Five Overlooked Sources of Bias
Martin Binks reminds us that many overlooked sources of bias get pushed out of view by a singular focus on bias associated with commercial financial interests. His commentary has just been published online in the International Journal of Obesity.
In another recent commentary — this one in the Journal of the Royal Society of Medicine — Richard Smith and colleagues tell us that bias comes in a thousand forms. Prodded by Binks, Smith, and their colleagues, we offer you five sources of bias worth your attention.
- Noncommercial organizations have their own biases regarding what they will and will not support and publish. Whereas commercial organizations generally have a straightforward profit motive, nonprofit organizations have agendas of their own that can be less obvious. For example, many are convinced that the Academy of Nutrition and Dietetics is biased by corporate funding. They are seriously mistaken. The Academy’s overwhelming bias is for the interests of the Registered Dietitian.
- White hat bias hides in plain sight. Experts with a righteous cause will question financial interests of those who disagree with them, but seldom acknowledge their own competing interests.
- Professional status is another source of obvious bias, routinely overlooked. Smith et al cite the example of the BMJ, which bars an employee of a commercial enterprise from writing in the journal on health policy issues. At the same time, general practitioners — who profit from their practices under the status quo — routinely write editorials for the journal.
- Publication imperatives shape the research literature in ways that are commonly underestimated. Increasing attention to publication bias is shaping the outcomes of rigorous evidence reviews and meta-analyses. But outside this very small sliver of the literature, this is not the case. Positive results are more likely to be published, published rapidly, published in English, published more than once, and cited by others. People who don’t like evidence reviews that challenge entrenched biases can and do fall back on positive studies to support their bias. In the case of the relationship between breastfeeding and obesity, publication bias has led many — even senior CDC officials — to believe that causality has been proven. Careful evidence review shows that this is not the case and that any effect, if it ever is proven, is likely to be small. Breastfeeding offers plenty of benefits, but obesity prevention is definitely not a biggie on the list.
- Policy agendas can bias the interpretation of research, especially in governmental and nonprofit institutions. A recent example occurred in dissemination of a recent publication in JAMA by CDC scientists about U.S. obesity trends. The peer-reviewed publication very clearly concluded that obesity rates were unchanged over the last decade. But in a press release, CDC leadership plucked a number from a subgroup analysis of toddlers and provided a headline to health reporters declaring that childhood obesity (in children 2-5) had dropped 43%. These headlines were nicely synergistic with the 4-year anniversary of the administration’s Let’s Move! initiative being celebrated in the same news cycle. Even though the statistics were insignificant, the public was delivered hundreds of millions of impressions for the message that childhood obesity is dropping like a rock.
Binks offers up the best antidote for all of these sources of bias — scientific rigor. He concludes:
We all need to base our evaluations on scientific data and not on predetermined opinions rooted in our own emotion-laden bias for or against specific funding sources.
“Two quite opposite qualities equally bias our minds – habits and novelty.” — Jean de la Bruyere
You can find the Binks editorial here and the commentary by Smith et al here.
The Invisible Bike, photograph © Aldo Cavini Benedetti / flickr
Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.
