Nutrition Studies Skewed by Industry Dollars?

For

anyone who tries to keep track of which foods provide which health benefits, life seemed a little more complicated this morning. The first major analysis of nutritional research found the science to be every bit as susceptible to sponsor bias as pharmaceuticals. In a paper published online Tuesday in PloS Medicine researchers from Children's Hospital Boston report that when studies linking beverages to health are funded entirely by industry, the conclusions are four to eight times more likely to support the sponsor's commercial interest than studies with no industry funding. And the implications of the findings, says senior author David Ludwig, are far-reaching. "Whereas conflicts of interest in pharmaceuticals could affect the millions of people taking drugs," he says, "conflicts of interest in nutrition could affect everybody — because everybody eats."

The research analyzes 206 studies on the health effects of juice, milk, and soft drinks — all of them published over the five-year period ending Dec. 31, 2003, and archived in the National Library of Medicine's online database, Medline. Ludwig says his team focused on beverages because they provided a discreet, easily analyzed sample to test the hunch that nutritional science might be skewed by industry dollars. (More studies would be needed to assess the impact of sponsorship on food research, which the study did not address.)

The study doesn't suggest that industry scientists are warping their studies deliberately or fudging numbers. But Ludwig says the scientific literature can become skewed overall toward commercial ends when certain hypotheses —those favorable to industry — are investigated more often than others, and when scientists only publish their results selectively. Industry scientists say that's nonsense. Greg Miller, executive VP of innovation for National Dairy Council, says sheer volume of results doesn't mean much. "As a scientist I'm more interested in the quality of the research," he says. And the PloS Medicine study looks only at the absolute number of favorable vs. unfavorable results, not at the nature of the research undertaken or at how much attention each publication received. "We [at National Dairy Council] bring good science to the table," Miller says. "If we don't, we won't get a seat at the table."

Major medical journals do publish industry-funded science as long as it meets the publications' quality standards. Those standards require scientists to disclose any potential conflicts of interest. But compliance is generally based on the honor system, leaving scientists to interpret the journal policies themselves. (The PloS Medicine paper noted a sharp uptick in the number of articles citing potential conflicts over the five-year study period — a trend Ludwig attributes to more stringent journal standards and better self-policing among scientists.)

No one ever died because he mixed two kinds of juice and got behind the wheel, of course. But Ludwig, who is director of the Optimal Weight for Life program at Children's Hospital Boston, worries that misleading nutritional data in a society obsessed with food, health, and weight can confuse the public and lead to poor public-health policy. He wants government to step up independent funding for nutritional science — and for consumers to know they'll have to work to keep a study's results in perspective when industry funding has been disclosed. "It might say, buyer beware. But then the buyer still needs to beware."