Industry-funded soda studies overwhelmingly downplay health risks

If you want to feel better about snacking on some Halloween treats, you could look to one of a handful of studies showing no connection between obesity and the consumption of sugary drinks and snacks.

There’s just one problem — those studies are likely to have been funded by food companies.

In fact, a new University of California study shows, studies with financial ties to sugary drinks are significantly less likely to find a link between those products and obesity or diabetes than are independent studies.

The result, said the study’s author, is that the idea of sugar consumption causing obesity is often reported as controversial, rather than definitive.

“That controversy has been manufactured by the industry itself, which is harnessing the science to its own ends,” said Dr. Dean Schillinger, professor at the University of California San Francisco.

For the latest study, Schillinger and colleagues combed PubMed, an online listing of scientific research, to find 60 experimental studies that examined the effects of sugar-sweetened beverages on metabolic outcomes like obesity and diabetes.

Thirty-four of those studies, a slight majority, found an association between sugary beverages and obesity; 26 found no association. But the latter group was entirely made up of industry-funded studies, while only one (2.9 percent) of the positive studies had ties to industry.

The funding — from companies like PepsiCo and Nestle — took the form of either direct financial support of the research, or the payment of authors as consultants. The results were published Monday in Annals of Internal Medicine.

“It’s interesting that they looked at experimental studies,” said Jennifer Harris, a social psychologist at the University of Connecticut Rudd Center for Food Policy and Obesity, who wasn’t involved in the study. “[They are] designed to show a causation effect. If you take out the studies that were industry-funded, there is no controversy. That makes this an important study.”

Adela Hruby, a Tufts University nutritional epidemiologist not involved in the study, is interested in what is not in it.

“Why only PubMed?” she asked. “And why are they limited to publications between 2001 and 2016?”

Hruby also suggested that publication bias could have affected either the current study, or the studies it studied. “People tend to want to publish a result, rather than a non-result. And the researchers could also have had their own biases.”

Schillinger’s research was funded by the National Institutes of Health. But he also discloses having been a paid expert for San Francisco County’s ordinance regarding billboard advertising for sugar-sweetened beverages.

But, Hruby said, the current study’s credibility is bolstered by a similar one published in 2013. “I don’t think this study is hugely novel, but it does show our susceptibility to bias.”

That susceptibility needs a closer look, said Schillinger, to determine how and when biased results occur.

“What was it about the conduct of the scientists that introduced the bias?” he wondered. “We need to pay attention to conflicts of interest. And if we’re going to use science to influence consumers’ health decisions, we have to have them think more critically by asking who funded the study.”