White hat bias will make you fat … or not

If you’ve ever found yourself questioning media reports that display evidence of medical researchers reinforcing each others’ work in an unscientific manner, then your feelings may be based on more than a hunch.

Take obesity, for example. Your powers of observation should confirm that we are, by and large, getting fatter, exercising less and not paying proper attention to our diets. And the evidence provided by official statistics is quite clear in this respect. Obesity is a growing problem.

But what about research which relates specific foods and drinks with weight loss and gain? In a paper published today in the International Journal of Obesity, Mark Cope and David Allison argue that discussions of obesity research may be misrepresented by other scientists because their bias distorts the original findings. The Birmingham, Alabama-based medical statisticians examine the role of predefined bias in influencing how research literature is cited by other scientists, and can lead to miscommunication in press releases and subsequent media coverage.

Referring to this misrepresentation of research literature as “white hat bias”, Cope and Allen look in detail at two papers which report the effects on weight of drinks known as “nutritively-sweetened beverages” (NSBs). Both of these original papers are said to report results that are inconclusive, and could be interpreted as meaning that NSBs have either no effect on, or only marginally decrease, their consumers’ weight.

Looking at the problem in broader scope, Cope and Allison judged that less than one-third of the papers which cited these two studies accurately reported the overall finding (of a null result), and that over two-thirds displayed white hat bias by supporting the idea that NSBs increase obesity.

White hat bias is defined by the authors as “bias leading to distortion of information in the service of what may be perceived to be righteous ends”. Their study also challenges claims which suggest that industry-funded papers are biased towards particular products.

Obesity has come to be demonised, say the researchers:

“Certain postulated causes [of obesity] have come to be demonized (for example, fast food, NSBs, formula feeding of infants) and certain postulated palliatives (for example, consumption of fruits and vegetables, building of sidewalks and walking trails) seem to have been sanctified. Such demonization and sanctification may come at a cost. Such casting may ignite feelings of righteous zeal.”

Of course, for policymakers, sending out the right message is often more important than credible scientific evidence. Look at the debate surrounding recreational drugs, for example. And with the spinning of lobby groups, and sensation-seeking journalists regurgitating dodgy press releases, it’s no wonder that white hat bias is amplified in the media. As Cope and Allison conclude, demonising bad stuff can be helpful when it comes to spurring positive social action, but it may be more harmful in the long run if it undermines scientific integrity.

Further reading

MB Cope & DB Allison, “White hat bias: examples of its presence in obesity research and a call for renewed commitment to faithfulness in research reporting”, Int. J. Obesity (2009)