
The researchers define “Presumptions about Obesity” as “unproved yet commonly espoused propositions.” They include
- “Regularly eating (vs. skipping) breakfast is protective against obesity.”
- “Snacking contributes to weight gain and obesity.”
Which of the affiliates, sponsors, or funders benefit the most from circulating these “unproved yet commonly espoused” propositions above? Might General Mills (the manufacturer of processed breakfast food products) have a vested interest in promoting eating breakfast regularly (confirming the presumption)? Might any number of the organizations listed above have a vested interest in disconfirming the presumption that snacking contributes to weight gain and obesity? That’s the thing about a presumption—it is neither confirmed proved nor disproved and thus open to interpretation. Which is ideal for advertising and marketing, isn’t it?
In some cases, the language of health news is literally a matter of life and death. Like in the recent health news reporting of deaths related to vaping. What do we do when there does not even yet exist research findings (let alone findings that can be influenced by corporations)? We demand for more research that is not strangled by the tentacles of corporate sponsorship. And hold corporations accountable for their role in perpetuating fake (health) news.
In the mean time, we as health information-seekers must do more on our own be more health media literate. Here is a list of health media literacy resources compiled by Denise O’Shea, Head of Access Services, Sprague Library, Montclair State University.
Here are also some fact-checking services that may help in disputing the health claims that circulate around social media.